The Oxford Handbook of Industrial Hubs and Economic Development 9780198850434, 0198850433
Industrialization supported by industrial hubs has been widely associated with structural transformation and catch-up. B
121 13 13MB
English Pages 1226 Year 2020
Cover
INDUSTRIAL HUBS AND ECONOMIC DEVELOPMENT
Copyright
Acknowledgements
Contents
List of Figures and Maps
Figures
Maps
List of Tables
List of Contributors
PART I: CONCEPTS AND METHODOLOGIES
Chapter 1: Industrial Hubs and Economic Development: An Introduction
1.1 Background to Industrial Hubs and Economic Development
1.2 Preparing the Handbook
1.2.1 Aims and Approaches
1.2.2 Preparation of the Handbook
1.2.3 Readership of the Handbook
1.3 Themes and Perspectives
1.3.1 The Concept of Industrial Hubs
1.3.2 Complexities and Connections between Industrial Hubs and Economic Development
1.3.3 Divergence in Development Perspectives and Variations in Empirical Evidence
1.3.4 Industrial Hubs and Implications for Latecomers
1.4 Organization of the Handbook
1.5 Conclusions
Acknowledgemen
References
Chapter 2: Industrial Hubs and Economic Development: A Literature Review
2.1 Background and Introduction to Literature Review of Industrial Hubs
2.1.1 Focus and Approach of Literature Review
2.2 Agglomeration Economies and Cluster Dynamics
2.2.1 Externalities and the Marshallian Trinity
2.2.1.1 Internal and external economies of scale
2.2.1.2 Marshallian localization economies
2.2.2 Industrial Clusters and Firm Competitiveness
2.2.3 Productive Cities and Industrial Hubs
2.3 Structural Transformation, Catch-up, and Industrial Policy
2.3.1 Structural Transformation as the Essence of Economic Development
2.3.1.1 Manufacturing as an engine of growth and structural change
2.3.1.2 The strategic role of exports
2.3.2 The Latecomer Advantage and Catch-up
2.3.3 Industrial Policies as Prime Vehicle of Transformation and Catch-up
2.4 Empirical Perspectives on Industrial Hubs and Recent Literature
2.4.1 The Industrial District Literature (1985 to 2000s)
2.4.2 Lessons from the Practice of Industrial Hubs
2.4.2.1 Concept of industrial hubs
2.4.2.2 The large wave of industrial hubs in Asia
2.5 Industrial Hubs and Economic Development: A Synthesis and Conclusions
2.5.1 Summary of Literature Review
2.5.2 Innovation and Technological Capabilities
2.5.3 Strategies of National Economies and Firms
Acknowledgements
References
Chapter 3: Heterodox Approaches to Industrial Policy and the Implications for Industrial Hubs
3.1 Introduction
3.2 Industrial Policy in History and Theory
3.3 Overview of Structuralist Development Economics and Structural Transformation
3.4 The Challenge of Premature Deindustrialization
3.5 From Assembly Plants to Plant Assembly: Classification and Structural Change
3.6 A Structuralist Approach to Hubs
3.7 Hubs and Structural Transformation: Difficulties and Policies
References
Chapter 4: The Economics of Innovation behind Cluster Dynamic Processes
4.1 Introduction
4.2 Cluster Dynamics in a Circular Flow Schema
4.3 Marshall and Penrose: Linking Intra-Firm and Inter-Firm Growth Dynamics
4.4 Kuznets, Penrose, and Sector Transitions
4.5 Veblen and Young: Cumulative Increasing Returns
4.6 Jane Jacobs and Marshall’s Large Cities
4.7 Industrial Districts and Rapid Growth in China
4.8 Policy Frameworks: The Capability Triad and Extra-firm Infrastructures
References
Chapter 5: Local Ecosystems and Social Conditions of Innovative Enterprise
5.1 Theoretical Approaches to the Firm–Locality Nexus
5.2 Lancashire Cotton and the ‘Workshop of the World’
5.3 The ‘Third Italy’ and the Emilia Romagna Industrial Ecosystem
5.4 ‘Silicon Valley’ as the Epicentre of a Global Technological Revolution
5.5 General Lessons about Economic Development from the Iconic Industrial Districts
5.5.1 Lessons from Lancashire
5.5.2 Lessons from the Third Italy and Emilia Romagna
5.5.3 Lessons from Silicon Valley
References
Chapter 6: Industrial Policy, Institutional Transformation, and the Development of Industrial Parks
6.1 Introduction
6.2 Neoclassical Economic Approaches to Industrial Parks
6.3 Neoclassical Economics and the Textile Parks of India: An Exploration
6.4 Alternative Theoretical Constructs for Building Industrial Policy and Parks
6.4.1 Structuralism
6.4.2 The Developmental State
6.4.3 The Capabilities Approach
6.4.4 Organizations and the Institutional Approach to Industrial Policy
6.5 Applications to Industrial Policy and Industrial Parks: Key Elements
6.6 Conclusions
References
Chapter 7: Industrial Hubs: The Viewpoints of Economic Geography and Empirical Economics
7.1 Introduction
7.2 Economic Geography and Spatial Economics Theories
7.2.1 The Driving Force of Agglomeration
7.2.2 Economic Development and Industrial Clustering
7.2.3 History and Expectations
7.3 Methodological Issues in Empirical Studies on Policy Evaluation
7.3.1 Causal Inference Framework
7.3.2 Selection Bias
7.3.3 Relocation and Spillover
7.3.4 Sample Selection Bias
7.3.5 External Validity
7.3.6 Reduced-form Approach and Structural Approach
7.4 Empirical Evidence on Economic Agglomeration and Policy Effectiveness
7.4.1 Agglomeration Economies
7.4.2 Industrial Zones
7.4.3 Infrastructure
7.5 Conclusion
References
Chapter 8: Industrial Hubs, Urban Systems, and Economic Development
8.1 Introduction
8.2 Industrialization and Urbanization: A Historical Perspective
8.3 Global Urban Dynamics
8.4 Urban Scale Economies and Productivity
8.5 Urban Systems: Theoretical Underpinnings
8.6 Industrial Location and the Urban System
8.7 Urban Diseconomies and Industrial Activity
8.8 The Global Urban System and Industrial Hubs
8.9 Enabling the Urban-industrial Economy
8.10 Concluding Remarks
References
PART II: HISTORICAL CONTEXT AND ANALYTICAL THEMES
Chapter 9: Industrialization, Hubs, and Catch-up: The World Economy in Historical Perspective
9.1 Industrialization and Industrial Hubs
9.2 The Era of Dramatic Changes: 1750–1913
9.3 The Decades of Continuity: 1913–73
9.4 Beginnings of Transformative Change: 1970–2017
9.5 Conclusion
References
Chapter 10: Industrial Districts in Europe: Lessons for Industrial Hubs and Industrialization
10.1 The Emergence of the Question
10.2 The Dynamics of Industrial Districts in Europe
10.2.1 The Success Stories of Southern European Industrial Districts
10.2.2 The Different ID Dynamics in Northern and Southern European Countries
10.2.3 The Introduction of Industrial Districts into the Economic Debate in Europe
10.3 The Industrial District Model
10.3.1 The Working Mechanisms
10.3.2 The Dynamic Model
10.3.3 Social Capability and Some Preliminary Conclusions
10.4 The Rise and Decline of Industrial Districts in Europe
10.4.1 The Golden Age of Industrial Districts
10.4.2 Transformation versus Crisis in Industrial District Organization
10.4.3 Learning and Innovation
10.4.4 Lessons from the Historical Perspective: The Risks of Collapse of the Model
10.5 The Forces of Globalization in the International Market: Impact on Industrial Districts
10.5.1 The International Context
10.5.2 Typologies of Transformation in Industrial Districts
10.6 Questions for Today: The Impact of the Economic Crisis and the New Internal Restructuring in Industrial Districts
10.6.1 The Governance of Transformation Processes
10.6.2 Collective Actions for the Reproduction of External Economies: Experiences in European Countries
10.7 Lessons for Emerging Countries and New Industrial Regions: Some Final Conclusions
References
Chapter 11: The Political Economy of Special Economic Zones: Pasts, Presents, Futures
11.1 Where is the Juncture? Industrialization-driven Development, SEZs, and Industrial Hubs
11.2 Free Trade or Industrial Districts? Predecessors of Special Economic Zones
11.3 Legacies of Imperialism? The Global Spread of Special Economic Zones
11.4 From Zones to Hubs? Learning to Curb Runaway Capital in SEZs
References
Chapter 12: Cluster Dynamics and Regional Networks: New Argonauts, Silicon Valley, and Route 128
12.1 Introduction
12.2 The New Argonauts and the Rise of Cross-regional Entrepreneurship
12.2.1 From Brain Drain to Brain Circulation
12.2.2 Entrepreneurship in the Periphery
12.3 Firm Dynamics: Cross-Regional Start-ups and Talent Gain
12.3.1 Looking Ahead
12.3.2 A Model for Others?
12.4 Agglomeration Economies and Industrial Clusters
12.5 A Network Approach to Regions and Industrial Hubs
12.5.1 Regional Networks and Shifting Industrial Systems
12.6 Synergy Between Regional Networks and Firm Production Systems
12.6.1 Start-ups and Spin-out Firms
12.6.2 Big Businesses and Cluster Dynamics
12.7 Conclusions
References
Chapter 13: Agglomeration of European Industries
13.1 Introduction
13.2 Agglomeration of European Industries: The Focus of this Chapter
13.3 External Imbalances in the European Union and the Role of Industrial Activity
13.4 The Agglomeration of IndustrialActivity in the ‘Central European Manufacturing Core’
13.5 The Role of Cross-border Production Networks and Patterns of Vertical Specialization
13.6 Conclusions
References
Chapter 14: The Global Experience in Special Economic Zones: A History and Review
14.1 Introduction
14.2 Definitions, Objectives, and Forms of SEZs
14.3 The Evolution of SEZs as Tools of Industrialization
14.4 Some Key Lessons Learned from SEZ Experiences
14.4.1 Sustained Political Support and Strategic Integration
14.4.2 Leveraging Comparative Advantage
14.4.3 Getting the Location Right
14.4.4 Identifying and Resolving Barriers to Investment in the Zones
14.4.5 Leveraging the Expertise of the Private Sector, but Avoiding Rigid Ideological Approaches
14.4.6 Getting Incentives Right
14.4.7 Delivering Quality Infrastructure
14.4.8 Connecting to the Local Economy
14.5 Conclusions
References
Chapter 15: Leapfrogging through Smart Manufacturing: The Role of Hubs in Industry 4.0
15.1 Introduction
15.2 Industrial Hubs: Looking Back, Drawing Lessons
15.3 Can Smart Manufacturing be a Leapfrogging Strategy?
15.3.1 Smart Manufacturing Explained
15.3.2 The Incremental and Modular Nature of Data-driven Innovation
15.4 Smart Manufacturing Strategies for Africa
15.4.1 Moving from Analogue to Digital/Smart Manufacturing
15.4.2 Some Lessons from Germany’s Industry 4.0 Framework
15.5 Drawing Lessons andRethinking the Role of Hubs and Policy in Industry 4.0
15.5.1 Starting Out: Integrating ICT Hubs with Manufacturing Potential
15.5.2 The Functions of Cross-sectional Industrial Hubs
15.5.3 Securing a Virtuous Cycle of Value Creation
15.6 Concluding Remarks
References
Chapter 16: The Servitization of Industrial Hubs in the Digital Era
16.1 Introduction
16.2 Megacities Ascending
16.3 Manufacturing Losing Ground
16.4 Can Special Zones Still Incubate Hubs?
16.5 Servitization and Digitization on the March
16.6 Concluding Observations on Servitized Industrial Hubs
References
Chapter 17: An Urban Planning Perspective on Industrial Hubs and Economic Development
17.1 Introduction
17.2 The Role of Cities in Economic Development: Conflicting Theoretical Perspectives
17.3 Urbanization and Industrialization: The African Experience?
17.3.1 Promoting Industrialization through Plan-led Urbanization
17.3.1.1 Cities as development nodes
17.4 Urbanization Trends in Ethiopia: Implications for Industrialization
17.4.1 ‘Vision 2025’ and Ethiopia’s Quest for Industrialization
17.4.2 Industrial Parks as Sub-national Development Nodes
17.4.3 Transforming Cities as Development Nodes and Clusters
17.4.4 Making Planned Urbanization and Industrialization Work: Key Challenges
17.4.5 Weak Policy Formulation and Implementation Capacity
17.4.6 Coordination Challenges
17.4.7 Lack of Coherence between Urban Development Policy and Industrial Policy
17.5 Corridor-based Development and Spatial Development Initiatives
17.5.1 Transboundary City Clusters along Transport Corridors
17.5.2 The Urban Dimension of Macro-regions
17.5.3 The SDI Methodology in Practice
17.5.4 Transforming Resource Corridors into Macro-regional Development Zones
17.6 Conclusion: Mainstreaming Urbanization in National Development Plans
References
Chapter 18: Special Economic Zones and Export-led Growth: An Industrial Policy Imperative
18.1 Introduction
18.2 Economic Zones (EZs) and Special Economic Zones (SEZs)
18.3 Strategic Industrial Policy Considerations
18.3.1 The Ability to Extract the Right Lessons
18.3.2 Static and Dynamic Benefits of SEZs and the Key Role of Productive Capacity
18.4 Industrial Policy Instruments’ Alignment with Industrialization Stage
18.4.1 Early Stages
18.4.2 Middle Stages
18.4.3 Late Stages
18.5 SEZs and the Policy Mix
18.5.1 Early Stages: Foreign Direct Investment (FDI)
18.5.2 Early Stages: Industrial Clustering
18.5.3 Middle Stages: Incubation, Entrepreneurship, and Imitation
18.5.4 Middle Stages: Linkages
18.5.5 Middle Stages: Participating in Global Value Chains (GVCs)
18.5.6 Late Stages: Innovation
18.6 Conclusions
References
Chapter 19: A Flowchart Approach to Industrial Hubs
19.1 Introduction
19.2 Economies of Sequence and the Flowchart Approach
19.2.1 Segments and the Economies of Sequence
19.2.2 The Flowchart Approach
19.2.3 Implications of the New Economic Geography for the Flowchart Approach
19.3 The Prototype of Asian Growth Models
19.3.1 Taiwanese Export Processing Zone as the Prototype of Asian Growth Models
19.3.2 State Economic Development Corporations in Malaysia
19.3.3 The Industrial Estate Authority of Thailand
19.4 Industrial Agglomerations in Northern Vietnam
19.4.1 National Highway Route 5 in 1993 and Haiphong Port in 1995
19.4.2 The Thang Long Industrial Park and Nomura Haiphong Industrial Zone in 1997
19.4.3 Institutional Reform of the Miyazawa Plan in 1999
19.4.4 Living Conditions in 2002
19.4.5 Canon and its Agglomeration in 2002
19.5 The Eastern Seaboard Development Programme in Thailand
19.5.1 Laem Chabang Port in 1984
19.5.2 Tax Incentives
19.5.3 Industrial Zones
19.5.4 Mitsubishi Motors and its Agglomeration in 1992
19.5.5 Industrial Agglomeration
19.6 The Automobile Industry in Guangzhou
19.6.1 The Honda Effect
19.6.2 The Nissan Effect
19.6.3 The Toyota Effect
19.7 The High-tech Industry Agglomeration in the Zhongguancun Science Park
19.7.1 Spin-offs
19.7.2 The Chinese Academy of Sciences (CAS)
19.7.3 The Zhongguancun Science Park
19.7.4 The Haidian Science Park (1992)
19.7.5 The Tsinghua Science Park
19.7.6 Regional Innovation Systems: The Beijing Municipal People’s Government and the Zhongguancun Science Park Management Committee
19.7.6.1 Ordinances of the Zhongguancun Science Park Management Committee (2000)
19.7.6.2 University–industry–government linkages: Innovation by partnerships between firms, universities, and research institutes
19.7.6.2.1 cases (b), (c), (e): company a
19.7.6.2.2 cases (d), (e), (g): tsinghua tongfang
19.7.7 From Patents to Products
19.8 Summary and Conclusions
References
PART III: INDUSTRIAL HUBS, LABOUR, GENDER, AND THE ENVIRONMENT
Chapter 20: Industrial Hubs and the IndustrialLabour Force inAfrica and Asia
20.1 The Expansion of Industrial Hubs and Industrial Employment
20.2 Hubs as Spaces of Labour Force Formation: Between Job Creationand Labour Control
20.3 The Effects of Industrial Hubs on Job Creation and Working Conditions
20.3.1 Employment
20.3.2 Remuneration
20.3.3 Non-wage Working Conditions
20.3.4 Industrial Hubs as Spaces of Gendered Labour Dynamics
20.4 Conclusion: Towards Decent Work in Industrial Hubs
References
Chapter 21: Gender, Industrialization, and Industrial Hubs
21.1 Introduction
21.2 Gender and IndustrialDevelopment
21.2.1 Overview
21.2.2 The Nexus of Gender Inequality and Industrialization: Theoretical and Empirical Linkages
21.3 Women, Gender, and Industrial Hubs
21.3.1 Effects of Hubs on Women and Gender Equality: An Analytic Framework
21.3.2 Effects of Hubs on Women and Gender Equality: Empirical Evidence
21.3.2.1 Employment
21.3.2.2 Wages
21.3.2.3 Working conditions
21.3.2.4 Rights and benefits
21.3.2.5 Empowerment
21.3.2.6 Social impact
21.4 Conclusions
References
Chapter 22: Women, Working Conditions, and Industrial Hubs
22.1 Introduction
22.2 Opportunities and Challenges for Female Employment in Industrial Hubs and GVCs
22.3 Governance of Work to Achieve Gender Equality
22.4 Working Conditions and Women’sEmpowerment in Industrial Hubs: The Business Case
22.5 Conclusions
References
Chapter 23: Special Economic Zones in Latecomer Countries: Time to Bring Environmental Sustainability to the Fore
23.1 Introduction
23.2 Growth, Structural Transformation, and Sustainability
23.3 The Broader Context of the Environmental Imperative
23.3.1 The Climate Change Factor
23.3.2 The Rise of Sustainability Standards
23.3.3 The Evolution of the Development Finance Landscape
23.4 The Strategic Dimension of Environmental Sustainability
23.5 Conclusions
Acknowledgement
References
Chapter 24: The Greening of Industrial Hubs: A Twenty-first-century Development Strategy
24.1 Introduction
24.2 Industrialization in the Twenty-first Century: Scale Effects
24.3 Green vs Black Tendencies:China’s Experience, and Globally
24.4 Global Expansion of Green Platforms and Cost Reduction
24.5 Industrial Hubs: Industrial Dynamics in Suprafirm Structures
24.6 Greening of Industrial Hubs
24.7 Concluding Remarks
References
Chapter 25: Greening Structural Transformation: What Role for Industrial Hubs?
25.1 Introduction
25.2 Greening Economies
25.3 Rethinking Industrial Policy from a Green Transformation Perspective
25.4 Green City Development
25.5 Green Clusters
25.6 Eco-Industrial Parks
25.7 Concluding Remarks
References
Chapter 26: Towards a New Generation of Special Economic Zones: Sustainable and Competitive
26.1 Introduction
26.2 The Universe of SEZs
26.3 The Impact of SEZs
26.3.1 Economic Impact of SEZs
26.3.1.1 Investment attraction
26.3.1.2 Export generation and diversification
26.3.1.3 Employment creation
26.3.1.4 SEZs and GVC participation
26.3.2 Industrial Development Ladder
26.3.3 Social and Environmental Implications
26.3.4 A Development Profit and Loss Statement
26.4 Triple Challenges for SEZs
26.5 Routes to a New Generation of SEZs
26.6 SDG Model Zones
26.7 Conclusion
References
PART IV: POLICY AND PRACTICE EMPIRICAL EVIDENCE FROM ASIA
Chapter 27: Industrial Hubs as Development Incubators: Asian Pioneers
27.1 Asian Pioneers and IndustrialHubs: An Introduction
27.2 East Asian Pioneers of Industrial Hubs: Taiwan, South Korea, and Singapore
27.2.1 Taiwan as a Pioneer of Industrial Hubs
27.2.1.1 The genesis of industrial hubs in Taiwan
27.2.1.2 Industrial deepening and restructuring
27.2.1.3 Technological hubs and technological catch-up
27.2.1.4 The second-mover advantage and its limits
27.2.2 South Korea: Industrial Hubs and Catch-up
27.2.2.1 Industrial hubs and industrial upgrading
27.2.2.2 Industrial hubs and technological catch-up
27.2.2.3 Summary
27.2.3 Singapore: Catch-up and Strategic Approach to Industrial Hubs
27.2.3.1 Industrial hubs and industrial upgrading
27.2.3.2 FDI-driven technological learning
27.2.3.3 Summary
27.3 China: Second Mover and Innovator of Industrial Hubs
27.3.1 The Formative Years (1978–85): Special Economic Zones as Experiment and FDI Vehicle
27.3.2 The Growth Stage (1986–90s): A Rising Global Exporter and Manufacturing Powerhouse
27.3.3 The Long March (1990s–): Technological Catch-up and Industrial Hubs
27.3.4 Industrial Policy Framework and Hubs Development
27.3.4.1 Tied to industrial policy framework
27.3.4.2 Policy and technological learning
27.3.4.3 The middle-income trap
27.3.5 Summary
27.4 A Synthesis of Industrial Hubs and Catch-up in East Asia
27.4.1 A Context: Japan as a Role Model and the External Environment
27.4.2 Strategic Approach and Industrial Policy Framework
27.4.3 Dynamics of Industrial Hubs
27.4.4 The Strategic Role of the State
27.5 Conclusions: Commonality, Unevenness, and Differentiation
27.5.1 Structural Transformation, Catch-up, and Industrial Hubs
27.5.2 Drivers of Technological Catch-up and Industrial Hubs
27.5.3 Policy Innovation and Learning
Acknowledgement
References
Chapter 28: Explaining Reform and Special EconomicZones in China
28.1 Introduction
28.2 The Origin, Varieties, and Evolution of SEZs in China
28.2.1 History
28.2.2 Varieties of Chinese SEZs
28.2.3 Evolution
28.3 SEZs as a Catalyst for the Transition from Planned Economy to Market Economy
28.4 SEZs as the Engine of Structural Transformation
28.4.1 Structural Transformation in China
28.4.2 The Role of SEZs
28.5 Distilling Ingredients of the Success of SEZs from a New Structural Economics Perspective
28.6 Conclusion
References
28.4.1 Structural Transformation in China
Chapter 29: Industrial Hubs in ‘Sphinx’ China
29.1 An Evolutionary Typology of Industrial Parks in China
29.2 Economic and Political Incentives under ‘Sphinx China’
29.3 An Entrepreneurial Analysis of Industrial Hubs
29.4 Fostering Initial Industrial Capacity
29.4.1 Alleviating the Tragedy of the Anti-commons from Rent-extraction by Officials
29.5 Concluding Remarks
Acknowledgement
References
Chapter 30: Innovative Firms and High-tech Industrial Hubs in China
30.1 Introduction
30.2 Beijing
30.3 Shanghai
30.4 Shenzhen
30.5 Lessons from High-tech Industrial Hubs in China
Acknowledgement
References
Chapter 31: Special Economic Zones in China and India: A Comparative Analysis
31.1 Introduction
31.2 Evolution of SEZs
31.3 Institutional Arrangement of SEZs
31.4 Economic Performance
31.5 Conclusion
References
Chapter 32: Taiwan’s Industrial Districts and Economic Development
32.1 Introduction
32.2 Antecedents of Clusters: Rural Industrialization
32.3 Sectoral Comparisons
32.3.1 Machine Tools
32.3.2 Bicycles
32.3.3 Information and Communication Technologies (ICT) Hardware Products
32.3.4 Integrated Circuits
32.4 A Cross-sectoral Comparison of Taiwan’s Clusters: Broader Policy Lessons
References
Chapter 33: Location Choice for Industrial Complexes in South Korea
33.1 Introduction
33.2 Urban Planning and Industrial Hubs
33.2.1 The Interface between Industrial Policy and National Territorial Planning (AC in Figure 33.1)
33.2.2 The Interface between Industrial Policy and Urban Planning (AB in Figure 33.1)
33.2.3 The Interface between Urban Planning and National Territorial Planning (BC in Figure 33.1)
33.3 Principles of Location Choice for Industrial Hubs: An Urban Planning Perspective
33.3.1 Two Variables
33.3.1.1 Industrial infrastructure
33.3.1.2 Living environment
33.3.2 Four Types
33.3.2.1 Industrial annex of metropolis
33.3.2.2 Town of factories
33.3.2.3 New industrial city
33.3.2.4 Metropolitan-immersed industrial hub
33.4 Understanding Korean Industrial Complexes from the Urban Planning Perspective
33.4.1 Guro Export Industrial Complex, an Industrial Annex of Seoul
33.4.2 Gumi Industrial Complex, a Town of Factories
33.4.3 Changwon, a New Industrial City
33.4.4 Guro Digital Industrial Complex, a Metropolitan immersedIndustrial Hub
33.5 Conclusion
References
Chapter 34: A Review of Industrial Clusters, Industrial Policy, and Industrializationin South Korea
34.1 Introduction
34.2 History of Industrial Complex Strategies and the Promotion of Industrial Development
34.2.1 Korean Industrial Complex Concepts
34.2.2 Historical Overview of Industrial Complex Development Strategy
34.2.3 A Timeline of Industrial Location Policies and Industrial Complex Development
34.2.3.1 The 1960s: Supporting light industry exports and establishing the foundations for industrial complexes
34.2.3.2 The 1970s: Building up a heavy chemicals industry and strengthening government leadership
34.2.3.3 The 1980s: Even distribution of industrial complexes for balanced regional growth
34.2.3.4 The 1990s: The creation of technology clusters and a realignment of industrial complex development systems
34.2.3.5 The 2000s: Diversifying industrial locations and sharpening the competitive edge of industrial complexes
34.3 Status and Objectives of CurrentIndustrial Location Policy
34.3.1 Legal Structure
34.3.2 Current State of Industrial Location PolicyAdministration
34.3.3 Converting Industrial Complexes into Industrial Clusters
34.4 Implications of Korean Industrial Location Policy
34.4.1 Strategy: Set Achievable Targets and Adopt an Appropriate Development Strategy
34.4.2 System: Establish Legislative Systems and Relevant Support Plans
34.4.3 Operation: Organize Government-led Provision of Land and Infrastructure Support
34.4.4 Relationships: Mediating Interests and Promoting Balanced Growth
References
Chapter 35: An Evidence-based Analysis of Industrial Hubs: The Singapore Narrative
35.1 Introduction and Overview
35.2 Historical Background of Singapore’s Economic Development
35.3 Building Resilient Industrial Hubs: Getting the Fundamentals Right
35.3.1 Vision and Key Institutions in Developing Industrial Hubs
35.3.2 Preparing the Physical Basic Infrastructure to Kick-start Industrial Hubs
35.3.3 Constant Upgrading of Industrial Hubs to Match Shifts in Global Production Networks
35.4 Six Decades of Industrial Hubs Development
35.4.1 The 1960s: Labour-intensive Industries
35.4.2 The 1970s: Skills-intensive Industries
35.4.3 The 1980s: Capital-intensive Industries
35.4.4 The 1990s: Technology-driven Industries
35.4.5 The 2000s: Knowledge-based Economy
35.4.6 The 2010s: Innovation-oriented Economy
35.5 Expansion and Synergies of Regional Industrial Hubs
35.6 Conclusion and Responding to Future Challenges to Industrial Hubs
Acknowledgement
References
Chapter 36: Industrialization and Industrial Hubs in Malaysia
36.1 Theoretical Considerations
36.2 Phases of Industrialization
36.2.1 Phase I (1958–90)
36.2.1.1 Import substitution
36.2.1.2 Export orientation
36.2.1.3 Heavy industrialization
36.2.1.4 Renewed focus on export orientation
36.2.2 Phase II (since 1991)
36.3 Industrial Hubs
36.3.1 Industrial Zones
36.3.2 Export Processing Zones
36.3.3 Science and Technology Parks
36.3.4 Regional Corridors
36.4 Conclusions
References
Chapter 37: Industrial Policy and the Evolutionof Industrial Hubs in Vietnam
37.1 The Background to Vietnam’s Economic Development Strategy and Industrialization Policy
37.2 The Birth and Evolution of Industrial Hubs in Vietnam
37.2.1 Prior to 1975: Before Unification
37.2.2 From 1986: The Beginning of the Economic Reform Process
37.2.3 From the First Export Processing Zone to the First Industrial Zone
37.2.4 From Border Gate Economic Zones to Coastal Economic Zones
37.2.5 High-tech Parks as a Means to Attract High-tech FDI
37.3 Special Incentives for Industrial Hubs
37.4 Industrial Hub Performance
37.4.1 Industrial Zones and Export Processing Zones
37.4.2 Border Gate Economic Zones
37.4.3 Coastal Economic Zones
37.4.4 High-tech Parks
37.4.5 Special Economic Zones
37.5 Contribution of Industrial Hubs to Vietnam’s Economic Development
37.5.1 Industrial Hubs Help Catalyse Economic Reforms in Vietnam
37.5.2 Encouraging Investment in Infrastructure Development
37.5.3 Encouraging Investment in Potential Industries
37.5.4 Promoting Economic and Labour Restructuring
37.5.5 Upgrading Vietnamese Industries
37.5.6 Promoting Economic Integration in Trade and Investment
37.5.7 Migration, Population Redistribution, Job Creation, and Skills Development
37.5.8 Generating Revenue for the State Budget
37.5.9 Radical Transformation of Urban and Rural Life
37.6 Drawbacks of the Economic Development Strategy and Industrial Hub Model in Vietnam
37.6.1 Expansion: Too Fast and Too Many
37.6.2 Inefficient Land Conversion
37.6.3 Reliance on Fiscal Incentives rather than Competitive Advantage
37.6.4 Low-skilled and Unskilled Labour
37.6.5 Low Technological Level, Limited Technological Transfer
37.6.6 Environmental Pollution
37.6.7 Slow Progress in Human Resource Quality and Science and Technology Capability
37.7 A Political-economy Explanation of Vietnam’s Development of Industrial Hubs
37.8 Lessions Learned from Successful Industrial Hubs in Vietnam
37.8.1 Location is Key to Success
37.8.2 Quality Infrastructure is a Prerequisite
37.8.3 Good Business Environment and Streamlined Administrative Procedures
37.8.4 Selecting a Competent and Resourceful IH Developer
37.8.5 Fiscal Incentives are Important, but Knowing How to Incentivize is Even More Important
37.8.6 Leadership Commitment and Consensus is Important to Produce Policy Consistency
37.9 Conclusion
References
PART V: POLICY AND PRACTICE: EMPIRICAL EVIDENCE FROM LATIN AMERICA
Chapter 38: Industrialization in Latin America: A Critical Review
38.1 Introduction
38.2 The Impact of Productive Development Policies in Latin America
38.3 Structural Change in Latin America: New Wine into Old Wineskins?
38.4 Towards a New Industrial Policy Agenda
38.5 Final Remarks
Acknowledgement
References
Chapter 39: The Development of Knowledge-based Export Clusters: Lessons from Costa Rica and Mexico
39.1 Thirty Years of Productive Transformation
39.2 The Origins of Export-oriented Manufacture in Costa Rica and Mexico
39.3 Background to Economic Transformation
39.3.1 Mexico: The Bumpier Road
39.3.2 Costa Rica: Graduality and Stability
39.4 High-Tech FDI: The Early Years
39.5 Policymaking and Public‒Private Dialogue in Costa Rica and Jalisco
39.5.1 Costa Rica: A Divided Society
39.5.2 Jalisco: Shared Vision, Collaboration, and Public Orientation
39.6 Recent FDI Attraction and Export Performance
39.7 Drivers of Economic Transformation: Some Striking Communalities
39.7.1 Labour Force: Quality and Productivity
39.7.2 Vertical Application of Horizontal Productive Development Instruments
39.7.3 Support of Self-revealing Winners rather than a priori Choice of Winners
39.7.3.1 The role of the private sector
39.7.3.2 The role of national CEOs
39.8 Limiting Factors
39.8.1 The Governance Structure of GVCs
39.8.2 Development of Domestic Suppliers and the SME Myth
39.8.3 Human Resource Development
39.8.4 Urban Development and Economies of Agglomeration
39.9 Conclusions
References
Chapter 40: Modern Industrial Policy in Latin America: Lessons from Cluster Development Policies
40.1 Key Traits of Modern Industrial Policy
40.2 Productive Agglomeration and Coordination Failures
40.2.1. Economies of Geographical Agglomeration
40.2.2 Coordination Failures
40.3 Cluster Development Policies as Examples of ‘Modern’ Industrial Policy
40.4 Cluster Development Policies in Latin America
40.5 An Example: The Tambaú Red Tile Cluster in the State of Sao Paulo, Brazil
40.6 Do Cluster Development Policies Work?
40.7 Conclusions
References
Chapter 41: Latin American Industrialization: Unfulfilled Expectations and Future Opportunities
41.1 The Latin American Industrialization Process
41.1.1 Unfulfilled Expectations and Future Opportunities
41.1.2 Back to Natural Resources, but from a Different Technological, Environmental, and Social Inclusion Perspective
41.2 Case Studies, ‘Appreciative Theorizing’, and a New Way of Approaching the Conceptualization of Industrial Policy Issues
41.2.1 Industry Inception
41.2.2 Industry Growth, Changes in Firm Size and Production Organization, New Institutions, and the Evolving Nature of Industrial Policies
41.2.3 Environmental Mismanagement and the Making of a Sanitary Crisis
41.2.4 Resilience: Response of Firms and Government Agencies to the Sanitary Crisis
41.2.5 Looking to the Future
41.3 Back to the Future: Macro and Micro Issues Underlying the Long-term Latin American Growth Process
References
Chapter 42: Industrial Hubs and Structural Transformation in Latin America: Lessons from Costa Rica
42.1 Introduction
42.2 Industrial Hubs and Structural Transformation: An Analytical Framework
42.3 Industrial Hubs: A Centrepiece of Costa Rica’s New Economic Model
42.3.1 Background: From ISI to a New Economic Model
42.3.2 Industrial Hubs: Policies, Institutions, and Context
42.4 Industrial Hubs and Structural Transformation
42.4.1 Transformation of the Structure of Production and Exports
42.4.2 FTZ Impacts on Employment, Foreign Exchange, and Taxes
42.4.3 Why Costa Rica?
42.4.4 Linkages and Knowledge Spillovers
42.5 An Uneven Playing Field: A Tale of Two Economies
42.6 Lessons from the Costa Rican Experience
Acknowledgements
References
Chapter 43: Industrial Hubs, Industrial Policy, and Economic Development in Mexico
43.1 Introduction
43.2 Industrial Hubs: A Typology Proposal for the Mexican Case
43.2.1 The Industrial Hub Based on the Foreign Maquila
43.2.2 The Industrial Hub Based on the Cluster Model
43.2.3 The Industrial Hub Based on Innovation
43.3 Case Examples of the Three Types of Industrial Hub
43.3.1 Baja California
43.3.2 Querétaro
43.3.3 Nuevo León
43.4 Discussion and Challenges
43.5 General Public PolicyRecommendations
References
Chapter 44: FDI, Industrial Hubs, and Latin American Industrialization:
44.1 Introduction
44.2 Setting the Stage: FDI as a Catalyst for Growth and Development
44.3 Overview of SEZs in LAC
44.4 Performance of SEZs
44.5 SEZs’ Critical Features to Attract FDI
44.5.1 SEZs Must be Embedded in National and/or Regional Development Plans, with Strong Political Support and Clear Objectives, and Aligned with Comparative Advantages
44.5.2 SEZs need to be Anchored in Specialized Regimes Embedded in Clear Legal Frameworks, Supported by Effective Institutions and Subject to Monitoring and Evaluation
44.5.3 Location is Key to Ensuring Exploitation of Advantages in Natural and Economic Geographies, while Potentially Spearheading Cluster Development
44.5.4 High-quality and Cost-effective Infrastructure and Services, as well as Efficient Management of Zones, are a Critical Driver of SEZ Performance
44.6 Global Businesses, the Future of SEZs and Opportunities for Latin American Countries
44.6.1 Emerging Trends
44.6.1.1 New technologies, global value chains, and the attractiveness of investment locations
44.6.1.2 Geopolitics, policy factors, and the push to reconfigure the governance of trade
44.6.1.3 Sustainability
44.6.2 Opportunities for Latin America to Attract Investment
44.7 Conclusions and Future Considerations
Acknowledgement
References
PART VI: POLICY AND PRACTICE: EMPIRICAL EVIDENCE FROM AFRICA
Chapter 45: A Strategic Approach to Industrial Hubs: Learnings in Ethiopia
45.1 Introduction to Learnings in Ethiopia: The Journey and Approach
45.2 Strategic Approach and the Anatomy of Industrial Hubs
45.2.1 Obstacles to Structural Transformation and Policy Learning
45.2.2 Strategies of Industrial Hubs
45.2.2.1 A focus on specialized industrial parks
45.2.2.1.1 Linkage effects
45.2.2.1.2 Industrial clustering
45.2.2.1.3 Skills formation and learning
45.2.2.1.4 Common facilities
45.2.2.1.5 Policy incentives
45.2.2.2 Adhering to sustainability standards and building eco-industrial parks
45.2.3 Financing, Construction, and Operations of Industrial Parks: The Formative Years
45.2.3.1 Private developers
45.2.3.2 Financing
45.2.3.3 Subsidies and incentives
45.2.3.4 Park development
45.2.3.5 Park operations
45.3 Location, Infrastructure, and Urban Ecosystem
45.3.1 Urban Ecosystem and Agglomeration
45.3.2 Location Selection and National Network of Industrial Parks
45.3.3 National Infrastructure Systems
45.4 Strategic Approach to Production Capability and Industrial Hubs
45.4.1 Industrial Policy and Productive Capacity
45.4.2 Production Investment: Manufacturing FDI as a Key Driver
45.4.3 Domestic Capabilities: Domestic Learning and Linkages
45.4.3.1 Building an industrial workforce
45.4.3.2 Building domestic firms and linkages
45.4.4 Anchor Firms and Cluster Dynamism
45.4.5 Network of Learning Institutions and Linkages
45.5 Strategic Approach: Governance and Institutional Framework
45.5.1 Institutionalization of a New Generation of Industrial Hubs
45.5.2 Legislative and Regulatory System: A Key Feature of the Formative Stage
45.5.2.1 Decrees and proclamations
45.5.2.2 Executive Directives of EIB
45.5.3 Instituting Executive Agencies
45.6 A Synthesis of Learning and Policy Innovation
45.6.1 Dynamics of Learning
45.6.2 Policy Dilemmas and Imbalances
45.7 Conclusions
45.7.1 A Summary of Lessons, Progress, and Constraints
45.7.2 Learning Dynamics and Latecomer Advantages
45.7.3 Complex Ecosystem and Web of Institutions
45.7.4 Industrial Policy Framework
Acknowledgements
References
Chapter 46: Hubs Development and Industrial Upgrading in Mauritius
46.1 Introduction
46.2 Theoretical Overview
46.3 Industrial Policy in Mauritius
46.4 The Mauritius Export Processing Zone
46.5 The Gender Dimension of EPZ Employment
46.6 Challenges and Upgrading
46.7 Recent Strategies
46.8 Conclusion
References
Chapter 47: Industrial Hubs and Technology Transfer in Africa’s Apparel Export Sector
47.1 Introduction
47.2 Development of Apparel Export Sectors within sub-Saharan African SEZs
47.3 Explaining the Extent of Technology Transfer
47.3.1 Mauritius
47.3.2 Madagascar
47.3.3 Kenya
47.3.4 Lesotho and Swaziland
47.3.5 Ethiopia
47.4 Conclusion
References
Chapter 48: Chinese Economic and Trade Cooperation Zones in Africa
48.1 Genesis: A New Platform for Attracting FDI and Industrialization
48.2 Successful Zone Development and Structural Transformation
48.3 Delays and Challenges in the China–Africa Cooperation Zones
48.3.1 Infrastructure
48.3.2 Administration and Management
48.3.3 Linkage with Local Industries
48.3.4 Policy and Strategy
48.4 Approaches to Improved Synergism
48.4.1 Chinese Initiatives
48.4.2 African Initiatives
48.5 Conclusion
Acknowledgement
References
Chapter 49: Legal Framework for Industrial Hubs: A Critical Review and Lessons for Africa
49.1 Introduction
49.2 The Theoretical and Legal Architecture of Industrial Hubs
49.2.1 International Sources of Law
49.2.2 Domestic Sources of Law
49.3 Industrial Hubs Law in Comparative Context
49.3.1 China
49.3.2 India
49.3.2.1 Simplified procedures
49.3.3 Republic of South Korea
49.3.3.1 Designation requirements
49.3.3.2 Development project operator
49.3.4 Mauritius
49.3.4.1 Labour standards
49.3.5 Ethiopia
49.3.6 Ukraine
49.4 Comparative Analysis and Lessons for Africa
49.5 Conclusion
References
Chapter 50: Industrialization and Industrial Hubs: Experiences in Kenya, Rwanda, and Tanzania
50.1 Introduction
50.2 The Impact of SEZs: Definitions and Conceptual Issues
50.2.1 What are SEZs?
50.2.2 Measuring the Role of SEZs in Industrialization and Economic Transformation
50.2.3 What Can We Learn from Successful SEZs?
50.2.3.1 Zone design
50.2.3.2 Zone performance
50.2.3.3 Zone linkages and productivity spillovers
50.3 Experiences in African SEZs
50.3.1 General Experiences in African Zones
50.3.2 Kenya
50.3.2.1 Zone design
50.3.2.2 Zone performance
50.3.2.3 Zone linkages and spillover benefits
50.3.3 Rwanda
50.3.3.1 Zone design
50.3.3.2 Zone performance
50.3.3.3 Zone linkages and spillover benefits
50.3.4 Tanzania
50.3.4.1 Zone design
50.3.4.2 Zone performance
50.3.4.3 Zone linkages and spillover benefits
50.4 Lessons from African SEZ Experiences
50.5 Conclusion and Implications
References
Chapter 51: Special Economic Zones in Sub-Saharan Africa: What Drives their Mixed Performance?
51.1 Definitions and Rationale
51.2 SEZs in Africa
51.3 Case Analysis of African SEZs
51.4 The Key Drivers of Successful SEZs
51.4.1 Legal and Regulatory Framework
51.4.2 Government Commitment and Public Services
51.4.3 Zone-level Governance
51.4.4 Industrial Infrastructure
51.4.5 Location and Connectivity
51.4.6 Human Resources
51.4.7 Market Demand
51.5 Key Lessons
Acknowledgements
References
PART VII: SYNTHESIS AND PATHWAYS TO THE FUTURE
Chapter 52: Cluster-based Policies: What Have We Learned?
52.1 Introduction
52.2 The Concept, the Economics, and the Rationale for Cluster Policies
52.3 The Cluster Approach under a New Light: Cluster Policy and Process as ‘Experimentalist Governance’
52.3.1 What is Experimentalist Governance?
52.3.2 How does this Conceptualization of Experimentalism Relate to Cluster Policies and Initiatives as a Type of Productive Development Policy?
52.4 Cluster Development Policies and their Governance: A Review of Two Experiences
52.4.1 Autonomous Community of the Basque Country
52.4.2 Agricultural Machinery Cluster in Argentina
52.5 Concluding Remarks
Acknowledgements
References
Chapter 53: Global Value Chains, Industrial Hubs, and Economic Development in the Twenty-first Century
53.1 Introduction
53.2 The Global Value Chains Framework
53.3 Industrial Hubs from a GVC Perspective
53.4 The Athletic Footwear GVC: From the Asian Production Model to Incipient Automation
53.4.1 The Asian-Based Production Model: Efficient but Slow
53.4.2 Sourcing Shifts: China Down, Vietnam Surging, and Indonesia Emerging
53.5 Adidas: The Need for Speed Drives Factory Automation
53.5.1 Adidas’s Speedfactory—Automation in the Footwear Industry
53.6 Nike: Using Technology to Connect with Consumers
53.6.1 The Nike-Flex Automation Partnership: A Cautionary Tale
53.7 The Electronics GVC: From Factory Cities to Lights-off Factories
53.7.1 Foxconn’s Industrial Hubs in China and Beyond
53.7.2 Foxconn’s Lights-off Factories
53.8 Conclusions
References
Chapter 54: The Impact of Decentralization on Industrial Hubs: Blockchain and the Digital Economy
54.1 Introduction
54.2 Blockchain as a Step in the Digital Technology Revolution
54.2.1 Digital Technology
54.2.2 Blockchain as a Digital Technology
54.2.3 How Policymakers Interpret the Technology–Growth Connection
54.3 Geography: Economics of Industrial Hubs and Global Value Chains
54.3.1 The WTO as a Promoter of Specialization
54.3.2 Fragmentation of GVCs
54.3.3 Blockchain Lacks Territoriality
54.4 Regulation: Blockchain, Hubs, and Multilateral Trade Rules
54.4.1 Data Governance
54.4.2 WTO Digital Provisions
54.4.3 Exceptions to Rules of Origin
54.5 Is Data Creating New Hubs?
54.5.1 Adoption
54.5.2 What Data is Captured?
54.5.3 New Hubs
54.6 Blockchain and Sustainability
54.6.1 Can Hubs Facilitate Sustainable Supply Chains?
54.6.2 The Little Problem of Standards
54.7 Conclusion
References
Chapter 55: The Economics of Agglomeration
55.1 Introduction
55.2 Understanding the Drivers of Economic Agglomeration
55.3 A Blueprint for Creating Agglomeration Economies
55.3.1 Clustering through Special Economic Zones
55.3.1.1 How best to do this?
55.3.2 Skills Enhancement Zones to Sustain Clustering
55.4 Conclusion
References
Chapter 56: Industry 4.0 and Impacts on Industrial Hubs
56.1 Introduction
56.2 Digital Technologies and the International Division of Labour
56.2.1 The Smile Curve
56.2.2 Digital Technologies, the Geography of Manufacturing, and the Provision of Services
56.2.2.1 Robotization and the geography of manufacturing
56.2.2.2 Other digital technologies and the geography of manufacturing
56.2.3 Digital Technologies and Remote Provision of Services
56.3 Digitalization and Industrial Hubs
56.3.1 Industrial Hubs as Pockets of Digital Infrastructure and Skills
56.3.2 Industrial Hubs as an Instrument of Digital Industrial Policy
56.3.2.1 Providing a data-management framework
56.3.2.2 Changing the approach towards FDI
56.3.2.3 Focusing on the demand side
56.3.2.4 Functioning as digital innovation hubs
56.3.2.4.1 China’s digital innovation hub: from shenzhen to ‘shanzhai’
56.3.2.4.2 Republic of korea: turning industrial complexes intoinnovation clusters
56.3.2.4.3 Thailand’s digital park: a public-private partnership
56.3.2.4.4 Digital innovation hubs and start-ups
56.3.3 Industrial Hubs and Levels of Industrialization
56.4 Conclusions: Industrial Hubs in a Digital World
References
Chapter 57: Industrial Hubs and Economic Development: Conclusions and Pathways to the Future
57.1 Industrial Hubs and Economic Development: An Overview
57.2 Drivers of Cluster Dynamism: Implications for Successes and Failures
57.2.1 Variety and Diversity as Key Features of Industrial Hubs
57.2.2 Strategic Perspectives on Industrial Hubs
57.2.3 Cluster Dynamics and Industrial Hubs
57.2.4 Sectoral Perspective and Industrial Hubs
57.2.5 Emerging Issues and Industrial Hubs
57.2.6 Environmental Sustainability and Industrial Hubs
57.2.7 Technological Advancement and Industrial Hubs
57.2.8 Urbanization, Urban Transformation, and Industrial Hubs
57.2.9 Shifts in the Global Economy and Industrial Hubs
57.3 Conclusion: Learning and Latecomer Advantage
Acknowledgement
References
Name Index
Subject Index
Recommend Papers
File loading please wait...
Citation preview
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
T h e Ox f o r d H a n d b o o k o f
I N DUST R I A L HUBS AND E C ONOM IC DE V E L OPM E N T
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
the oxford handbook of
INDUSTRIAL HUBS AND ECONOMIC DEVELOPMENT Edited by
ARKEBE OQUBAY and
JUSTIN YIFU LIN
1
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
1 Great Clarendon Street, Oxford, ox2 6dp, United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © Oxford University Press 2020 The moral rights of the authors have been asserted First Edition published in 2020 Impression: 1 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2020931358 ISBN 978–0–19–885043–4 Printed and bound by CPI Group (UK) Ltd, Croydon, cr0 4yy Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
Acknowledgements
Working on this exciting yet challenging scholarly project has involved eighty-two authors. We have enjoyed the full support of contributors who put significant time and effort into this volume. The Handbook has benefited from constructive comments provided by internal and anonymous external reviewers. In particular, the editors are very grateful to all the contributors to this Handbook and thank them sincerely for their exceptional efforts and their personal commitment to ensuring the success of this volume. Special thanks are due to our commissioning editor, Adam Swallow, and his team at Oxford University Press, as well as to the entire production team for their advice, support, and encouragement. We also wish to thank the anonymous reviewers and delegates of OUP who approved our book proposal. We are grateful to Professor Mohamed Salih, Erasmus University, Netherlands and all the peer reviewers, who provided invaluable guidance on the structure and content of the book. This gratitude is also extended to all the participants at both the inception and chapter review workshops held in Addis Ababa in December 2018 and April 2019 for their insightful presentations, discussions, feedback, and collaboration. Our sincere appreciation goes to the Bill and Melinda Gates Foundation for the grant received for this project and to Mr. Haddis Tadesse and his team for their continued personal support. We also thank Mr. Tewolde Gebremariam, Group CEO Ethiopian Airlines, and his team, and Skylight Hotel, for their support. We also wish to thank IFC Ethiopia Office, especially Mr Adamou Labara, for supporting this project’s inception workshop; and Professor Tan Kong Yam (Nanyang Technological University, Singapore) and Mr. Tan Wooi Leong (Jurong Pte Ltd, Singapore) for their support in the cover design. Special thanks are due to Keith Povey Editorial Services Ltd for the pre-publication copyediting of the volume. And finally, our appreciation and gratitude goes to our project and research coordinator, Deborah M. Kefale, and the team of Helena Alemu, Tsion Kifle, Meron Tilahun, Samuel Arkebe Oqubay, and Binyam Arkebe Oqubay for their dedicated support in the preparation of the manuscript and assistance throughout the project.
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
Contents
List of Figures and Mapsxiii List of Tables xix List of Contributors xxiii
PA RT I C ONC E P T S A N D M E T HOD OL O GI E S 1. Industrial Hubs and Economic Development: An Introduction
3
Arkebe Oqubay and Justin Yifu Lin
2. Industrial Hubs and Economic Development: A Literature Review
15
Arkebe Oqubay
3. Heterodox Approaches to Industrial Policy and the Implications for Industrial Hubs
40
Christopher Cramer and Fiona Tregenna
4. The Economics of Innovation behind Cluster Dynamic Processes
64
Michael H. Best
5. Local Ecosystems and Social Conditions of Innovative Enterprise
77
Antonio Andreoni and William Lazonick
6. Industrial Policy, Institutional Transformation, and the Development of Industrial Parks
98
Howard Stein
7. Industrial Hubs: The Viewpoints of Economic Geography and Empirical Economics
114
Hisaki Kono
8. Industrial Hubs, Urban Systems, and Economic Development Edlam Abera Yemeru
130
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
viii Contents
PA RT I I H I S TOR IC A L C ON T E X T A N D A NA LY T IC A L T H E M E S 9. Industrialization, Hubs, and Catch-up: The World Economy in Historical Perspective
147
Deepak Nayyar
10. Industrial Districts in Europe: Lessons for Industrial Hubs and Industrialization162 Gioacchino Garofoli
11. The Political Economy of Special Economic Zones: Pasts, Presents, Futures190 Patrick Neveling
12. Cluster Dynamics and Regional Networks: New Argonauts, Silicon Valley, and Route 128
206
AnnaLee Saxenian
13. Agglomeration of European Industries
227
Michael A. Landesmann and Joris M. Schröder
14. The Global Experience in Special Economic Zones: A History and Review
243
Thomas Farole
15. Leapfrogging through Smart Manufacturing: The Role of Hubs in Industry 4.0
262
Padmashree Gehl Sampath
16. The Servitization of Industrial Hubs in the Digital Era
281
Shahid Yusuf
17. An Urban Planning Perspective on Industrial Hubs and Economic Development
301
Fantu Cheru and Aklilu Fikresilassie
18. Special Economic Zones and Export-led Growth: An Industrial Policy Imperative
323
Ludovico Alcorta and Taffere Tesfachew
19. A Flowchart Approach to Industrial Hubs Akifumi Kuchiki
345
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
Contents ix
PA RT I I I I N DUS T R I A L H U B S , L A B OU R , G E N DE R , A N D T H E E N V I RON M E N T 20. Industrial Hubs and the Industrial Labour Force in Africa and Asia
381
Carlos Oya and Florian Schaefer
21. Gender, Industrialization, and Industrial Hubs
401
Fiona Tregenna and Özge İzdeş
22. Women, Working Conditions, and Industrial Hubs
425
Arianna Rossi
23. Special Economic Zones in Latecomer Countries: Time to Bring Environmental Sustainability to the Fore
438
Giovanni Valensisi
24. The Greening of Industrial Hubs: A Twenty-first-century Development Strategy
454
John A. Mathews
25. Greening Structural Transformation: What Role for Industrial Hubs?
471
Tilman Altenburg and Kasper Vrolijk
26. Towards a New Generation of Special Economic Zones: Sustainable and Competitive
491
James X. Zhan, Bruno Casella, and Richard Bolwijn
PA RT I V P OL IC Y A N D PR AC T IC E : E M PI R IC A L E V I DE NC E F ROM A SI A 27. Industrial Hubs as Development Incubators: Asian Pioneers
523
Arkebe Oqubay
28. Explaining Reform and Special Economic Zones in China
559
Justin Yifu Lin, Jiajun Xu, and Junjie Xia
29. Industrial Hubs in ‘Sphinx’ China
574
Zonglai Kou and Jun Zhang
30. Innovative Firms and High-tech Industrial Hubs in China Yin Li
592
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
x Contents
31. Special Economic Zones in China and India: A Comparative Analysis
607
Yu Zheng and Aradhna Aggarwal
32. Taiwan’s Industrial Districts and Economic Development
623
Douglas B. Fuller
33. Location Choice for Industrial Complexes in South Korea
635
Jung Won Sonn and Sang Hee Kim
34. A Review of Industrial Clusters, Industrial Policy, and Industrialization in South Korea
650
Sanghoon Kim and Hah-Zoong Song
35. An Evidence-based Analysis of Industrial Hubs: The Singapore Narrative
673
George Yeo, Tan Khee Giap, Tan Kong Yam, and Wilfred Loo
36. Industrialization and Industrial Hubs in Malaysia
701
Rajah Rasiah and Gopi Krishnan
37. Industrial Policy and the Evolution of Industrial Hubs in Vietnam
723
Vu-Thanh Tu-Anh and Do-Thien Anh-Tuan
PA RT V P OL IC Y A N D PR AC T IC E : E M PI R IC A L E V I DE NC E F ROM L AT I N A M E R IC A 38. Industrialization in Latin America: A Critical Review
749
Sebastián Torres
39. The Development of Knowledge-based Export Clusters: Lessons from Costa Rica and Mexico
762
Jorge Cornick
40. Modern Industrial Policy in Latin America: Lessons from Cluster Development Policies
783
Carlo Pietrobelli
41. Latin American Industrialization: Unfulfilled Expectations and Future Opportunities Jorge Katz
799
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
Contents xi
42. Industrial Hubs and Structural Transformation in Latin America: Lessons from Costa Rica
815
Eva Paus
43. Industrial Hubs, Industrial Policy, and Economic Development in Mexico836 Jorge Carrillo and Saúl de los Santos
44. FDI, Industrial Hubs, and Latin American Industrialization
855
Anabel González
PA RT V I P OL IC Y A N D PR AC T IC E : E M PI R IC A L E V I DE NC E F ROM A F R IC A 45. A Strategic Approach to Industrial Hubs: Learnings in Ethiopia
877
Arkebe Oqubay and Deborah M. Kefale
46. Hubs Development and Industrial Upgrading in Mauritius
914
Ramola Ramtohul
47. Industrial Hubs and Technology Transfer in Africa’s Apparel Export Sector931 Lindsay Whitfield and Cornelia Staritz
48. Chinese Economic and Trade Cooperation Zones in Africa
950
Tang Xiaoyang
49. Legal Framework for Industrial Hubs: A Critical Review and Lessons for Africa
967
Won Kidane and Belachew M. Fikre
50. Industrialization and Industrial Hubs: Experiences in Kenya, Rwanda, and Tanzania
985
Josaphat Kweka and Dirk Willem te Velde
51. Special Economic Zones in sub-Saharan Africa: What Drives their Mixed Performance? Douglas Zhihua Zeng
1008
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
xii Contents
PA RT V I I S Y N T H E SI S A N D PAT H WAYS TO T H E F U T U R E 52. Cluster-based Policies: What Have We Learned?
1027
José Manuel Salazar-Xirinachs
53. Global Value Chains, Industrial Hubs, and Economic Development in the Twenty-first Century 1049 Gary Gereffi and Xinyi Wu
54. The Impact of Decentralization on Industrial Hubs: Blockchain and the Digital Economy
1069
Alisa DiCaprio, Maria V. Sokolova, and Jacqueline Yang
55. The Economics of Agglomeration
1086
Célestin Monga
56. Industry 4.0 and Impacts on Industrial Hubs
1104
Jörg Mayer and Rashmi Banga
57. Industrial Hubs and Economic Development: Conclusions and Pathways to the Future
1121
Arkebe Oqubay and Justin Yifu Lin
Name Index Subject Index
1133 1150
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
List of Figures and Maps Figures 3.1 Average manufacturing value added as a share of GDP, developed and developing countries (1970–2017)
48
3.2 Aggregate manufacturing value added as a share of GDP, developed and developing countries (1970–2017)
49
4.1 Cluster growth dynamics
65
4.2 Capability triad of competitive advantage
74
7.1 Backward linkage effects and forward linkage effects
116
7.2 Economic development and spread of industrial clusters
119
7.3 Illustrative example of relocation effects and biases in empirical studies
123
8.1 Percentage of population at mid-year in urban areas, 1950–2050
133
A12.1 Total high technology employment, Silicon Valley and Route 128, 1985–90 223 A12.2 Fastest-growing electronics firms, Silicon Valley and Route 128, 1985–90
223
13.1 Composition of the current account of the balance of payments, 2002–18, in per cent of GDP
230
13.2 External trade, 2003–17, in per cent of GDP
231
13.3 Shares in GVA of manufacturing 2000–2 vs. changes 2000/2–14/16
235
13.4a Employment shares in manufacturing (NUTS-2 regions) and GDP per capita levels
236
13.4b Employment shares in advanced tradable services (NUTS 2 regions) and GDP per capita levels
236
13.5 Shares in 2000/2 and change in shares of GVA of manufacturing 2000/2–14/16238 13.6 Complementarities in functional specialization within the Central European manufacturing core
239
14.1 Hierarchy of issues and link to SEZ models
247
16.1 Urban population in low- and middle-income countries, 1960–2017
282
16.2 Population by major regions 1950–2030 with 95 per cent prediction intervals284 16.3 Global share of manufacturing value added, 1994–2016
287
16.4 Share of manufacturing in employment
288
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
xiv List of Figures and Maps
16.5 Share of servitized firms by industry, France, 1997–2007
295
16.6 Share of services in industry output, France, 1997–2007
295
16.7 Labour productivity growth in manufacturing and services before and after 2000
296
17.1 Key drivers of structural transformation
304
17.2 Urbanization and GDP growth per capita by region, averages 1980–2014
305
17.3 Resources economic linkage potential
317
19.1 Flowchart approach: segment building
347
19.2 The electronics industry cluster in northern Vietnam
350
19.3 The sequence of the Eastern Seaboard Region: the case of Mitsubishi Motors360 19.4 Roles of Guangzhou City
363
19.5 Flowchart approach to innovation
369
19.6 Management structure of Zhongguancun Science Park
373
19.7 Partnerships between universities and companies in the Zhongguancun Science Park
375
23.1 Composition of wealth, by country income group
441
24.1 Shift eastwards of global manufacturing value added
456
24.2 China’s rise as premier manufacturing power
457
24.3 The black face of China’s energy transformation
458
24.4 China’s green shift: proportion of WWS electric power
458
24.5 China now a renewables superpower, 2018
459
24.6 Global electricity capacity, 2001–18: rise of WWS sources
460
24.7 Guitang closed-loop sugar-based industrial system
461
24.8 Learning curves for silicon solar cells and lithium-ion batteries
462
24.9 Top 100 industrial clusters in China
466
26.1 Historical trends in SEZs and current size
493
26.2 Distribution of SEZs by geographic regions
493
26.3 Distribution of SEZs by type
494
26.4 Impact of SEZs on trade and GVC integration: illustrative analysis
498
26.5 The SEZ development ladder
499
26.6 UNCTAD’s SEZ Sustainable Development Profit and Loss Statement
501
26.7 Towards a new generation of SEZs
502
26.8 SDG model zone: three dimensions of policy action
507
28.1 State-level SEZs
561
28.2 ETDZs and HIDZs in China
562
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
List of Figures and Maps xv
28.3 Emerging state-level SEZs
563
28.4 Number of state-level SEZs
563
28.5 Number of province-level SEZs
564
28.6 The evolution of SEZs in China
565
28.7 Structural change in China
568
28.8 Growth miracle in China
569
29.1 Number of industrial hubs in China, 1984–2018
577
29.2 Percentage of GDP of industrial hubs in China, 2002–16
578
29.3 Industrial hubs in the eastern, and central and western regions, 1984–2018 579 29.4 Number of industrial hubs by city, 1992 and 2018
580
29.5 Number of industries covered by national industrial hubs, 1984–2018
580
29.6 Number of industrial hubs by industry, 2018
581
29.7 GDP of industrial hubs by industry, 2002, 2008, and 2016
582
29.8 Distribution of the same industries in each province’s industrial hubs, 1998, 2008, and 2018
583
29.9 Distribution of industry similarity of each province’s industrial hubs, 1998, 2008, and 2018
584
30.1 R&D investment and R&D intensity in Beijing, Shanghai, and Shenzhen, 2009–17593 30.2 Shares of domestic patent grants by Beijing, Shanghai, and Shenzhen, 2008–17594 31.1 Contribution of SEZs to China’s economy, 2001–17
615
31.2 EPZ employment and share in manufacturing exports in India
617
31.3 Trends in SEZ employment and investment growth rate in India, per cent
618
33.1 Industrial hub development at the cross-section of three areas of public policy636 33.2 Land use in four industrial complexes
643
34.1 Integration of industrial complex laws
657
34.2 Legal system of industrial complex development in South Korea
661
34.3 Types of cluster projects
662
34.4 Industrial Complex of Korea Cluster Programme (ICCP) outline
663
34.5 Targets of the industrial cluster programme
665
34.6 Advantages and disadvantages of private versus public location planning
670
36.1 Institutional framework for supporting industrial upgrading
704
36.2 Debt service as a share of gross national income, Malaysia, 1970–2016
708
36.3 Value added in output, manufacturing, Malaysia and South Korea, 1968–2015709
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
xvi List of Figures and Maps
36.4 Value added in output, electric and electronics, selected countries, 1970–2015712 36.5 Value added in output, food and beverages, selected countries, 1970–2015, per cent
712
37.1 Distribution of industrial hubs and coastal economic zones, Vietnam
730
37.2 Contribution to GDP by economic sector
734
37.3 Labour structure by economic sectors, 1989 and 2018
734
37.4 The rise of industrial clusters in Vietnam
736
38.1 Per capita income relative to United States (PPP in 2011 dollars, multiple benchmarks)752 38.2 Productivity (TFP level at current PPP, United Sates = 1)
752
38.3 Informal employment in Latin America, per cent of total activity
753
38.4 Exports, product concentration index (1 = most concentrated)
754
38.5 Economic complexity index
755
38.6 Technological content of exports (Lall classification)
755
38.7 Research and development expenditure, per cent of GDP
756
38.8 Divergence of growth paths
758
38.9 Where is the value added along supply chains?
759
39.1 FDI, Costa Rica and Jalisco, 2000–18, US$ millions
771
39.2 FDI per capita, Costa Rica and Jalisco, selected years
772
39.3 Merchandise exports, Costa Rica and Jalisco, 2007–17
772
39.4 Merchandise exports per capita, Costa Rica and Jalisco, 2010 and 2015, US773 39.5 Merchandise and service exports, Costa Rica, 1999–2018
773
39.6 Competitive base of Costa Rica merchandise exports, 2007–17
774
41.1 Argentina, stop-and-go cycle, 1980–2012
801
41.2 Brazil, stop-and-go cycle, 1980–2012
801
41.3 Korea, 1980–2010
802
42.1 FDI inflows (US$ millions)
824
43.1 Location of priority industries in three industrial hubs by Mexican state
841
45.1 National network of industrial hubs, Ethiopia 2019
889
47.1 Apparel exports from top sub-Saharan Africa exporter countries, 2000–17 (US$ thousands)
936
50.1 EPZ employment and share in total manufacturing employment, Kenya, 1998–2018993 50.2 EPZ exports (US$ millions), share in total Kenya’s exports, 2008–18
993
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
List of Figures and Maps xvii
50.3 Key trends in SEZ development, 2007–17
999
53.1 Share of footwear made by Adidas and Nike in China and Vietnam, 2007–171056 56.1 Stylized potential effects from digitalization on the smile curve
1106
Maps 10.1 Map of Italian industrial districts in the 1990s
173
10.2 Map of French industrial districts in the 1990s
176
13.1 Countries’ shares in ‘value-added exports’ (VAX)—in manufacturing (rhs) and in total European exports
232
13.2 Shares of employment in manufacturing in total employment by NUTS-2 regions
234
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
List of Tables
9.1 The distribution of manufacturing production in the world economy: 1750–1913, in percentages: triennial annual averages
151
9.2 Changes in the distribution of industrial production in the world: 1913–73, in percentages
154
9.3 Share of selected countries and country-groups in world manufacturing value added, in percentages
157
10.1 Strategies, bifurcations, and transformations in industrial districts
181
10.2 Development trajectories for clusters in developing countries
186
11.1 The global spread of EPZs/SEZs since 1975
199
14.1 Continuum/forms of economic zones
246
15.1 Moving from analogue to smart production
271
15.2 Germany’s Industry 4.0 framework
273
17.1 Current and projected population distribution of Ethiopian urban centres for 2020 and 2025
308
17.2 Number of industrial parks by functionality and number of jobs created 310 17.3 Current and expected functions of the key urban centres
312
18.1 Types of economic zones (EZs)
325
18.2 Changes in policy in Korea 1960 onwards
332
19.1 Three theories of industrial agglomeration
346
19.2 Segment building
348
19.3 FDI, approved amounts, in US$ thousands, per cent
352
19.4 Number of IZs and FTZs in Malaysia
353
19.5 IZs in Thailand
354
19.6 Thang Long Industrial Park, list of tenants
359
19.7 Automobile industry agglomeration in Guangzhou City
364
19.8 Industrial zones and segment building
365
19.9 Honda’s related companies in Guangzhou
367
19.10 Enterprises invested in by research institutes/universities
370
21.1 Analytic framework of effects of hubs on women and gender equality
407
A26.1 The universe of SEZs, 2018
510
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
xx List of Tables
27.1 Phases of Taiwan’s industrialization
527
27.2 Synergy between industrial hubs and industrialization, South Korea
531
27.3 Trade performance outcome: technological classification of Chinese exports, 1985–2017
541
27.4 R&D intensity (investment in R&D as per cent of GDP)
543
A27.1 Taiwan: summary of industrial hubs
553
A27.2 South Korea: summary of industrial hubs
554
A27.3 Singapore: (a) Industrial hubs in Singapore; (b) Composition, area, and facility of industrial parks
554
A27.4 China: a summary of industrial hubs
555
31.1
Expansion of SEZs in China
610
31.2
Structure of India’s EPZ/SEZ sector
611
31.3
Status of manufacturing and IT/ITES SEZs in 2018
619
33.1
Typology of industrial hubs based on their locations
639
34.1
Korean industrial complexes
651
34.2
Overview of Korea’s industrial location policies
659
34.3
Implementation of development and spatial strategies in South Korea
666
34.4
South Korean industrial complex legislation
668
34.5
Evolution of the South Korean vocational training system
669
35.1
Number of manufacturing establishments and workers, 1960–2016
676
35.2
Value added of manufacturing establishments by industry, 1980–2016
676
35.3
Number of workers in manufacturing by industry, 1980–2016
677
36.1
Phases of industrialization in Malaysia
713
36.2
Investment and employment, regional corridors, Malaysia
718
37.1
Summary of special incentives for investors in industrial hubs, Vietnam
727
37.2
High-tech park statistics
731
39.1
Economic complexity indices and ranks, selected countries, 1995 and 2017763
39.2
Mexico, composition of goods exports, 1995 and 2017
763
39.3
Costa Rica, composition of goods exports, 1995 and 2017
764
42.1
The foreign trade zone regime in Costa Rica
820
42.2
Basic economic data for Costa Rica, 1980–2017
823
42.3
Goods exports by technology intensity and export regime
825
42.4
Composition of FTZ merchandise and services exports
826
42.5
Percentage composition of FTZ employment
827
42.6
Income tax payer status of companies in FTZ regime, December 2015
827
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
List of Tables xxi
42.7 Suppliers to FTZ producers
830
42.8 Key economic variables by firm size
832
42.9 Labour productivity by sector, 2012 (millions of colones)
832
43.1 SWOT analysis, aerospace industry, Baja California
843
43.2 Global map of the CTI system in the state of Querétaro
845
43.3 SWOT analysis of the automotive and auto parts areas of Nuevo León
847
A43.1 Typology of industrial hub features
851
A43.2 Industrial hub types, competitiveness index
853
45.1 Institutionalization framework
897
A45.1 List of industrial parks in Ethiopia (2019)
907
A45.2 Performance of industrial parks in Ethiopia: export and employment data
909
47.1 SEZ characteristics and technology transfer conditions and outcomes
933
48.1 Progress of China–Africa ETCZs (2019)
953
48.2 Expansion of Chinese zones into metropolitan areas
954
50.1 Estimated impact of moving into an SEZ: the case of Kigali SEZ
996
50.2 Comparison of firms in special economic zones with those outside
1000
50.3 Summary of experiences of African zones
1003
51.1 African zone programmes by decade of launch
1010
51.2 Estimates of direct employment and exports in zones in select African regions, 2004–6
1011
51.3 Selected SEZ cases in Africa
1013
51.4 A qualitative assessment of selected SEZs in Africa based on the key drivers for success
1019
52.1 The standard and experimentalist models of public policy
1036
52.2 Clusters recognized by the ACBC cluster policy
1041
53.1 Top Asian athletic and casual footwear manufacturers, 2018 (selected)
1054
53.2 Foxconn’s industrial hubs in mainland China
1062
54.1 Connecting blockchain and industrial hubs
1075
54.2 Conceptual mapping of sources of uncertainty to blockchain architecture1080 55.1 A typology of economies of scale
1092
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
List of Contributors
Arkebe Oqubay, PhD, is a senior minister and special adviser to the prime minister of Ethiopia and has been at the centre of policymaking for over twenty-five years. He is the former mayor of Addis Ababa and winner of the Best African Mayor of 2006 by ABN, and finalist for the World Mayor Award 2006, for transforming the city. He is a recipient of the Order of the Rising Sun, Gold and Silver Star presented by the Emperor of Japan. He currently serves as board chair of several leading public organizations and international advis ory boards. He is an ODI Distinguished Fellow and a research associate at the Centre of African Studies in the University of London, and holds a PhD in development studies from SOAS, University of London. His recent works include the path-breaking Made in Africa: Industrial Policy in Ethiopia (OUP, 2015); The Oxford Handbook of the Ethiopian Economy (co-edited with Fantu Cheru and Christopher Cramer, OUP, 2019); How Nations Learn: Technological Learning, Industrial Policy, and Catch-up (OUP, 2019); China–Africa and an Economic Transformation (OUP, 2019); African Economic Development: Evidence, Theory, and Policy (OUP, 2020); and The Oxford Handbook of Industrial Policy (OUP, 2020). He was recognized as one of the 100 Most Influential Africans of 2016, and a ‘leading thinker on Africa’s strategic development’ by the NewAfrican, for his work, both theoretical and practical, on industrial policies. His research focus includes structural transformation, catch-up, industrial policy, and policymaking, with a special emphasis on Africa. Justin Yifu Lin is dean of the Institute of New Structural Economics and Institute of South–South Cooperation and Development, and professor and honorary dean of the National School of Development at Peking University. He was the senior vice president and chief economist at the World Bank, 2008–12. He had previously served for fifteen years as founding director and professor of the China Centre for Economic Research (CCER) at Peking University. He is a councillor of the State Council and a member of the Standing Committee, Chinese People’s Political Consultation Conference. He is the author of more than twenty books, including Beating the Odds: Jump-starting Developing Countries (Princeton University Press, 2017); Going beyond Aid: Development Cooperation for Structural Transformation (Cambridge University Press, 2017); The Quest for Prosperity: How Developing Economies Can Take Off (Princeton University Press, 2012); New Structural Economics: A Framework for Rethinking Development and Policy (World Bank, 2012); Against the Consensus: Reflections on the Great Recession (Cambridge University Press, 2013); and Demystifying the Chinese Economy (Cambridge University Press, 2011). He is editor of The Oxford Handbook of Africa and Economics (OUP, 2015) and The Oxford Handbook of Structural Change (OUP, 2019). He is a corresponding fellow of the British Academy and a fellow of the Academy of Sciences for the Developing World.
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
xxiv List of Contributors Aradhna Aggarwal, professor at Copenhagen Business School, Denmark has published widely on international trade and business, SEZs, and technology transfers and innov ations, and is a senior editor of the International Journal of Emerging Markets. She is the key resource person for an annual training programme for SEZ officials from South-east Asia and China co-organized by ADB and Asia-Pacific Finance and Development Institute, China. She has participated in research projects on SEZs with ADB, World Bank, UN ESCAP, UNCTAD, and UNDP, and actively participates in international expert panel discussions, public forums, and research and training workshops on SEZs. Ludovico Alcorta, PhD, is Professorial Fellow on Technology, Innovation and Sustainable Development at UNU-MERIT, School of Business and Economics, University of Maastricht. He has been director of industrial policy, statistics, and strategic research at UNIDO, and is chief technical adviser on industrial strategy to the government of Oman. Ludovico’s research and publications are in the areas of new technologies, innovation, industrialization, and development. He has bachelor’s and master’s degrees in economics from Universidad del Pacifico, Lima, Peru, and an MPhil in development studies and a PhD from the University of Sussex. Tilman Altenburg, PhD, is an economic geographer and head of the Sustainable Economic and Social Development Department at the German Development Institute. Since 1986 he has carried out empirical research and published books and journal articles on issues of competitiveness, industrial and innovation policy, SME promotion, and value chain development. Antonio Andreoni, PhD, is associate professor of industrial economics at University College London (UCL) and head of research of the UCL Institute for Innovation and Public Purpose (IIPP). He is also visiting associate professor in the fourth industrial revolution, South African Research Chair in Industrial Development, University of Johannesburg, and served as adviser for the World Bank, OECD, DFID, UNIDO, UNCTAD, ILO, UNDP, UNECA, UNU-WIDER, EU Growth and Innovation, governments in South Africa, Tanzania, and the United Kingdom. His work on production, technology, and innovation dynamics, financialization, corruption, structural change, governance, and industrial policy appeared in Cambridge Journal of Economics, Development and Change, Cambridge Journal of Regions Economy and Society, Structural Change and Economic Dynamics, Oxford Review of Economic Policy, European Journal of Development Research and Energy Policy. Do-Thien Anh-Tuan is a lecturer in macroeconomics, public sector economics, financial analysis, and development finance at the Fulbright School of Public Policy and Management, Fulbright University Vietnam. His current research focuses on macroeconomic policy, public finance, development finance, and the banking system. In addition to teaching and research at Fulbright School, he participates in policy discussions and provides consultancy to the Vietnamese government and National Assembly on issues of economic development policy and reforms in budget, public investment, finance, banking, and state-owned enterprises.
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
List of Contributors xxv Rashmi Banga, PhD, is a senior economic affairs officer in the Unit on Economic Cooperation and Integration amongst Developing Countries, Division of Globalisation and Development Strategies, UNCTAD. She is a former adviser and head of the Trade Competitiveness in Commonwealth Secretariat, where she provided policy support to more than thirty countries. She was a professor of economics in Jawaharlal Nehru University, India. She received her doctorate from Delhi School of Economics and has published extensively on issues relating to international trade, FDI, and the digital economy. She has been awarded an international research award by Export-Import Bank of India and received two gold medals from the Global Development Network, the World Bank. Michael H. Best’s How Growth Really Happens: The Making of Economic Miracles through Production, Governance, and Skills (Princeton University Press, 2018) won the 2018 Schumpeter Prize. Professor Best has participated in development projects with UNIDO, the World Bank, and governments in more than twenty countries. He is also the author of The New Competition: Institution of Industrial Restructuring (Harvard, Polity, 1990); The New Competitive Advantage: The Renewal of American Industry (OUP, 2001); and The Politicized Economy (with William Connolly, 1982). He lives in Oxford, England. Richard Bolwijn is the head of the investment research branch in UNCTAD’s Investment and Enterprise Division, based in Geneva. As such, he is the research coordinator for UNCTAD’s annual flagship World Investment Report, as well as for a range of other publications on investment and enterprise development. He was lead author of WIR-theme chapters on GVCs, international tax, corporate nationality, and the digital economy. Prior to joining UNCTAD in 2010 Richard was a senior manager in McKinsey & Company, based in the Milan office, advising corporate clients worldwide on strategy and business development. He also worked in the economics group of KPMG Consulting, based in the London office. Richard holds degrees in international economics and management from Bocconi University in Milan, and in International Relations from Groningen University in the Netherlands. Jorge Carrillo, PhD, is a full-time researcher and co-founder of El COLEF, a research institute located in Tijuana, Mexico, since 1982. He has a PhD in sociology from El Colegio de Mexico. He has research visiting fellowships in Spain, France, Japan, the United Kingdom, and the United States, and is the author of sixteen books, twenty-seven edited books, 129 book chapters, and eighty articles in academic journals in Spanish, English, German, Portuguese, Italian, French, Japanese, and Chinese. He has participated in seventy research projects. He is active in international academic networks such as GERPISA-France and CRIMT-Canada. He is a specialist in technology, innovation, and employment in multi national corporations. Bruno Casella is senior economist at UNCTAD’s Investment and Enterprise Division. He is a core member of the team preparing the UN flagship World Investment Report. His main areas of expertise include trends and patterns in foreign direct investment and their impact
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
xxvi List of Contributors on international production, global value chains, and development. Prior to joining UNCTAD, he was manager at McKinsey & Company where he served C-level executive clients worldwide. He holds a PhD in statistics and a degree in economics at Bocconi University. He has published extensively in the area of international business and finance, economics, statistics and probability, both in refereed academic journals and in specialized books and reports. Fantu Cheru is emeritus professor of international relations, American University, Washington, DC and a senior researcher at the African Studies Centre, Leiden University, the Netherlands. Between 1998 and 2001, Professor Cheru was the UN special rapporteur on foreign debt for the Human Rights Commission in Geneva. He was associate senior fellow at the Stockholm International Peace Research Institute (Sweden), and the North-South Institute (Ottawa, Canada). From 2007 to 2012, Cheru was research director at the Nordic Africa Institute in Uppsala, Sweden. Professor Cheru was a member of UN SecretaryGeneral Kofi Annan’s Panel on Mobilizing International Support for the New Partnership for African Development (2005–7) as well as convener of the Global Economic Agenda Track of the Helsinki Process on Globalization and Democracy (Finland). Professor Cheru has served as both adviser and consultant to many governments and donor institutions and is on the editorial board of several scholarly journals. Jorge Cornick, PhD, University of Wisconsin, Madison, has more than thirty years’ experi ence as an international consultant on economic development, fiscal policy, and productive development policy, working for clients such as the Inter-American Development Bank, the International Labour Organization, and other international and domestic institutions. He has extensive public service experience in the fields of tax administration, public utilities regulation, financial services supervision, and national development planning. He is a partner at Eureka Comunicación, a strategic communications firm, and a founding partner of DRP Trejos & Cornick, an economics and economic policy consultancy firm. Christopher Cramer is professor of the political economy of development at SOAS, University of London. He is a vice-chair of the Royal Africa Society and chair of the Scientific Committee of the African Programme on Rethinking Development Economics. His publications include Civil War Is Not a Stupid Thing: Accounting for Violence in Developing Countries (Hurst, 2006), African Economic Development: Evidence, Theory, Policy (with Sender and Oqubay, OUP, 2020), and The Oxford Handbook of the Ethiopian Economy (OUP, 2019, co-edited with Cheru and Oqubay). He led the research project ‘Fairtrade, Employment, and Poverty Reduction in Ethiopia and Uganda’. Saúl de los Santos is a consultant and researcher, founder and CEO of Axis Strategic Intelligence Center, a private firm located in Tijuana, Mexico. He has a master’s degree in engineering with a specialization in quality and productivity from ITESM and an MBA degree with specialization in public administration from CETYS University. He works for government agencies as well as research centres, industrial associations, and clusters, specializing in assessing capabilities and generating strategies for sectors and regional development, with a particular focus on mid- and high-tech industries.
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
List of Contributors xxvii Alisa DiCaprio, PhD, is the head of trade and supply chain at R3 in New York City. Her primary focus is building blockchain capability and infrastructure into the global trade ecosystem. She also covers consortium governance and global standards initiatives in the trade space. Prior to R3, Alisa was a senior economist at Asian Development Bank working on digital trade, trade finance, and innovation. She has also worked in both the public and private sectors on export promotion, trade negotiations, and labour issues. Her PhD is from MIT, and she holds a BA and MA from Johns Hopkins University. Thomas Farole is a lead economist at the World Bank. He is the author of several books, including: Special Economic Zones in Africa; Making Foreign Direct Investment Work for sub-Saharan Africa (with Deborah Winkler); Special Economic Zones: Progress, Challenges, and Future Directions (with Gokhan Akinci); and The Internal Geography of Trade: Lagging Regions and Global Markets. His recent articles have appeared in IZA World of Labor; The Journal of Economic Geography; and World Development. Tom holds a PhD in economic geography from the London School of Economics and Political Science (LSE), an MSc in local economic development from the LSE, and a BSc in economics from the Wharton School of the University of Pennsylvania. Belachew M. Fikre, PhD, who served as commissioner on the Ethiopian Investment Commission, holds higher-level degrees from Addis Ababa University, the Royal Institute of Technology in Sweden, and University of Essex, King’s College London, and University of Surrey in the United Kingdom. Dr Belachew previously served as Deputy Commissioner for the Industrial Parks Division at the Commission, mandated to attract, support, and regulate industrial park developers, enterprises, and operators. He was founding staff and faculty dean of Bahir Dar University School of Law before joining Addis Ababa University Centre for Human Rights as lecturer and coordinator for research and publications. He currently works for the Deutsche Gesellschaft für Internationale Zusammenarbeit GmbH (GIZ) as an Adviser in Germany. Aklilu Fikresilassie is a qualified urban and development planning, monitoring, and evalu ation specialist with a Master of Arts degree in urban management and development from Erasmus University, the Netherlands (2006), and a Bachelor of Arts degree in economics from Addis Ababa University (2001). He is currently a PhD candidate in the School of Architecture, Planning and Environmental Policy, University College Dublin, Ireland. Aklilu’s research focuses on decentralized urban management and its effect on environmental management of common pool resources in Lake Zewai and Lake Hawassa, Ethiopia. Douglas B. Fuller is an associate professor in the Department of Asia and International Studies at City University of Hong Kong. He previously taught at Zhejiang University, King’s College London, Chinese University Hong Kong, and the American University in Washington, DC. His research spans comparative political economy of development, comparative capitalism (comparative political economy of wealthy societies), and technology policy, with a geographic focus on Greater China/East Asia. He has written Paper Tigers, Hidden Dragons: Firms and the Political Economy of China’s Technological Development (OUP, 2016, 2019) amongst other works. Gioacchino Garofoli is professor of economic policy at Insubria University (Varese, Italy); his previous appointments were in the universities of Bari and Pavia. He has been president
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
xxviii List of Contributors of the Association of Neolatin Economists since 2005, and a member of the editorial boards of the journals Regional Studies and Entrepreneurship and Regional Development since the mid-1990s. His publications include: Development on the Ground (Routledge); Endogenous Development and Southern Europe (Avebury); Economia e politica economica in Italia. Lo sviluppo economic italiano dal 1945 ad oggi (Angeli); Economia del territorio (Etas); Modelli locali di sviluppo (Angeli), and Industrializzazzione diffusa in Lombardia (Angeli). Padmashree Gehl Sampath is a leading expert on trade policy, innovation policy, and economic development. She is currently a senior adviser to the Global Access in Action programme and a fellow of the Berkman Klein Center, Harvard University. She is also a visiting fellow at GDAE Tufts University, an Adjunct Professor at the Department of Social Sciences, University of Aalborg, Denmark and a professional fellow at the United Nations University‒MERIT. Padmashree’s research interests focus on the economic and social implications of trade and technology in different contexts. She studies trade technology and industrialization; trade technology and inequality; technology, trade, and market power; and governance frameworks for technology, innovation, and industrialization. In her recent work on the digital economy, she analyses how questions of data infrastructure, architecture, and design impact stewardship and the exercise of rights, the persistence of exclusion, and potential governance solutions, looking at industrial policy, intellectual property, and technology policy. Gary Gereffi is professor of sociology and director of the Global Value Chains Center at Duke University. He received his BA degree from the University of Notre Dame and his MPhil and PhD from Yale University. Gereffi has published numerous books and articles on globalization, industrial upgrading, and social and economic development, and he is one of the originators of the global value chains framework. Recent books include: Handbook on Global Value Chains, co-edited with Stefano Ponte and Gale Raj-Reichert (Edward Elgar Publishing, 2019); Global Value Chains and Development: Redefining the Contours of 21st Century Capitalism (Cambridge University Press, 2018); Local Clusters in Global Value Chains: Linking Actors and Territories through Manufacturing and Innovation (Routledge, 2018); Brazilian Industry in Global Value Chains (Portuguese and English) (Elsevier, 2014); and Global Value Chains in Post Crisis World: A Development Perspective (World Bank, 2010). Tan Khee Giap, PhD, is co-director of the Asia Competitiveness Institute and associate professor at Lee Kuan Yew School of Public Policy, National University of Singapore. He is also chairman of the Singapore National Committee for Pacific Economic Cooperation. Dr Tan graduated with a PhD from the University of East Anglia, United Kingdom in 1987. He has consulted extensively with government ministries and statutory boards in Singapore. Dr Tan is the lead author of more than twenty books and has published extensively in international refereed journals. His current research interests include competitiveness of subnational Asian economies, global liveable cities, and the ease-of-doing-business index. Anabel González is a consultant on trade and investment, and non-resident senior fellow at the Peterson Institute for International Economics. She is former senior director of the
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
List of Contributors xxix World Bank’s Global Practice on Trade & Competitiveness and served as minister of trade of Costa Rica. She also worked as director, Agriculture Division, World Trade Organization; senior consultant on trade and investment, Inter-American Development Bank; and dir ector general, Costa Rican Investment Promotion Agency. She is a member of the World Economic Forum Global Future Council on Trade and Investment and has written extensively on trade and investment and presented in over sixty countries around the world. Özge İzdeş is an assistant professor of economics at Istanbul University—Cerrahpasa. She is also an affiliated instructor on the gender studies programme of the Women’s Studies Centre at Istanbul University. İzdeş holds BA and MSc degrees in economics from Istanbul University and received her PhD in economics from the University of Utah, United States. Her research areas include gender and macroeconomics, economic development, and labour economics. Her publications are mostly on crisis and gendered employment effects, employment-oriented macro policies and labour market inequalities, decent employment, and care economics. Jorge Katz, DPhil, received his first degree in economics at the University of Buenos Aires in 1964 and completed his tertiary education in England, where he obtained a DPhil in economics at Nuffield College, Oxford, in 1967. From 1994 until 2003 he was director of the Division of Production, Productivity and Management, at ECLAC, the UN Economic Commission for Latin America and the Caribbean, in Santiago de Chile. He now teaches at the Faculty of Economics and Business of the University of Chile in Santiago. Deborah M. Kefale is a researcher and development practitioner with extensive experience in designing, implementing, and evaluating social development, private sector and economic research programmes. She has worked with different bilateral and multilateral development agencies in Ethiopia and abroad. She has been a policy specialist at the UNDP Norway, Oslo Governance Centre, a programme specialist at the UNDP Liaison Office to the African Union and deputy programme manager at DFID Ethiopia. She has also served as an adviser to the Ethiopian Investment Commission. She holds an MA degree in development studies from Addis Ababa University. Won Kidane is a tenured associate professor of law at the Seattle University School of Law and a practitioner in international litigation and arbitration. Professor Kidane is the author or co-author of four books and dozens of law review articles in leading international law journals. He is a member of many organizations. He holds a Doctor of Juridical Science (SJD) degree from Georgetown, a Juris Doctor degree (JD) from the University of Illinois, a Master of Laws (LLM) from the University of Georgia School of Law, and a Bachelor of Laws (LLB) from Addis Ababa University. Sang Hee Kim, PhD, is director and CEO of Urban Regeneration Plus, a London-based design, planning, and regeneration consultancy. She obtained a BSc and MSc in architectural engineering at Seoul National University, and holds a PhD from the Bartlett School of Planning, University College London. Prior to the PhD degree, she worked in Korea for several architectural design practices. She is driven by an interest in the interrelationship
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
xxx List of Contributors between urban space and people, and the impact of planning policy on it. As a researcher and an architect, she is engaged in urban regeneration, design, and management projects. Sanghoon Kim, DPhil, is a research fellow at Korea’s Institute for Industrial Economics and Trade (KIET) and currently an adviser to the minister at the ministry of trade, industry, and energy. His research areas are technology/art management and policies, which include KBCs (IPs), technology evaluation/financing and cross-nation analysis in the context of SMEs, Industry 4.0 for manufacturing, culture (music and design), and the finance sector. He is author/co-author of more than seventy publications in the areas of materials science and technology management. He is also the core inventor of both KTRSTM (technology rating system) and TAGRTM (technology-market forecasting system), which are used for standard models in both public and private sectors nationwide. He holds a DPhil in materials science from the University of Oxford, master’s and bachelor’s degrees in metallurgy from Yonsei University, and master’s degrees in design management from KIDP and in technology management from Korea University. Hisaki Kono is an associate professor in the Graduate School of Economics, Kyoto University. He was previously research fellow at the Institute of Developing Economies and a visiting scholar at Harvard University. He studies various issues of economic development and poverty based on microeconomic theory, econometrics, and statistical methods including firm productivity, misallocation of resources, estimation of trade costs and linkage setup costs in agricultural trade, contract design for microcredit and insurance schemes, migration, education, and intra-household decision-making. Zonglai Kou, PhD is professor of economics and deputy dean of the School of Economics at Fudan University. His main research fields include industrial economics, innovation and intellectual property, digital economy and corporate finance. He has published many papers in refereed journals including Journal of Industrial Economics, Journal of Population Economics, Economics Letters and Economic Research Journal (jingji yanjiu). He has also chaired several research projects supported by the National Natural Science Foundation of China and the National Social Science Foundation of China. Gopi Krishnan is an associate director at the Institute for Capital Market Research, an affi li ate organization to the Securities Commission (SC) of Malaysia. His research is focused on macroeconomic and market analytics of the Malaysian capital market. He obtained his doctorate in economics from the University of Malaysia in 2019. In addition to his career as a researcher, he has also published academic works in the areas of innovation, institutions, and development policy. Akifumi Kuchiki was a staff member of the Japan International Cooperation Agency, a visiting research fellow at University of Pennsylvania, a senior economist at the World Bank, and the director-general of Japan External Trade Organization, Institute of Developing Economies. He also held positions as a visiting professor at Hiroshima University and Nagoya University, and was a specially appointed professor at Tokyo University. He was also executive vice president of the Japan External Trade Organization. Mr Kuchiki now works in the College of Bioresource Sciences at Nihon University.
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
List of Contributors xxxi Josaphat Kweka, PhD, is a lead consultant and founder of Talanta International, an advisory and research firm based in Tanzania. He is also a senior research associate with the Economic and Social Research Foundation (ESRF) and REPOA. With twenty-four years of professional experience, he has extensive knowledge of the Tanzanian and East African economies, and strong competency in undertaking development policy research. He has worked and published on a wide range of development policy issues, including tourism economics, industrialization, SMEs, SEZs, and trade and regional integration. He has conducted a range of advisory work for various clients, including the government of Tanzania; international organizations such as the World Bank, UNCTAD, UNDP, and AGRA; donor agencies for Denmark, Norway, Canada, Finland, and DFID; and local and international research organizations such as the UONGOZI Institute, ESRF, REPOA, Overseas Development Institute, International Food Policy Research Institute (FPRI), and the International Growth Centre (IGC). Formerly he worked with the World Bank as senior economist, and later with TradeMark East Africa as the Country Director for Tanzania. Michael A. Landesmann was scientific director of the Vienna Institute for International Economic Studies from 1996–2016 and is still attached to that institute. He is also a professor of economics at Johannes Kepler University where he is head of the Department of Economic Theory and Quantitative Research. His research focuses on international economic relations, European economic integration, structural change and economic growth, globalization and labour markets, and migration. He was a member of the Group of Economic Policy Advisors under the chairmanship of former EU Commission President Romano Prodi and has undertaken research for a wide range of international organizations (including World Bank, OECD, ILO, UNIDO, and UNCTAD). He has a DPhil from the University of Oxford, was a lecturer, fellow, and research officer at the University of Cambridge, and has held visiting professorships at Harvard University (where he was a J. Schumpeter Professor 1988–9 and Pierre Keller Visiting Professor 2008–9) and a range of other universities. He is a trustee of Cambridge University China Development Trust and a corresponding member of the Italian Academy of Science. William Lazonick, PhD, is an economist who studies innovation and competition in the global economy. Dr Lazonick’s research seeks to understand how, on the basis of innovative enterprise, a national economy can achieve stable and equitable economic growth. He is the originator of the theory of ‘innovative enterprise’. He also conducts cross-national comparative research on the social conditions that enable or proscribe innovative enterprise, focusing in particular on the economies of Britain, Japan, and China, as well as the United States. He received his Bachelor of Commerce degree from the University of Toronto, then attended the London School of Economics, where he was awarded an MSc in economics with distinction, and received his PhD from Harvard University. Yin Li, PhD, is an assistant professor of public policy at Fudan University’s School of International Relations and Public Affairs, where he studies innovation and development, science and technology policy, and institutional economics. His research has been published in Research Policy, Technovation, and Journal of Technology Transfer. Li holds a bachelor’s
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
xxxii List of Contributors degree in economics from Renmin University of China, and a PhD from Georgia Institute of Technology. Wilfred Loo has over thirty years of experience in urban planning and design practice, both in the public and private sectors in Singapore and internationally. He was formerly the senior vice president (planning) and head of the planning division in JURONG Consultants, and was involved in strategic industrial positioning studies, urban planning and urban design of residential townships, special economic zones, industrial parks, and specialized parks in Singapore and overseas. Wilfred is also active in the planning community in Singapore and is a fellow of the Singapore Institute of Planners. John A. Mathews is professor emeritus in the Faculty of Business and Economics at Macquarie University, Sydney. He was a professor of strategy at Macquarie Graduate School of Management for twenty years, retiring from active teaching in 2018. From 2009 to 2012 he held concurrently Eni Chair of Competitive Dynamics and Global Strategy at LUISS Gardo Carli University in Rome. For the past several years Professor Mathews has focused on the greening of industry with an emphasis on the role of China. The year 2017 saw the publication of Global Green Shift by Anthem Press in London. In July 2018 Professor Mathews was awarded the bi-annual Schumpeter Prize in recognition of his work and most recent book. Jörg Mayer, PhD, is a senior economist at UNCTAD, where he is part of the team that prepares UNCTAD’s annual ‘Trade and Development Report’. Before joining the United Nations in 1991, he was an economist in the international department of the German central bank. His areas of expertise include development strategies and the interrelationships between macroeconomic policies, trade, investment, employment, and growth. He has published in peer-reviewed journals, books, and conference volumes. He holds a PhD in economics from the European University Institute in Florence, Italy. Célestin Monga, PhD, is Senior Economic Adviser/Director at the World Bank Group and Visiting Professor of Economics at the University of Paris 1 Panthéon-Sorbonne (France) and Peking University (China). He has held various board and senior executive positions in academia, financial services, and international development institutions, serving most recently as Vice-President and Chief Economist of the African Development Bank Group and Managing Director at the United Nations Industrial Development Organization (UNIDO). He previously served on the Board of Directors of the Sloan School of Management’s Fellows Program at the Massachusetts Institute of Technology (MIT), and taught economics at Boston University and the University of Bordeaux. He also worked with the Banque Nationale de Paris Group as Department Head and Branch Manager at the BICEC Bank in Cameroon. Dr. Monga has published extensively on various dimensions of economic and political development. His books have been translated into several languages and are widely used as teaching tools in academic institutions around the world. His most recent works include The Oxford Handbook of Structural Transformation (2019), with J. Y. Lin; Beating the Odds: Jump-Starting Developing Countries (Princeton University Press, 2017), with J. Y. Lin; the two-volume Oxford Handbook of Africa and Economics (2015), with J. Y. Lin; and Nihilism
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
List of Contributors xxxiii and Negritude: Ways of Living in Africa (Harvard University Press, 2016). Dr. Monga holds graduate degrees from MIT, Harvard University, and the Universities of Paris and Pau. Deepak Nayyar is emeritus professor of economics at Jawaharlal Nehru University, New Delhi, and an honorary fellow of Balliol College, Oxford. He was formerly distinguished university professor of economics at the New School for Social Research, New York. He has also served as vice chancellor of the University of Delhi and chief economic adviser to the government of India. He has a DPhil in economics from the University of Oxford. His most recent books include Catch Up: Developing Countries in the World Economy (OUP, 2013 and 2016), Resurgent Asia: Diversity in Development (OUP, 2019), and Asian Transformations: An Inquiry into the Development of Nations (OUP, 2019). Patrick Neveling, PhD, is a researcher at the Historical Institute, University of Bern, and a fellow in the Department of Social Anthropology, University of Bergen. Previously, he held positions and fellowships in development studies, social anthropology, and global history at the universities of Basel, Halle, Hamburg, London, Manchester, and Utrecht. He is the author of six special issues and edited volumes, amongst them Capitalism and Global Anthropology (2018) and The Making of Neoliberal India (2014). In more than fifty publications, Patrick has analysed the historical incorporation of Mauritius into the capitalist world system since 1800 and the global history of special economic zones. Carlos Oya, PhD, is a reader (associate professor) in the political economy of development at SOAS, University of London, and a development economist by training. His main research interests are labour markets, poverty, agrarian political economy, political economy of development, development policy, development aid, and research methodology. Carlos has done extensive field-based research on contemporary labour market dynamics in various African economies. He is currently leading a major ESRC research project on the employment dynamics and effects of Chinese firms in sub-Saharan Africa, within a comparative framework focused on manufacturing and construction in Angola and Ethiopia. Carlos has published widely in journals such as Journal of Development Studies, Journal of Agrarian Change, Journal of Peasant Studies, Feminist Economics, Journal of Development Effectiveness, Review of African Political Economy, Journal of Modern African Studies, and Third World Quarterly, and recently co-edited the 2015 Routledge volume Rural Wage Employment in Developing Countries. Eva Paus is professor of economics on the Ford Foundation at Mount Holyoke College in Massachusetts, United States. She has published widely on development and globalization, often with a focus on Latin America. She is the author or editor of seven books and more than forty book chapters and articles in refereed journals, including World Development, Journal of Development Studies, Journal of Policy Reform, and Studies in Comparative International Development. Her most recent publications include Confronting Dystopia: The New Technological Revolution and the Future of Work (Cornell University Press, 2018, edited), ‘Innovation Strategies Matter: Latin America’s Middle-income Trap Confronts China and Globalization’ (Journal of Development Studies, 2019), and “Firm Innovation and the Middle Income Trap: Insights from Latin American Economies,” (CEPAL Review, forthcoming). She has received numerous national grants and has consulted for different UN organizations.
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
xxxiv List of Contributors Carlo Pietrobelli, PhD, is a professor and policy adviser on innovation and industrial development and policy. He is currently Professor of economics and Department Head at University Roma Tre, Italy, Professorial Fellow at UNU-MERIT, Maastricht, and adjunct professor at Georgetown University, Washington, DC. During 2009–16 he was a lead economist at the Inter-American Development Bank, where he led the preparation of development loans in Latin America and the Caribbean on innovation and industrial policies, and on cluster and value chains programmes, SMEs, and local economic development. His research interests range from development economics to innovation, trade, industry, and natural resources in developing countries. He has published widely in international journals and his books have been published by Harvard University Press, Edward Elgar, Palgrave, and Routledge. Ramola Ramtohul, PhD, is a senior lecturer in the Department of Social Studies of the Faculty of Social Sciences and Humanities at the University of Mauritius. She has a PhD in gender studies from the African Gender Institute at the University of Cape Town. Her research interests are in gender, politics, and citizenship in multicultural contexts. Some of her publications include: ‘Contested Terrain: Identity and Women’s Suffrage in Mauritius’ (Journal of Southern African Studies, 2016); ‘High Net Worth Migration in Mauritius: A Critical Analysis’ (Migration Letters, 2016); ‘Divided Loyalties and Contested Identities: Citizenship in Colonial Mauritius’, in Emma Hunter (ed.) Citizenship, Belonging and Political Community in Africa: Dialogues between Past and Present (Ohio Press, 2016); and ‘The Mauritian Paradox: Fifty Years of Development, Diversity and Democracy (University of Mauritius Press, 2018, with Thomas Hylland Eriksen). She has also received research fellowships from the University of Cape Town, American Association of University Women, University of Cambridge, and University of Pretoria. She is currently co-editor of the Journal of Contemporary African Studies and a member of the CODESRIA Executive Committee. Rajah Rasiah currently holds the position of distinguished professor of economics at the University of Malaya. His previous positions were professor of industrial organization at National University of Malaysia, and senior research fellow at UNU-INTECH, Maastricht. He obtained his doctorate in economics from the University of Cambridge in 1992. He was awarded the Celso Furtado Prize by the World Academy of Sciences in 2014 for advancing social science thought (economics). Arianna Rossi, PhD, is Senior Research and Policy Specialist for the International Labour Organization‒International Finance Corporation (ILO‒IFC) Better Work programme. Her work covers policy research, impact assessment, and gender inequality, with a particular focus on working conditions and labour rights in global value chains and the garment industry. She holds a PhD from the Institute of Development Studies at Sussex University, UK, an MSc from the London School of Economics and Political Science, and a degree in economics from the University of Ferrara, Italy. José Manuel Salazar-Xirinachs was regional director of the ILO Regional Office for Latin America and the Caribbean (2015–18), ILO assistant director general for employment glo bally (2005–14), director of the Trade Unit, Organization of American States (1998–2005),
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
List of Contributors xxxv and Costa Rica’s minister of foreign trade (1997–98). He has a PhD in economics from the University of Cambridge. He has published widely on economic growth and productive transformation policies, trade and competitiveness policy, and employment and skills pol icies. Recent books include Políticas de Formalización en América Latina: Avances y Desafíos (OIT, 2018, with J. Chacaltana) and Transforming Economies: Making Industrial Policy Work for Growth, Jobs and Development (2014, with I. Nubler and R. Kozul-Wright). AnnaLee Saxenian is a professor of economic development at the University of California, Berkeley. She was dean of the School of Information from 2004–19. She is the author of Regional Advantage: Culture and Competition in Silicon Valley and Route 128 (Harvard, 1994), and The New Argonauts: Regional Advantage in a Global Economy (Harvard, 2006), and has published in journals of economic geography, regional development, and industrial change. She chaired the Advisory Committee for Social, Behavioral and Economic Sciences at the National Science Foundation from 2010–15. She holds degrees from MIT, Berkeley, and Williams College. Florian Schaefer, PhD, is a fellow in the Department of Development Studies at the London School of Economics and Political Science. He previously worked as a postdoctoral researcher at SOAS, University of London, from where he also obtained his PhD. His main research interests are the political economy of industrialization and agrarian change, employment relations, the development of labour markets and entrepreneurship, and empirical research methods. Much of this research focuses on Ethiopia, where Florian has lived and worked for several years. Joris M. Schröder is an economist at the Vienna Institute for International Economic Studies. His research focuses on regional economic inequality, civil society, and labour markets. He holds an MSc in socioeconomics from the Vienna University of Economics and Business, and a BA in development studies from the University of Vienna. In addition, he is enrolled in the MA programme in development studies at the University of Vienna. Maria V. Sokolova, PhD, is a trade and market intelligence consultant at the International Trade Center. Her expertise lies in market access and sustainable trade policies. Her primary research focus is regional governance in international trade and macroeconomic effects of trade integration. She has experience working in both the private sector and academia and has also held positions at the International Monetary Fund and the UN Secretariat. She holds a PhD in international economics with a minor in international law from the Graduate Institute of International and Development Studies, Geneva, an MSc from the University of St Andrews, and a BSc from the Russian Economic Academy. Hah-Zoong Song is professor emeritus in the Department of Public Administration, KyungHee University. Dr Song served as chairman of the Presidential Commission on Policy Planning in 2005–6 and was also the president of the Korean Association for Policy Studies. He is a recipient of the Red Stripes, Order of Service Merit from the South Korean government and the Iljin Award of Excellence from the National Academy of Engineering of Korea. He holds a PhD from the John F. Kennedy School of Government, Harvard
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
xxxvi List of Contributors University. With a career focus on science and technology policy, Dr Song continues to research the interfaces of the natural and social sciences. His recent works include ‘Why Do South Korea’s Scientists and Engineers Delay Returning Home?’, ‘Strategy for Establishment of a National E-Science System’ and ‘Business Strategies and Policy Recommendations for Industrial Technology Progress’. Jung Won Sonn, PhD, is a lecturer in urban economic development at Bartlett School of Planning, University College London. Prior to this appointment, he studied at Seoul National University and the University of California, Los Angeles. He also had visiting posts at Seoul National University and the Chinese Academy of Science. His research is mainly about the interactions between the economy and urban planning in the United States, South Korea, China, and Indonesia. His research projects have been funded by the North American Regional Science Council, the European Commission’s Marie Curie Fellowship, and the Social Science Korea Grant, amongst others. Cornelia Staritz is tenure track professor in development economics at the Department of Development Studies at the University of Vienna, Austria. She is also a research associate at the Austria Foundation for Development Research (ÖFSE) and at Policy Research in International Services and Manufacturing at the University of Cape Town, South Africa. Her research focuses on economic development, international trade, global value chains and production networks, trade and industrial policy, and commodity-based development. Howard Stein is a professor in the Department of Afro-American and African Studies (DAAS) and Epidemiology at the University of Michigan. He is a development economist educated in Canada, the United States, and the United Kingdom. He has held university appointments in Japan, Tanzania, Netherlands, the United Kingdom, the United States, Canada, Ireland, and Portugal. He is the editor or author of more than a dozen books and edited collections. His research has focused on foreign aid, finance, structural adjustment and neoliberalism, health, gender and development, institutional transformation, industrial policy, export-processing zones, agricultural policy, poverty and rural property right transformation, and Chinese‒African economic relations. Taffere Tesfachew, PhD, is an international consultant on trade- and development-related issues. Until May 2016, he was director of the Division on Africa and Least Developed Countries, UNCTAD, where he led a team of economists who conducted research and analysis for two major annual reports, the ‘Economic Development in Africa Report’ and the ‘Least Developed Countries Report’. He has also authored and co-authored articles on a range of topics. Dr Tesfachew holds a PhD in development economics from the Institute of Development Studies, University of Sussex. Dirk Willem te Velde, PhD, is a principal research fellow and head of the International Economic Development Group at the Overseas Development Institute. He directs the Supporting Economic Transformation Programme and is research leader in the DFID-ESRC Growth Research Programme. He has written and edited a dozen books, thirty peer-reviewed articles and forty book chapters related mainly to investment, trade, and economic transformation. His research has featured on the BBC, and in China Daily, The Economist, Financial Times, and The
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
List of Contributors xxxvii Guardian, in African newspapers and TV, and in other media. He has advised donor agencies (including the EU, DFID, Sida, and the Dutch Ministry of Foreign Affairs), the UK Parliament and government ministers, developing-country ministers, and multilateral bodies (including IFIs and UN agencies). He holds a PhD from Birkbeck College, University of London. Sebastián Torres, PhD, works as an economist (development coordinator officer) at the United Nations Office in Uruguay. During the past decade, he was national director of industry at the Ministry of Industry, Energy and Mining of Uruguay, director of planning at the Presidency of Uruguay, research associate at the Trans National Institute (TNI), director of the National Research and Innovation Agency (ANII) of Uruguay, and a visiting research fellow at the London School of Economics and United Nations World Institute for Development Economics Research (UNU-WIDER). He has a post-doc from the Economic and Social Research Council (UK), a PhD from the University of Leicester, an MA from the Institute of Social Studies (The Hague), and a BA from the University of the Republic of Uruguay. Fiona Tregenna holds the DST/NRF South African Research Chair in Industrial Development and is also a professor of economics at the University of Johannesburg. She is a part-time member of the Competition Tribunal where she adjudicates competition (antitrust) cases, serves on a number of boards, councils, and advisory panels, and consults for various international organizations. Fiona has a PhD in economics from the University of Cambridge, a master’s degree in economics from the University of Massachusetts, and earl ier degrees from the Universities of the Witwatersrand and Natal (in South Africa). Her primary research interests are structural change, deindustrialization, and industrial development. Vu-Thanh Tu-Anh, PhD, is the dean of the Fulbright School of Public Policy and Management, Fulbright University Vietnam, and a senior research fellow at the Harvard Kennedy School, Harvard University. His primary research interests include political economy of development, public finance, institutional economics, and industrial policy. His recent publications include ‘Does WTO Accession Help Domestic Reform: The Political Economy of SOE Reform Backsliding in Vietnam’, and ‘The Political Economy of Industrial Development in Vietnam: Impact of State-Business Relationship on Industrial Performance’. Dr Vu-Thanh Tu-Anh is a member of the Economic Advisory Group for Vietnam’s prime minister during the period 2017–21. Giovanni Valensisi is economic affairs officer at UNCTAD, where he is part of the team in charge of the ‘Least Developed Countries Report’ and ‘Economic Development in Africa Report’. He previously worked on regional and multilateral trade negotiation issues at UNECA and carried out various assignments for UNCTAD, UNDP-Syria, and several international NGOs. He holds an MSc in environmental and natural resource economics from the Toulouse School of Economics, and a PhD in economics from the University of Pavia, where he was also a research fellow. Kasper Vrolijk is a researcher at the German Development Institute, with research interests in development economics, structural change, and political economy. He holds a PhD in economics from the School of Oriental and African Studies, University of London and a
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
xxxviii List of Contributors double-degree master’s in public policy from the Maastricht School of Governance and UNU-MERIT, Maastricht. Lindsay Whitfield is professor MSO in global studies and leader of the Centre of African Economies in the Department of Social Sciences and Business, Roskilde University, Denmark. She is author of several books on African politics and economics, including The Politics of African Industrial Policy: A Comparative Perspective (Cambridge University Press, 2015) and Economies after Colonialism: Ghana and the Struggle for Power (Cambridge University Press, 2018). Xinyi Wu previously a Research Assistant at Duke Global Value Chain Centre, is a Senior Analyst at AlixPartners. She received her BA in literature from Wuhan University and her MA in East Asian studies from Duke University. She has published ‘Amazon and Alibaba: Business Models, Internationalization Strategies and Internet Governance in the Digital Economy’ (Progress in International Business Research, 2018, with Gary Gereffi). Junjie Xia, PhD, is an assistant professor at the Institute of New Structural Economics at Peking University. His research lies mainly in fields related to macroeconomics. His current research work connects theoretical mechanisms with salient microeconomic data in the fields of financial market imperfections and liquidity reallocations; resource misallocation and its implications for international trade and income distribution; automation, labour share, and industrial upgrading; housing and wealth mobility; and SOEs and industrial policy in China. Dr Xia obtained his PhD in economics from the University of Southern California in 2017. Tang Xiaoyang, PhD, is vice-chair in the Department of International Relations at Tsinghua University and deputy director at the Carnegie-Tsinghua Centre for Global Policy. His research interests include political philosophy, China’s engagement in Africa, and the modernization process of the developing countries. He has published extensively on Asia‒Africa relations. He completed his PhD in the philosophy department at the New School for Social Research in New York. He has also worked as a consultant for the World Bank, USAID, and various research institutes and consulting companies. Before he came to Tsinghua, he worked at the International Food Policy Research Institute in Washington, DC. Jiajun Xu, DPhil, is an assistant professor and the executive deputy dean of the Institute for New Structural Economics at Peking University. Xu worked at the United Nations and World Bank and currently acts as the general secretary of the Global Research Consortium on Economic Structural Transformation. Her research focuses on development financing, industrial parks, and global economic governance. She has published in top academic journals in the field of international development such as the Journal of International Development and Institute of Development Studies Bulletin. She is the guest editor of China Economic Review and the lead author of the ‘Global Development Financing Report’ (2015). Her academic monograph Beyond US Hegemony in International Development was published by Cambridge University Press in 2017. She has led several policy-oriented development
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
List of Contributors xxxix projects aimed at leveraging special economic zones to achieve structural transformation in developing countries such as Benin, Djibouti, and Nigeria. Xu holds a DPhil (PhD) from the University of Oxford. Tan Kong Yam is professor of economics at the Nanyang Technological University. He is a board member of Changi Airport Group. From June 2002 to June 2005, he was a senior economist at the World Bank office in Beijing where he worked on the 11th Five Year Plan with the State Council. Prior to that, he was the chief economist of the Singapore government (1999–2002) and head of department of strategy and policy at the National University of Singapore business school. He is a graduate of Princeton and Stanford Universities. His research interests are in international trade and finance, economic and business trends in the Asia Pacific region and economic reforms in China. Jacqueline Yang studies political economy at the New York University Stern School of Business. She is also a market intelligence intern at R3, where her research focuses on blockchain supply chain initiatives and global trade standards. Prior to R3, she worked at CNBC monitoring emerging technology and blockchain market news for the digital news team. She has also been involved with public-sector efforts to promote the facilitation of trade finance, specifically through the use of distributed ledger technology. Edlam Abera Yemeru is the chief of the Urbanization and Development Section at the United Nations Economic Commission for Africa. In this capacity, she leads the delivery of technical assistance to African countries to strengthen linkages between spatial and economic planning through national and sector policies and investments. Previously she led a number of normative and operational programmes on urbanization and economic development at UN-Habitat, and at academic institutions in Africa and Europe. Edlam holds an MPhil in development studies from the University of Cambridge and a PhD in human geography from the University of London. George Yeo is a former Singapore Cabinet Minister who held various portfolios from 1991 to 2011 (Information and the Arts, Health, Trade and Industry, Foreign Affairs). From 2012 to 2019, he was Chairman of Kerry Logistics in HK. He is a Member of the Vatican Council for the Economy and a Board Member of AIA and Pinduoduo. He also serves on various advisory bodies. Shahid Yusuf, PhD, is currently chief economist of the Growth Dialogue at the George Washington University, School of Business in Washington, DC; a non-resident fellow of the Center for Global Development in Washington, DC; and adjunct professor at the Paul H. Nitze School of Advanced International Studies, Johns Hopkins University. Prior to joining the Growth Dialogue, he was on the staff of the World Bank. Dr Yusuf has written extensively on development issues, with a special focus on East Asia, and has also published widely in various academic journals. Douglas Zhihua Zeng, PhD, is a senior economist at the World Bank, and a global expert on spatial economy, technology innovation, and the knowledge economy. Since joining the Bank in 1998, he has worked on many countries globally. He has published many
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
xl List of Contributors eer-reviewed papers/books, including ‘Promoting Dynamic and Innovative Growth in p Asia’ (2016); Building Engines for Growth and Competitiveness in China: Experiences with SEZs and Industrial Clusters (World Bank, 2010); Knowledge, Technology, and Cluster-Based Growth in Africa (World Bank, 2008); Promoting Enterprise-Led Innovation in China (World Bank); and ‘China and the Knowledge Economy’ (2007), amongst others. Prior to joining the World Bank, he worked as a scholar at the Chinese Academy of Social Sciences and Stanford University. James X. Zhan is senior director at UNCTAD. He leads the UN World Investment Report and is editor-in-chief of the Transnational Corporations Journal. He chairs the Governing Board of the United Nations Sustainable Stock Exchanges Initiative (with all major stock exchanges worldwide as members). He has directed extensive research on key issues and facilitated the formulation of outcomes at various international summits. He has also provided policy advice to governments (including heads of states) and parliaments in over one hundred countries. He led the formulation of global guidelines for a new generation of investment policies and the establishment of the UNCTAD World Investment Forum. He is chief strategist for the World Association of Investment Promotion Agencies. Dr. Zhan has held several advisory positions with academic institutions (e.g. Cambridge University, Columbia University, Oxford University, and the University of Geneva). He was research fellow at Oxford University and Trade and Investment Council member of the World Economic Forum. He has published extensively on trade and investment-related economic and legal issues. Jun Zhang is a Chinese economist and currently serves as dean of the School of Economics at Fudan University in Shanghai. He received his BA, MSc, and PhD degrees from Fudan University. His fields of research are the Chinese economy, economic growth, the economics of transition, development economics, industrial economics, and property rights and institutional changes in China. Dr Zhang is the founding director of the China Centre for Economic Studies (CCES), a Shanghai-based think-tank for the Chinese economy. He is also a member of the special advisory committee to the Shanghai municipal government. Over the past twenty-five years he has published numerous books. In January 2018, he was awarded the Bergson Prize by the American Association for Comparative Economic Studies for his paper published in Comparative Economic Studies in 2015. Also in 2015, along with Justin Yifu Lin and Fan Gang, he was awarded the China Economics Innovation prize. Yu Zheng is a professor at the School of International Relations and Public Affairs, Fudan University. His research interests include international development, aid and investment, industrialization, and business‒government relations. He is the author of Governance and Foreign Investment in China, India, and Taiwan: Credibility, Flexibility, and International Business (University of Michigan Press). His publications have also appeared in journals such as Comparative Politics, International Studies Quarterly, Public Opinion Quarterly, Socio-Economic Review, and Studies in Comparative International Development.
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
pa rt I
C ONC E P T S A N D M ET HOD OL O GI E S
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
chapter 1
I n dustr i a l H u bs a n d Economic Dev el opm en t An Introduction Arkebe Oqubay and Justin Yifu Lin
1.1 Background to Industrial Hubs and Economic Development Industrialization, supported by industrial hubs, has been widely associated with structural transformation and catch-up in developing, emerging, and advanced economies.1 The notion of industrial hubs used in this volume is a generic expression of economic agglomeration and industrial clusters of economic activities that have evolved since the industrial revolution. Since World War II, successful export-led industrialization and enhancement of technological capabilities in developing countries have been driven by industrial hubs, which have in turn made significant contributions to economic development through direct and indirect economic benefits such as employment creation, exports, FDI promotion, and government revenue. The number of industrial hubs in Asia increased dramatically after the 1980s, with the Asia-Pacific region alone accounting for more than 65 per cent of jobs and exports globally, while Chinese special economic zones (SEZs) accounted for 80 per cent of cumulative FDI in the nation. In 2018, there were around close to 6,000 (out of which 5,400 SEZs were existing SEZs and about 500 were additionally planned SEZs) spread across 147 economies, with a high concentration in Asia (UNCTAD 2019).2 There were also more than 400 science and technology parks in the 2010s.3 These SEZs had created nearly 70 million direct jobs, total exports of close to a trillion dollars, and US$500 billion of direct trade-related value added.
1 See Ocampo et al. (2009) and Oqubay et al. (2020). See also Chapters 2 and 3. 2 See also FIAS (2008). 3 See Chapter 26.
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
4 Arkebe Oqubay and Justin Yifu Lin While these direct economic benefits are significant, the value of industrial hubs lies first and foremost in their contribution as incubators of technological capability and innovation, which in turn facilitates sustained growth and the climbing of the development ladder. Development of technological capabilities and domestic capacity is a key pathway to escaping the middle-income trap. In both emerging and advanced economies industrial complexes and high-tech firms are commonly to be found concentrated in hubs, which have proved critical conduits for domestic linkages and technological advancement.4 Industrial hubs have also become important components of industrial policies and development strategies in many developing and emerging economies. Industrial parks and/or SEZs or export-processing zones (EPZs) have been widely used in the catch-up and economic transformation of new industrializing economies (NIEs) in East Asia. The theories of agglomeration economies and the concept of industrial districts evolved during the time of Alfred Marshall in the late nineteenth century.5 Modern industrial hubs became the defining feature of late industrialization and economic catch-up after World War II, and many countries have adopted the industrial hubs concept and integrated policies into their national development paths accordingly. China used SEZs as a strategy for reform and opening up, as well as promoting industrialization and fostering its technological capabilities (Lin 2012; Oqubay and Lin 2019). Nevertheless, diverse conceptual interpretations cause some confusion. There is a paucity of literature offering both a comprehensive theoretical review of industrial hubs and economic development, and empirical evidence encompassing diverse experiences and multiple regions. Research on industrial hubs and economic development lacks the underpinning of an interdisciplinary approach. This book contributes to filling these gaps.
1.2 Preparing the Handbook 1.2.1 Aims and Approaches The Oxford Handbook of Industrial Hubs and Economic Development has three key objec tives: first, to provide a volume that examines the conceptual underpinnings, theoretical perspectives, and research methodologies involved in industrial hubs and economic development. Second, to extract lessons relevant to policymakers and practitioners from empirical evidence of policies and practices, supported by diverse national experiences and case studies. Third, to provide alternative perspectives and approaches embedded in an industrial policy framework to promote structural transformation and technological catch-up (Oqubay 2020). The volume’s editors have sought to adhere to the interdisciplinary approach, with insights from such disciplines as economic geography, urban economics, sociology, social anthropology, and political economy. However, the Handbook’s central theme is embedded predominantly in the development economics tradition and in heterodox perspectives on industrialization and catch-up, as well as the structuralist approach to economic 4 See Lee (2019). 5 See Marshall ([1890] 1920). Alfred Marshall lived from 26 July 1842 to 13 July 1924.
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
An Introduction 5 transformation and industrial policy. Rather than ideologically driven hypotheses, it is based on empirical evidence and economic history, which are widely explored in the book: the ultimate tests of industrialization, structural transformation, and catch-up associated with industrial policy frameworks and strategic approaches to industrial hubs, and guided by the developmental role of the state as the organizer of production.6 The volume’s focus on regional and country case studies from developing and emerging economies in Asia, Latin America, and Africa will have significant relevance for policymakers, scholars, and practitioners alike. The volume’s seven complementary sub-themes, which include conceptual perspective, context, empirical evidence, regional and country case studies, and synthesis, will provide a comprehensive understanding of industrial hubs and economic development. The volume combines conceptual and theoretical perspectives with context and empirical evidence. In addition to past, present, and future perspectives on industrial hubs, it has a vast regional coverage focusing on developing and emerging economies of the South. The focus is neither traditional loose clusters of micro and small enterprises, nor the ‘Third Italy’ notion of industrial districts, but rather, industrial hubs as policy instruments facilitating industrialization and promoting FDI, domestic learning, technological upgrading, catch-up, and structural transformation, not least in countries with inadequate infrastructure or business environment. The seven parts of the volume are:
(1) Concepts and methodologies (2) Historical context and analytical themes (3) Industrial hubs, labour, gender, and the environment (4) Policy and practices: empirical evidence from Asia (5) Policy and practices: empirical evidence from Latin America (6) Policy and practices: empirical evidence from Africa, and (7) Synthesis and pathways to the future.
1.2.2 Preparation of the Handbook A critical step in the preparation of the Handbook was the identification and selection of potential contributors, a highly rewarding process providing opportunities to widen a global research network. More than 120 eminent scholars were initially approached, of whom a total of eighty-two were selected and contributed fifty-seven original chapters. These present an authoritative yet accessible state-of-the-art analysis and authentic references. Anonymous reviewers made important contributions and raised a number of issues for consideration.7 A book project team, which has provided support throughout the process, organized an initial workshop in 2018, involving presentations, brainstorming sessions, and a review of all abstracts, to help develop the volume’s common themes. Draft chapters were revised 6 See Best (2018); Oqubay (2016). 7 A reminder to increase coverage of Latin American experiences, and a focus on extracting lessons from failures and ‘follies’.
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
6 Arkebe Oqubay and Justin Yifu Lin following external and peer reviews.8 A chapter review workshop held in 2019 provided feedback to authors, ensured the complementarity of various chapters, and reduced overlaps. Pre-publisher copy-edits improved readability and consistency of style. The publisher’s rigorous publication process has perfected the volume, spotting errors and improving quality.
1.2.3 Readership of the Handbook The primary readership of the Handbook will be scholars and researchers, policymakers, and practitioners. We anticipate that it will be a major research source for international, regional, and national research centres and think-tanks, as well as private-sector organizations and NGOs involved in hubs development. It may also be used as core or supplementary course material for undergraduates as well as graduate students in universities worldwide with a special interest in economic development and developing and emerging economies.
1.3 Themes and Perspectives The aim has been to produce an influential volume that brings together dispassionate, critical, and positivist views in a systematic and cohesive manner. In addition to looking back, the book aims to be forward-looking, and contribute to future research agendas on industrial hubs and economic development.
1.3.1 The Concept of Industrial Hubs The notion of industrial hubs is a generic expression of economic agglomeration and industrial clusters of economic activities that have evolved since the industrial revolution resulting in reduced transaction costs, external economy of scale, learning and innovation, and linkages in the development of industrialization and capitalism.9 Contributors whose chapters examine empirical perspectives and national experiences have been encouraged to use terms suitable for their context. The industrial districts of the eighteenth and nineteenth centuries were a specific form of industrial hub in which small firms in similar industries agglomerated to gain from positive externality, where support institutions were under developed. The term ‘industrial districts’ was used to represent the business model in regions of central and northern Italy, comprising industries such as apparel, footwear, and furniture, which became known as the Third Italy and was associated with post-Fordist flexible production systems.10 However, a combination of the external environment (globalization of production networks) and changes within industries and economic structure has seen a gradual diminution in the role of such industrial districts. Similarly, the term ‘industrial 8 Two reviewers (in some cases three) reviewed each draft chapter, shared written comments, and presented their reviews in person at the review workshop. 9 See also Krugman (1993). 10 See Chapter 2.
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
An Introduction 7 clusters’ includes a range of meanings from a cluster of small firms or the organic development of industrial hubs to cluster dynamics. Other widely used terms representing different concepts and national contexts include free zones, export-processing zones, special economic zones, economic zones, industrial zones, industrial parks, and science and technology parks. These terms, which vary across countries, are defined in national or provincial legislation where not only the rights and obligations of different players, but also incentives and supports are implied. ‘Free zones’ has been used for centuries to describe ports that are tax-free and provide transhipment platforms. Export-processing zones became popular after World War II, and especially in the 1970s and 1980s, as vehicles of export-led industrialization and conduits of globalization. In East Asian economies, export-processing zones have been used as part of industrial policy frameworks to promote exports, employment, and technological capabilities (Oqubay et al. 2020). On the other hand, they have also been associated with simple assembly and processing, and often operate outside the national labour regime to avoid the recipient country’s legislation. The new term ‘special economic zones’ (SEZs) was first used in China in 1980. These zones were intended to attract investment and manufacturing exports, develop technological capabilities, and encompass different types of industrial hubs, combining the limited scope of export-processing zones with larger national objectives (such as technology, manufacturing capability, and industrial upgrading) and ‘Chinese characteristics’. Special economic zones are defined variously in different countries, so the term cannot be used generically. The terms ‘industrial parks’, ‘industrial zones’, and ‘economic zones’ have also been used in different countries (including Singapore, South Korea, Vietnam, the Philippines, and recently Ethiopia) with the aim of developing manufacturing capability and exports. Other terms (science and technology park, innovation hub, and industrial complex) are associated with specific uses.
1.3.2 Complexities and Connections between Industrial Hubs and Economic Development The standard view of industrial hubs, which is to approach them as isolated entities, ignores the vital role of the industrial ecosystem, the web of institutional networks, and the symbiosis with broader dimensions and policy frameworks of industrial hubs. Industrial hubs are embedded in social structure, urban systems, politics and political economy, macro-economy, and the development of physical and human capital infrastructure.11 However, these critical dimensions and interactions occur within a constantly evolving external environment and domestic context. These complexities and connections are a common theme running through the various chapters of the Handbook. Gender issues in the industrial workforce are amongst the central aspects of the social fabric of industrial hubs, and are reviewed in depth.12 Urban systems and urban structures represent another important component of industrial hubs and are examined thoroughly.13 Climate change, green industrialization, and sustainable industrial hubs, which are becoming increasingly central to the industrialization agenda, 11 See Breschi and Malerba (2005) on clusters and networks. 12 See in Chapters 21 and 22. 13 See Chapters 8, 17, and 31. See also Jacobs (1969) on productive cities and innovation.
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
8 Arkebe Oqubay and Justin Yifu Lin are thoroughly examined in thematic chapters as well in many country experiences.14 Sustainability as a cross-cutting issue appears in several other chapters. Emerging changes in technology, sectoral shifts, global economy, and global value chains are examined extensively.15
1.3.3 Divergence in Development Perspectives and Variations in Empirical Evidence The editors of this volume believe that the ultimate test of ideas is practice-rooted in empirical evidence, rather than dogma. The book incorporates divergent perspectives from various traditions and disciplines, with views that have implications for policymaking and policy innovation supported by empirical evidence and economic history. Diversity notwithstanding, two broad views of industrial hubs and economic development emerge: a widely shared standard and static view; and a dynamic and strategic perspective embedded in long-term dynamics and trajectories of growth and economic transformation. This is particularly evident in the various chapters describing empirical experiences from advanced economies, and from the developing and emerging economies of Asia, Latin America, and Africa. These two broad perspectives reflect broad development paths and paradigms. A standard and oversimplified view that ignores the wider complexities and critical role of the industrial ecosystem, cluster dynamics, and complex connections, perceives industrial hubs as a bullet medicine or recipe for industrialization. While this view enjoys significant influence over policymakers and practitioners, a much more influential perspective emphasizes the direct benefits of economic liberalization (such as foreign direct investment, job creation, and exports), and sees hubs as a conduit of economic globalization in the 1980s and after. This view, widely popularized by international financial institutions and the international donor community who promote neoliberal viewpoints, is problematic because it is not driven by structural transformation and technological catch-up. A much deeper analytical perspective founded on the neoclassical tradition sees industrial hubs responding to market failures, such as deficiency of infrastructure or a weak business environment, and has the state playing a facilitating role (Lin 2012). The main emphasis is on industrial hubs as a vehicle for (a) removing bottlenecks in hard and soft infrastructure so as to turn the nation’s latent comparative advantages into the nation’s competitive advantages, and (b) facilitating the further upgrading of the nation’s latent comparative advantages through the rapid accumulation of human as well as physical capital brought out by the dynamic growth (Lin 2017). Recent contributions have also laid increasing emphasis on the middle-income trap, which is linked to inadequate focus on the development of domestic capability and the lack of development of technological and innovation capabilities.16 An alternative view focuses on long-term trajectories, strategic perspectives, and a more holistic and symbiotic view. This perspective is informed by the successful catch-up experi ences of the East Asian Tiger economies and advanced economies such as the United States, Germany, and Japan, and reinforced by experiences of countries that have suffered from the middle-income trap or premature deindustrialization (such as Malaysia, Brazil, and South 14 See Chapters 23, 24, and 25. 15 Chapters 15, 26, 53, 54, and 56. See also Gereffi (2018). 16 See Chapters 27, 29, 36, and 42, amongst others.
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
An Introduction 9 Africa). The Sino-centrism of the global economy has increased the intensity of the challenge. China’s accelerated industrialization, its rise as a global manufacturing powerhouse, and its continuing focus on technological and innovation capability is the subject of intense international interest as the transition unfolds and the country shifts to an innovation-led economy. Industrial hubs may develop strategic capabilities, fostering structural transformation and technological catch-up. In line with this perspective, this book demonstrates that an understanding of industrial hubs should be embedded in a production-centred paradigm not only looking to market failure but also taking a dynamic approach. Industrial hubs can also be development incubators that provide ecosystems to develop technological capability.17 This is important as technology, stored within organizations, is linked to learning by doing and new capabilities. Developing production, technological, and innovation capability is a long-term process that requires an effective development path and industrial and innovation policies that are adapted and connected to broader complementarities of policies and strategies.18 Hence, the industrial policy framework needs to focus on the transformation and development of production capability, the upgrading of exports and international trade ranking, and innovation capability. Industrial hubs that enhance these capabilities are therefore essential for climbing development ladders. The need to develop domestic capability, domestic linkages, and domestic learning lies at the heart of this perspective. The role of the state goes beyond ‘fixing’ market failures to becoming the organizer of production and propeller of technological capabilities. Learning, both policy innovation and learning by the state, and technological learning at firm level, is central to this perspective (Oqubay and Ohno 2019). The state’s role is not only to facilitate the competitiveness of the private sector but also to shape it by creating a conducive environment, providing mission-oriented support, and promoting the dynamism of firms. Incentives should be designed to augment firm- and industry-level capability development. Policy learning has become increasingly important if the state is to play a strategic and leading role as organizer of production rather than merely optimizer of resources and exchanges. In a dynamic political economy, the state is not immune from external influences, but rather is constantly adapting and learning from its interactions with sectoral, national, and international actors.19
1.3.4 Industrial Hubs and Implications for Latecomers Alternative perspectives on industrial hubs and economic development that emerge from various chapters of this book may have strategic implications for policymaking and innovation. Geographical stickiness and location specificities have become increasingly important for accelerating and scaling up innovation. Increased awareness on the part of consumers and the multilateral environmental protection dialogue clearly indicate that low-carbon industrialization is the only way forward. Urbanization has accelerated in developing economies and has become a major socioeconomic force shaping the future and generating increasing externality for production and innovation. The way forward is 17 See Best (2001, 2018) on cluster dynamics and the production capability triad as analytical framework. See Amsden (1989), Oqubay and Ohno (2019). 18 See Hirschman (1958). 19 See Lin and Zhang (2019).
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
10 Arkebe Oqubay and Justin Yifu Lin towards sustainability-oriented, innovation-driven, concentrated industrial hubs co-existing with liveable urban systems. The focus will be on understanding and enhancing interdependencies and linkages, complementarities, and externalities. The application of policy instruments is influenced by international law, the rise of global value chains (GVCs), and the increased role of multinational corporations (MNCs) which help to build domestic linkages and domestic learning. It is of paramount importance that developing countries focus on attracting dynamic FDI firms in targeted sectors, and adopt policies and incentives that enhance high-value activities, promote know-how and technology transfer, and deepen domestic linkages. Mega-production driven by industrial clusters is an increasingly important element in firm strategies.20 The Sino-centric world and China’s rebalancing strategy offer opportunities to attract productive investment and tap into export potential, while also increasing competitive pressure for new innovation-driven industrialization that builds domestic capabilities and develops technological and innovation capabilities to escape the middle-income trap (Oqubay and Lin 2019; Lin 2017). Despite the increased challenge presented by hyper-globalization and premature deindustrialization, the new global context magnifies the importance of export-led industrialization and industrial policies. Industrial hubs should focus on developing manufacturing capabilities and turning the country’s latent comparative advantages into competitive advantages. The rapid pace and increasing prominence of technological development increases the role of technological learning and innovation capability. The emergence of the digital economy requires regulatory capacity and an understanding of its impact on various sectors and value chains. The quality and sustained growth of the services sector is arguably dependent on productive investment driving the dynamism of manufacturing and technological capabilities.21 Neither economic history nor robust empirical studies support the abandoning of the manufacturing sector, with significant implications for the development path chosen by emerging economies and developing countries.
1.4 Organization of the Handbook One of this book’s aims is to create clarity on the conceptual approach to industrial hubs. The literature is currently ambiguous, with alternative terms, such as industrial districts, industrial estates, special economic zones, export-processing zones, and industrial parks having different connotations and reflecting specific historical and national contexts (see Chapter 2). A common theme of the various chapters has been to examine and explain the strategic approach to industrial hubs development associated with latecomer advantage. With technological advancement becoming more central, the volume focuses on examining innovation and cluster dynamics. The chapters which present a review of experiences from specific regions have also attempted to focus on learning from failures and mistakes. The first, second, and third parts of this volume review the conceptual, historical, and key analytical themes. The fourth part dives into gender and environmental sustainability. 20 See Porter (1990). 21 See Kaldor (1967) on special properties of the manufacturing sector as engine of structural change.
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
An Introduction 11 The next three parts examine policy and practice of industrial hubs and economic development, with a focus on regional and country case studies from Asia, Latin America, and Africa. Empirical evidence is drawn from developing countries, emerging economies, and advanced economies. Part I: Concepts and Methodologies lays the conceptual and theoretical foundation of industrial hubs and economic development with a review of concepts, theoretical traditions, and methodologies. Oqubay’s chapter reviews literature on industrial hubs collated from development economics and related disciplines. Cramer and Tregenna present an alternative perspective on structural transformation and catch-up, and implications for industrial hubs. Best reviews cluster dynamic processes and the economics of innovation, and this is complemented by Andreoni and Lazonick with a discussion focused on the nexus between innovation, agglomeration economies, and industrial hubs. Stein examines links between industrial policy, institutional transformation, and industrial hubs. Kono introduces an economic geographic approach to industrial hubs. Yemeru considers the role of urban economic systems, and the implications of urbanization trends for industrial hubs. Part II: Historical Context and Analytical Themes consists of chapters that provide a deeper understanding of industrial hubs and their connections with related themes. Nayyar expounds the long historical perspectives of developing countries’ catch-up and industrial hubs as a critical context for understanding the contemporary world. Garofoli puts the evolution of industrial districts in Western Europe into perspective. Neveling details the genesis and review of special economic zones in the twentieth century, and Farole complements this with a review of global experience of special economic zones. Saxenian presents the genesis and unique features of the notable innovation hub of Silicon Valley, while Landesmann and Schröder review agglomeration of European industries.22 Gehl Sampath reviews the transformation of the new economies, with a focus on innovation and industrial hubs. Yusuf reviews the growth of the services sector or ‘servicification’ and its implications for hubs development. Cheru and Fikresilassie present an urban planning perspective of industrial hubs, and Alcorta and Tesfachew present an industrial policy imperative on industrial hubs. The final chapter by Kuchiki presents his pioneering approach to the life cycle of industrial hubs, which he terms the ‘economy of sequencing’. This part builds on and complements the conceptual perspectives, laying the foundations for the empirical evidence and case studies of the next three parts. Part III: Industrial Hubs, Labour, Gender, and the Environment reviews important and multidimensional aspects of industrial hubs with immense social and political significance, including labour dynamics and gender, and climate change, environmental protection, and sustainability. Oya and Schaefer explore the social dimension with a review of the historical evolution of the industrial workforce, examining its centrality in the development of capitalism. İzdeş and Tregenna discuss gender and the industrialization process, while Rossi considers the future of gender and industrial hubs. The remaining chapters of the section look at climate change and sustainability. Valensisi provides a critical review of trends in sustainability with a particular focus on developing countries, showing why sustainability should be the central thrust of industrial hubs. Mathews examines the broader and deeper connection between climate change, industrialization, and environmental sustainability, putting forward green capitalism as the new paradigm. Altenburg and Vrolijk explore the 22 See also Saxenian (1996) on pioneering work in Silicon Valley and Route 128 in the United States.
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
12 Arkebe Oqubay and Justin Yifu Lin effects of incorporating sustainability and environmental protection into strategic approaches to industrial development. Zhan, Casella, and Bolwijn review the advances in industrial hubs development and pathways to a new generation of special industrial zones. These labour, gender, and environmental sustainability themes are further explored throughout the volume, in particular in the chapters that include empirical evidence and national and regional experiences. Part IV: Policy and Practices: Empirical Evidence from Asia presents lessons from regional and national experiences. Asia is a dynamic region where the greatest economic trans formation and catch-up has occurred since World War II. Oqubay reviews the Asian experience with a focus on the pioneers of industrial hubs (Taiwan, South Korea, Singapore, and China). Lin, Xu, and Xia examine China’s reforms and special economic zones in depth, while Kou and Zhang focus on industrial policies and industrial hubs. This is complemented by Li’s review of wider transformation and industrial hubs in China. Zheng and Aggarwal present a comparative review of special economic zones in China and India. Fuller reviews industrial hubs and catch-up in Taiwan, while Sonn and Kim review the South Korean industrial complex applying an urban planning approach. Kim and Song consider industrial hubs, industrial policies, and industrialization in South Korea, while Yeo, Giap, Yam, and Loo examine industrial hubs and industrialization in Singapore.23 Rasiah and Krishnan look at the Malaysian journey towards industrialization and industrial hubs, while Tu-Anh and Anh-Tuan examine Vietnam’s reforms, industrialization, and industrial hubs. The wide range of chapters in this section offers lessons for policymakers on how industrial hubs are interwoven with industrial policies, and on the increasing importance of innovation and technological capability. Part V: Policy and Practices: Empirical Evidence from Latin America explores the Latin American experience, beginning with a critical review by Torres of industrialization in the region. Cornick’s chapter examines industrial hubs and economic development in the region, with a special focus on Costa Rica and Mexico. Pietrobelli reviews clusters, GVCs, and industrial hubs in Latin America, and Katz’s chapter focuses on a review of agglomer ation economies and industrial hubs in Latin America. The chapter by Paus reviews the regional experience and presents an in-depth analysis on Costa Rica. Carrillo and de los Santos review industrial policy and industrial hubs in Mexico, and González reviews FDI and Latin American industrialization. The focus in Part VI: Policy and Practices: Empirical Evidence from Africa is on the specific characteristics of Africa; a range of country case studies is related to successful experiences from elsewhere. Lessons from recent experiments in industrial hubs are presented by Oqubay and Kefale. Ramtohul reviews the pioneering experiences of Mauritius from the 1970s to recent times. Whitfield and Staritz examine the evolution and features of apparel and textile industrial hubs across Africa. Xiaoyang presents the genesis and outcomes of Chinese economic and trade cooperation zones across the continent. Kidane and Fikre review the legislative framework for industrial hubs in Africa, with recent lessons for late latecomers. Kweka and te Velde explore special economic zones and economic transformation with a focus on industrialization in East Africa, while Zeng focuses on learning from successes and failures in special economic zones in Africa.
23 See also Lee (2018).
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
An Introduction 13 The final section, Part VII: Synthesis and Pathways to the Future, reviews emerging trends in industrial hubs and economic development in the twenty-first century. Salazar-Xirinachs presents a review of lessons gleaned from cluster-based policies and approaches Gereffi and Wu’s chapter reviews GVCs and their connection to industrial hubs and economic development. DiCaprio, Sokolova, and Yang review block chain and the digital economy and its impact on industrial hubs, and Monga reviews the economics of agglomeration. Mayer and Banga discuss Industrial Revolution 4.0 with a focus on the digital economy and how techno logical development is influencing GVCs and hubs development in the twenty-first century. The final chapter brings together the key conclusions and indicates pathways to the future.
1.5 Conclusions Interest in industrial hubs heightened during the final phase of the preparation of the Handbook. UNCTAD’s (2019) World Investment Report was devoted to industrial hubs development, which was discussed at the Trade and Development Board in Geneva. Policymakers and practitioners are increasingly seeking a better understanding of the nexus between industrial hubs and economic development. Readers, ultimately, will judge whether the Handbook lives up to its ambitious aims. However, despite the editors’ wish to produce a comprehensive volume covering all key aspects with the inputs of our eighty-two contributors, some subjects remain either not covered or insufficiently covered. The editors believe this will encourage future research on these topics, and would like to emphasize the importance of interdisciplinary research and cross-fertilization of knowledge, pushing the frontiers of research still further. National governments, international organizations, businesses, industrial associations, and research institutes have an important role to play in fostering these research undertakings. Universities and training centres might also consider incorporating research outcomes into future courses on industrial hubs, industrialization, and economic development.
Acknowledgement The authors are grateful to Adam Swallow for his constructive comments, and to Deborah Kefale, Samuel Arkebe, and Binyam Arkebe for inputs to improve the draft, and for their continued support.
References Amsden, Alice (1989) Asia’s Next Giant: South Korea and Late Industrialization. Oxford: Oxford University Press. Best, Michael (2001) The New Competitive Advantage: The Renewal of American Industry. Oxford: Oxford University Press. Best, Michael (2018) How Growth Really Happens: The Making of Economic Miracles through Production, Governance, and Skills. Princeton, NJ: Princeton University Press. Breschi, Stefano and Franco Malerba (eds) (2005) Clusters, Networks, and Innovation. Oxford: Oxford University Press. FIAS (2008) Special Economic Zones: Performance, Lessons Learned, and Implications for Zone Development. Washington, DC: World Bank. Gereffi, Gary (2018) Global Value Chains and Development. Cambridge: Cambridge University Press.
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
14 Arkebe Oqubay and Justin Yifu Lin Hirschman, Albert (1958) The Strategy of Economic Development. New Haven, CT: Yale University Press. Jacobs, Jane (1969) The Economy of Cities. New York: Vintage Books. Kaldor, Nicholas (1967) Strategic Factors in Economic Development. Ithaca, NY: Cornell University Press. Krugman, Paul (1993) Geography and Trade. Cambridge, MA: MIT Press. Lee, Keun (2018) The Art of Economic Catch-up: Barriers, Detours, and Leapfrogging in Innovation Systems. Cambridge: Cambridge University Press. Lee, Keun (2019) ‘The Origin of Absorptive Capacity in Korea: How Korean Industry Learnt’, in Arkebe Oqubay and Kenichi Ohno (eds) How Nations Learn: Technological Learning, Industrial Policy, and Catch-up. Oxford: Oxford University Press, pp. 125–48. Lin, Justin Yifu (2012) The Quest for Prosperity: How Developing Economies Can Take Off. Princeton, NJ: Princeton University Press. Lin, Justin Yifu (2017) ‘Industrial Policies for Avoiding the Middle-income Trap: A New Structural Economics Perspective’, Journal of Chinese Economic and Business Studies 15(1): 5–18. Lin, Justin Y. and Jun Zhang (2019) ‘China: Learning to Catch up in a Globalized World’, in Arkebe Oqubay and Kenichi Ohno (2019) (eds) How Nations Learn: Technological Learning, Industrial Policy, and Catch-up. Oxford: Oxford University Press, pp. 149–72. Marshall, Alfred ([1890] 1920) Principles of Economics. London: Macmillan. Ocampo, José A., Cordina Rada, and Lance Taylor (2009) Growth and Policy in Developing Countries: A Structuralist Approach. New York: Columbia University Press. Oqubay, Arkebe (2016) Made in Africa: Industrial Policy in Ethiopia. Oxford: Oxford University Press. Oqubay, Arkebe (2020) ‘The Theory and Practice of Industrial Policy’, chapter 2, in Arkebe Oqubay, Christopher Cramer, Ha-Joon Chang, and Richard Kozul-Wright (eds) The Oxford Handbook of Industrial Policy. Oxford: Oxford University Press. Oqubay, Arkebe and Justin Yifu Lin (eds) (2019) China–Africa and an Economic Transformation. Oxford: Oxford University Press. Oqubay, Arkebe and Kenichi Ohno (eds) (2019) How Nations Learn: Technological Learning, Industrial Policy, and Catch-up. Oxford: Oxford University Press. Oqubay, Arkebe, Christopher Cramer, Ha-Joon Chang, and Richard Kozul-Wright (eds) (2020) The Oxford Handbook of Industrial Policy. Oxford: Oxford University Press. Porter, Michael (1990) The Competitive Advantage of Nations. New York: Free Press. Saxenian, Annalee (1996) Regional Advantage: Culture and Competition in Silicon Valley and Route 128. Cambridge, MA: Harvard University Press. UNCTAD (2019) The World Investment Report: Special Economic Zones and Development. Geneva: UNCTAD.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
chapter 2
I n dustr i a l H u bs a n d Economic Dev el opm en t A Literature Review Arkebe Oqubay
2.1 Background and Introduction to Literature Review of Industrial Hubs Historical accounts show that the clustering of economic activities took a different form after the advent of the eighteenth-century industrial revolution in England. Industrial clusters and the concentration of firms in England during the eighteenth and nineteenth centuries have been described as industrial districts, whereas in Europe post-Fordist mass production after the late 1970s has been described in the literature associated with ‘Third Italy’ and Emilia-Romagna regions. Industrial clusters, with high concentrations of small and medium enterprises, have been widely observed in developing countries. New terms that have surfaced since World War II, such as ‘export-processing zone’, have been associated with the globalization of production, the expansion of international trade, the rise of transnational corporations, and the internationalization of foreign direct investment. In 2018, there were nearly 6,000 industrial hubs worldwide (of which 500 were newly planned) with prime concentration in developing economies particularly in Asia (UNCTAD 2019).1 However, in terms of scale and scope, the largest increase of industrial clusters has been observed in East Asia since the mid-1960s in the form of export-processing zones, industrial zones, industrial parks, and industrial complexes. The term special economic zone became a mainstream concept with wide application in China.2 East Asian economies designed innova tive policies on industrial hubs to develop their manufacturing sectors and industrialization, 1 In the 2010s, there were more than four hundred science and technology parks (UNCTAD 2019). 2 On the use of the notion of industrial hubs and various alternative terms and applications, see Chapter 1.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
16 Arkebe Oqubay increase their exports to gain a greater share of international trade, diversify their economies, and build technological capabilities. Despite this major shift, it was evident that the process and context of the development of industrial hubs differed between East Asian countries. Unlike the case in East Asia, US technology hubs have always been leaders in the field, with hubs like Silicon Valley and Boston’s Route 128 focusing on developing technological and innovation capabilities. Whether their purpose is to catch up or to maintain leadership, industrial hubs are found across the world in both developed and developing economies. Conservative estimates show that industrial hubs created more than 70 million manufacturing jobs and generated more than US$1 trillion worth of exports in the early 2000s.3 However this was not evenly distributed globally, with the Asia-Pacific region accounting for 65 per cent of jobs and exports created. These industrial hubs were not like the industrial districts of the ‘Third Italy’ or the clusters of small enterprises that had been idolized in many developing countries. Nor were they limited to labour-intensive manufacturing, but were, rather, widely found in heavy and chemical industries (for example in South Korea, Singapore, and China), and even more importantly as science and technology parks or innovation hubs.4 This is impressive progress in view of their remoteness from innovation frontiers (such as Silicon Valley) compared with Europe or other regions in the United States. Contrary to standard perceptions, industrial hubs can be used as vehicles for technology transfer, technological upgrading, and innovation. Moreover, the development of industrial hubs has occurred within the wider development strategy and within industrial policy frameworks, in conjunction with other policies (such as joint ventures or access to domestic markets) to develop domestic production linkages.
2.1.1 Focus and Approach of Literature Review Industrial hubs are at the centre of economic development. However, not all countries or hubs have succeeded equally.5 Policy coherence has not been equally effective in all countries and hubs, and industrial hubs have faced diverse challenges and opportunities. Moreover, the literature on industrial hubs is fragmented and characterized by diverse conceptual and methodological approaches. This volume seeks to fill gaps in the understanding of industrial hubs and economic development. The guiding objectives of this chapter are first, to provide a synopsis of the literature on the theory and practice of industrial hubs and economic development drawn from various intellectual traditions. Second, to provide an overview of the conceptual and empirical discussions contained in the volume, emphasizing the importance of an interdisciplinary approach to include development economics, political economy, economic geography, spatial economics, and urban economics. Third, to review key themes drawn from structuralist development economics, with a special focus 3 For instance, see FIAS (2008). 4 The Jurong Petrochemical Hub in Singapore is amongst the top global petrochemical hubs. Industrial complexes in South Korea were the key vehicles for developing heavy and chemical industries (HCI) in the 1970s and 1980s along the South Korean coast. 5 The Malaysian development of a multimedia super corridor is an example of mixed outcomes. Similar examples are available in other South Asian, African, and Latin American economies (see the various national and regional experiences).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
A Literature Review 17 on structural transformation and catch-up. Perspectives drawn from structuralist development economics emphasize the centrality of structural transformation and catch-up, and associated strategies and policies that are of particular relevance for policymaking, a primary aim of this Handbook. The chapter sets the scene in terms of three key aspects: a review of the most important concepts and analytical categories; an overview of literature on industrial hubs; and linking the literature on industrial hubs to broader debates on structural transformation and economic development. Section 2.2 focuses on the genealogy of ‘agglomeration economies’ from classical political economists like Smith and Marshall to contributions from other important disciplines (economic geography, spatial and regional economies, industrial economics etc.), tracing core analytical categories (agglomeration economies, localization economies, thriving cities, etc.) that are relevant or have been widely used to analyse industrial hubs of different kinds. Classic works by Jacobs and Porter, amongst other well-known names in the field, are summarized here. Section 2.3 links to key contributions from structuralist development economics, especially the foundation of heterodox development economics built on key concepts and arguments like the role of manufacturing in long-term growth, the importance of exports, latecomer advantages and the centrality of technological catching up, all combined with a strong case for deliberate state interventions in the form of coordinated industrial policy.6 The idea is to make the case for integrating this well-known literature on industrialization and economic development with classic contributions on industrial hubs. The key emphasis of structuralism is that the structure or composition of the economy should constantly change from low to high productivity and knowledge-intensive economic activities to remain a dynamic whole (Pasinetti 1981, 1993). This perspective is embedded in the production-centric paradigm.7 Section 2.4 provides a chronological overview of empirical contributions to the literature on industrial hubs, partly connected to the theoretical insights explored in sections 2.2 and 2.3, partly as a set of additional ideas and contributions. The chapter ends with section 2.5, an attempt to synthesize some of these insights with a focus on what these mean for core policy priorities, especially on innovation and technological capabilities.
2.2 Agglomeration Economies and Cluster Dynamics By integrating structuralist development economics perspectives with those of economic geography, urban economics, spatial economics, industrial economics, and political economy, we can combine the study of firms, industries, and markets with the interactions between firms’ strategies and industrial hubs. This allows for a better understanding of the 6 See Chang (1994) and Oqubay (2015) on industrial policies. 7 See Best (2001, 2018) on the production-centric paradigm or ‘production capability triad’ consisting of production capability, skills development, and business model (governance). Best highlights how cluster dynamics and spatiation interact with specialization and drive production capability. See also Mazzucato (2018) on wealth creation and extraction paradigms.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
18 Arkebe Oqubay urbanization–industrialization nexus—how space shapes economic activities and the spatial dimension of industrial hubs, and how urban areas are in turn shaped by industrial hubs. Understanding the dynamics of global production networks and global power structures enables policies to be designed that incorporate global and regional dynamics. Research findings from other streams, such as social anthropology, can contribute to a richer understanding of industrial hub dynamics. The distribution of economic activities and people has always varied across space and time, and agglomeration economies are the rule for all economic activities. Economic history shows that the concentration and divergence of economic activities during the development of industrial capitalism has been unparalleled. This has been shaped by shifts in the nature of firms, industries, and technological advancement, the globalization of economic activities, and policy changes across regions, nations, and sub-nations.8 Interdisciplinary research helps us to understand the origins of industrial hubs, their effectiveness, and their selection for use as policy instruments, as well as how those policy instruments are shaped over time. Further questions that lead to a comprehensive understanding of industrial hubs and their outcomes are: what policy instruments in particular were used and why? Why did policy towards industrial parks change over time in particular cases? What were the institutional design features of industrial parks that might help to explain their relative success and failure?
2.2.1 Externalities and the Marshallian Trinity 2.2.1.1 Internal and external economies of scale Industrial production is associated with the concentration of economic, technological, and entrepreneurial activities, which is shaped by the constant motion of forces of concentration (centripetal) and forces of dispersion (centrifugal). The key benefits of agglomeration economies are the reduction of transaction costs and increasing returns to scale, which offer an opportunity for productivity gains (Young 1928). A third equally important benefit is the enhancement of knowledge and innovation that is critical to sustaining long-term competitiveness, which has pivotal implications for production and innovation capabilities. Lastly, fostering linkages and economy of speed is increasingly important in market and production frontiers.9 An understanding of the sources of agglomeration economies is critical for business decisions and also enables the development of policies that promote positive externalities and mitigate negative externalities. Agglomeration economies may be broadly divided into internal and external economies of scale. In classical political economy, Adam Smith ([1776] 1976) argued that specialization and division of labour are central to firm productivity and emphasized that unit cost of output depends on the size of the firm, which is shaped by the size of markets. Smith also highlighted that productive activity and sectors are distinct, and are interdependent amongst different activities by output and inputs. Smith ([1776] 1976: 47) highlighted ‘The greatest improvement in the productive powers of labour, and the greater part of the 8 See Proost and Thisse (2017). 9 Economy of speed is the increasingly shrinking delivery cycle to meet changing tastes and fashions of consumers. While cost and quality are minimum requirements, speed to market has become the prime competitive factor in fashion-driven industries such as apparel and footwear.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
A Literature Review 19 skill, dexterity, and judgement with which it is anywhere directed, or applied, seem to have been the effects of the division of labour.’
2.2.1.2 Marshallian localization economies The second type of agglomeration economies is external economies of scale, which are characterized by externalities of the firms. These may be classified as localization economies, which are economies of scale limited to a specific local industry or sector, or urbanization economies, which are associated with external economies of scale across varied industries in the local economy. Alfred Marshall in Principles of Economics ([1890] 1920) shed light on the principles and law of external economies, and his concept of localization economies (named the Marshallian Trinity) continues to serve as the foundation of the theory of industrial hubs and industrial capitalism. The term ‘localization economies’ defines firms concentrated in the same industry located within a certain vicinity, and in which the cluster provides an advantage to both producers and buyers. Marshall ([1890] 1920: 273) highlights the importance of diverse industries, ‘large towns or large industrial districts in which several distinct industries are strongly developed’, (showing that his concept was not limited to localization economies, namely externality within an industry). Marshall ([1890] 1920: 225) states that ‘when an industry has thus chosen a locality for itself, it is likely to stay there long: so great are the advantages which people following the same skilled trade get from near neighbourhood to one another’.10 The three laws of the Marshallian Trinity are the fundamental advantages gained from the pool of skilled labour, the availability of intermediate inputs and services, and the know ledge and technological spillover. The pool of skilled labour and intermediate inputs and services would reduce transaction costs and contribute to production efficiency or produc tivity gains. Technological spillover and learning have received rather less attention, despite the fact that development of technological and innovation capability is the key determinant for long-term competitiveness. Marshall ([1890] 1920: 156) states: ‘The mysteries of the trade become no mystery . . . good work is rightly appreciated, inventions and improvements in machinery, in processes and the general organization of business have their merits promptly discussed: If one man starts a new idea, it is taken up by others and combined with suggestions of their own; and thus it becomes a source of further new ideas.’11 Unlike localization economies, external economies of scale may involve multiple industries and sectors, where firms in one industry attract firms or economic activities in other industries. This typically occurs in major urban centres and large cities where the size of the city leads to an increase in productivity, rather than in small towns, where localization economies are more common. Initial observations on this subject were made by Ohlin (1933) in Interregional and International Trade.12 10 The European emergence comes from the original Marshallian industrial districts characterized by thousands of small firms clustered by a simple division of labour (such as clothing firms with fabric mills, and accessory and machinery producers). The Third Italy industrial districts featured a blend of cooperation and competition. 11 The Marshallian industrial district is a systemic one in which technology diffuses in rudimentary ways such as learning by doing and its spillover to other workers; science and technology infrastructure and deliberate investment in technological capability play an insignificant role. 12 Industrial hubs enhance agglomeration economies that stem from specialization based on scale and scope. Economy of scale increases concentration with a greater command over reducing unit price with lower coordination costs. Economy of scope increases the division of labour with higher coordination
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
20 Arkebe Oqubay Diversity is the key driver of innovation and knowledge spillover. This is common in technology firms and services, and especially in corporate headquarters that rely on a wide range of services and institutions. Increasing costs of land and labour, increased congestion, and environmental protection are negative externalities that may push manufacturing firms (who primarily rely on firms and services within their industry) to secondary cities.13 This is true for both capital-intensive and labour-intensive industries.
2.2.2 Industrial Clusters and Firm Competitiveness The existing industrial organization and business literature provides an added dimension to our understanding of industrial hubs.14 Michael Porter’s (1990) The Competitive Advantage of Nations enriched the concept by introducing the competitive strategy of firms, supported by empirical evidence on industrial clusters. According to Porter (1990: 253–4), ‘Clusters are geographic concentrations of interconnected companies, specialized suppliers and service providers, firms in related industries, and associated institutions (e.g., universities, standard agencies, and trade associations) in particular fields that compete but also cooperate.’ Porter’s contribution to understanding the dynamics of agglomeration and industrial clusters has been significant. According to Porter, the four key determinants of national competitive advantage are: firm strategy, structure, and rivalry; demand conditions; factor conditions; and the domestic geographical concentration of related industries. The presence of clusters, or related and supporting industries, shapes the creation of factors and upgrades the national advantages, which further stimulates government agencies, educational and research institutions, and firms. Porter (1990: 154) cites empirical evidence from Italy, Germany, and Japan in which ‘competitors in many international successful industries, and often entire clusters of industries, are often located in a single town or region within a nation’.15 Porter adds that while concentration of competitors, customers, and suppliers promotes efficiencies and special ization is important, the influence of geographical concentration is critical for improving product quality, service, and innovation. Clusters attract resources and economic activity from isolated industries, and national competitive advantage rests not only in specific industries but also in clusters or geographic hubs. Porter emphasizes that clusters exist in various forms in different countries, industries, and regions because of variations in institutions, strategy and structure of firms, and information flow systems.
costs, and dissimilarities in production increase the role of scope. The management of the districts offers easier mechanisms to amortize investment and markets for scope-based specialists. This situation, supported by proximity, institutional networks, and intense learning amongst firms, accelerates the linkage dynamics. 13 According to the new economic geography perspective, externalities are influenced by ‘falling transport costs, industrialization, and growing economies of scale’, and partly by the life cycle of industries (Krugman 1993: 52). 14 Edith Penrose’s The Theory of the Growth of the Firm (1959) focuses on the firm as creators of wealth through continuous capability development. See also Chandler (1990). 15 Japanese firms applied the mother-driven industrial hub (such as the automotive industry) to implement a flexible manufacturing system, a variant that differed from the US production system (Best 2001).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
A Literature Review 21 Amongst Porter’s key insights, five are of particular significance. First is his emphasis on interrelationships and dynamic change—‘how each determinant is influenced by the others’—and geographical industry concentration (1990: 131). Second, Porter shows that clusters vary in their form and change due to globalization of competition, increased complexity of the global economy, and the evolution of a knowledgebased society. Third, Porter (1990: 157) builds on the three drivers of agglomeration economies, with more emphasis on innovation and the sustainable sources of competitiveness: ‘geographic concentration of an industry acts as a strong magnet to attract talented people and other factors to it . . . spin-offs have a tendency to locate near the original company . . . proximity increases concentration of information . . . the process of clustering, and interchange among industries in the cluster, also works best when the industries involved are geographically concentrated.’ More importantly, ‘the mutual reinforcement within also leads to surges in innovation (and international competitive position) in whole sectors of a national economy’ (164). Fourth, in contrast to the denial of the role of government (though peripheral) in mainstream literature, Porter suggests that governments play a critical role in the development of industrial clusters by supporting localized initiatives, building around existing institutions, and understanding the interconnectedness of industries, with regional policies focusing on principles of clusters rather than generic support. Fifth, Porter points to cooperation and competition, commonalities and complementarities, social capital and institutional networks, and path dependency (such as the role of historical incidences or change in the location of clusters) as the key features of clusters.16
2.2.3 Productive Cities and Industrial Hubs The role of cities in economic development has been well documented by Paul Bairoch in his (1988) classic work Cities and Economic Development: From the Dawn of History to the Present. In The Economy of Cities and Cities and the Wealth of Nations, Jane Jacobs (1969, 1984) moved beyond the ‘standard’ urban economics discourse, which underlines that cities are hubs of ideas and civilizations, where new is created. Jacobs emphasizes the ‘productive city’ (or ‘thriving cities’) and innovation as key drivers of economic development. Jacobs (1969: 122) states: We now know a few general things about the economy of cities: that cities are settlements where much new is added to older work and that this new work multiplies and diversifies a city’s division of labour; that cities develop because of this process, not because of events outside of themselves; that cities invent and reinvent rural economic life; that developing new work is different from merely repeating and expanding efficiently the production of already existing goods and services, and thus requires different, conflicting conditions from those required from efficient production . . . that the past development of a city is no guarantee of future development because the city can stop vigorously adding new work into the economy and thus can stagnate. 16 See also Breschi and Malerba (2005) and Bottazi et al. (2005) on cluster dynamics, institutional networks, and innovation.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
22 Arkebe Oqubay This view has wider implications than the viewpoint that considers cities for their passive host or urban population rather than productive role. Jacobs recognized the causal relationship between industrialization and urbanization, in which agglomeration of production and services shapes the tapestry of cities, and extends their influence. Hence, the development of industrial hubs is inseparable from that of cities, with significant implications for associated policies.17 In summary, Jacobs’ key perspectives are first, that productive work and innovation are the source of the wealth of nations. Jacobs (1969: 77) highlights: ‘A country’s basic wealth is its productive capacity, created by the practical opportunities people have to add new work’ (emphasis added). Second, Jacobs underlines the importance of manufacturing as the engine of growth: it is the ‘core activity that organizes other economic activities around which the other activities centre’ (1969: 236). This view helps to exploit the intra-sectoral linkages with services and agriculture sectors, and to build technological capabilities. Third, ‘thriving cities’ are characterized by the export-multiplier and import-replacement effects, which emphasize the need to constantly create new products. Fourth, Jacobs highlights the importance of understanding the ecosystem of cities and that urban and economic planning should be viewed as a process of creative thinking and learning rather than a technical exercise.
2.3 Structural Transformation, Catch-up, and Industrial Policy A complementary conceptual theme for understanding industrial hubs and economic development is structuralist development economics, which contributes ideas and concepts that do not appear prominently in the conventional literature of industrial hubs. Industrial hubs must be viewed as an integral part of the industrialization process and from a long-term perspective embedded in economic history rather than a neoclassical abstraction of the ‘efficiency of resource allocation’ laws. Where development of industrial hubs is concerned, three fundamental ideas and concepts underlie the structuralist approach to development economics: structural transformation, catch-up and late development, and industrial policies as framework.
2.3.1 Structural Transformation as the Essence of Economic Development 2.3.1.1 Manufacturing as an engine of growth and structural change Since the first industrial revolution, the development of industrial capitalism has been associated with the development of tertiary sectors, especially in manufacturing, and 17 On roles of cities, the links between industrialization and urbanization, and economic restructuring, see Bairoch (1988), Lefebvre (1996); and Castells (1991), Glaeser (2011), and Henderson (1974, 1988) on urban systems and economic contribution of cities. See also Von Thünen ([1863] 2009) on the logic of towns and economy (‘the Thünen model’), and Losch (1940) and Hoover (1948) on cities and locational theory.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
A Literature Review 23 industrialization. Key contributors to classical political economy, such as Adam Smith and Friedrich List in the eighteenth and nineteenth centuries, and to development economics since the 1950s, have explored manufacturing as an engine of structural change and industrial capitalism making a prime contribution to value and wealth creation principally through production and technology. In the structuralist approach, significant long-term socioeconomic transformation comes about through structural transformation in the form of production, sectoral, and institutional shifts. Structural transformation involves a shift in employment and output from low to high productivity and continuous diversification of economic activities, characterized by a deepening of knowledge and the development of technological capability.18 Much of the literature suggests that it is manufacturing’s special properties that make it the engine of growth and structural transformation (Kaldor 1967; Thirlwall 2013; Ocampo et al. 2009; Amsden 1989). Growth in manufacturing is causally linked to growth in the national economy, and is associated with productivity gains due to dynamic returns to scale. Manufacturing is a prime driver of innovation, and the technological and knowledge spillover has a dynamic effect in creating both direct and indirect employment. Manufacturing creates strong linkages within and across manufacturing industries, and broader inter-sectoral links, offering gains from complementarities. An in-depth understanding of these features calls for a sectoral approach to economic growth with implications for policymaking (Kaldor 1967).19 Technological advancement is linked to sectoral competitiveness and accumulates in specific industries or related sectors. Various researchers suggest that technological innova tion leads to change across sectors in a constant but uneven way, and that these advances diminish the contribution of manufacturing. Nevertheless, manufacturing as a sector continues to be the most important magnet for technological innovation and shows the highest productivity. Other important economic activities that show similar properties to manufacturing include high-productivity services (such as information technology or logistics), high-value horticulture, and industrialization of agricultural activities. These activities rely on capabilities shared with manufacturing and rely on manufacturing to be competitive. Many activities that were traditionally part of manufacturing firms are now carried out by service providers, and the growth of the service sector is linked to that of manufacturing.
2.3.1.2 The strategic role of exports Global exports and international trade have shown rapid growth, especially since the 1970s. This has been a key feature of the globalization of economic activities and is associated with the rise of global production networks. The demand for goods for export creates growth in manufacturing, and the technological composition of these goods is shaped by the sophistication of the export market. There are several reasons for the strategic importance of exports. First, exports are a source of international learning as they offer continuous market opportunities and put competitive pressure on the technological capabilities of firms. 18 See also Monga and Lin (2019) for essays with various perspectives on transformation. 19 See also Hirschman (1958: 170) who highlights the rationale for sectoral distinction: ‘the accumulation of technical knowledge does not occur at the same rate for all economic activities, and receptivity to innovation also varies widely among different sectors, so that at any one time the scope for growth is much larger in some activities than in others.’
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
24 Arkebe Oqubay Exports can thus be a channel for developing dynamism in firms, industries, and national competitiveness.20 A larger export market offers opportunities to increase economies of scale and productivity gains, while the sophistication of the market is an impetus to develop technological capabilities. Second, exports relax the constraints on balance of payment, thereby limiting the scope of economic growth.21 During rapid economic growth the demand for equipment imports and inputs increases, and the constraints on balance of payments becomes the prime restraint on growth dynamics. Third, the export market enables development of import-substitution industries and the exploitation of natural resources as viable economic options. The rise of global production networks and global value chains has affected the global division of labour as well as governance systems for production and markets. This has focused debate on the footloose or embedded nature of clusters, on prioritizing technological capability, and on using foreign direct investment as a catalyst (easier said than done).22 Communications and transport have significantly affected the localized concentration of economic activities and globalization in which transnational firms play a critical role across countries and regions. Some influential thinkers have claimed that in a world of post-2008 economic crisis, global slowdown and protectionism, and the rise of China, an export-led strategy is no longer a policy option for a developing country.23 However, empirical evidence and economic history do not support this proposition although it has implications for strategy design and policies that focus on building competitiveness. In summary, the strategic importance of exports has significant implications for industrial hub policymaking. It is increasingly important to design policies that deepen domestic learning and linkages, and ensure local embeddedness. The acquisition of knowledge and technological capability is critical for improving capacity to climb the global value chain ladder. An understanding of the dynamics of the global production network in a specific industry is crucial for hub development. Continuous improvement and adjustment of export-led industrialization policies should be based on a policy-learning ecosystem in which research centres and learning institutions are linked with industries. Policies that transform the development of human capital and the quality of education, and continuously improve infrastructure, will be essential. Recent research has pointed to the dangers of the ‘middle-income trap’ if nations fail to develop the technological and production capabilities to ensure catch-up.
2.3.2 The Latecomer Advantage and Catch-up Economic history shows that unevenness is an essential characteristic of capitalism and is reflected across countries and regions. The pioneering work of Gerschenkron (1952, 1962) 20 See also Hirschman (1958) on how integration into export markets offers the opportunity to build linkages through ‘import swallowing’. 21 See also Hirschman’s concept of exportability of outputs, of both existing goods and new products. Hirschman (1958: 169) highlights ‘if exportability is high, growth will not lead to balance-of-payments pressures and may actually be accompanied by export surpluses’, and argues the importance of fiscal and exchange policy that are conducive to enhancing exportability. 22 See Akyüz (2017) on the complexity of managing FDI. See also Hymer (1976) on the theoretical foundation of FDI. 23 For instance, Joseph Stiglitz (2018) amongst others.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
A Literature Review 25 and Hirschman (1958) shows that backwardness and late development is not necessarily an evil, and nations may chart their own strategy to catch up by learning from their forerunners and without any ‘prerequisites’: ‘to use institutional innovations to create, mobilize, and concentrate resources with maximum speed and effort . . . the sources of the latecomer advantage was the opportunity to follow a more deliberate and less spontaneous process, in the footsteps of the forerunners . . . a process facilitated by multiple forms of contacts between the latecomer and forerunners’ (Oqubay 2015: 23–4).24 Latecomers successfully catch up if they can create competitive advantages in knowledge and technological innova tions (Kaldorian strategy), although initially it may be necessary to rely on a strategy of the comparative advantage of cheap labour and learning by doing or imitation.25 Recent contributions have shown that catch-up has been a rare phenomenon, industrial strategies may not result in catch-up, and some will be subject to stalled industrialization or premature deindustrialization. Since the 1980s, a stream of research that builds on technological catch-up, the Schumpeterian tradition of creative destruction, and the leading role of learning and technology has filled a critical gap in catch-up literature, focusing on the innovation dimension of industrial hubs (Schumpeter 1934, 1954; Nelson 1993; Dosi 1984; Nelson and Winter 1982, amongst others). Recent research has emphasized that the ‘middle-income trap’ is associated with a low focus on technological capabilities and innovations. It has been emphasized that economic catch-up is shaped and determined by technological leapfrogging, and this may coincide with the development of new technologies.26 In summary, the key implications and contributions of this literature in a continuation of the structuralist development economics tradition are: a) Technical change or technological revolution is a key determinant of economic development, and creative destruction in products and production processes has implications for firms and industries. b) Technological learning at firm level, and technological capabilities and innovation systems at sectoral and economy level, are complementary. c) Innovation is not limited to inventions or products but may occur in various forms including leapfrogging, and production and organizational improvements. d) A holistic perspective has developed the importance of skill formation, production capabilities (more importantly technological management capabilities), and business organization as components of a critical framework for understanding firm and sec toral dynamics (Best 2001, 2018). e) Industrial clusters have been emphasized as central in creating an ecosystem that enables firms’ innovation, together with a dense institutional network in which universities, industrial institutions, research centres, large firms, and start-up firms collaborate. f) Industrial and innovation policies, and national innovation systems at both sectoral and national level, are critical for cluster dynamics. Policies have to be distinctive at different stages of the life cycle of industrial clusters.27 24 See also Oqubay and Ohno (2019). 25 See Schwartz (2010) for the rationale of relying on a Kaldorian or Ricardian strategy. 26 See Lee (2019). 27 See Breschi and Malerba (2005).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
26 Arkebe Oqubay
2.3.3 Industrial Policies as Prime Vehicle of Transformation and Catch-up All advanced nations and developing countries have used industrial policies for economic catch-up and to sustain growth and industrialization, as documented by both economic history and empirical evidence.28 This is especially true for latecomers of the nineteenth and twentieth centuries, including the United States, continental European countries, and Japan. The US government took deliberate action to promote industrialization and economic catch-up. In his Report on Manufactures (1791), Alexander Hamilton, the first US Treasury Secretary, proposed manufacturing as the key to wealth creation, highlighted the essential protection of infant industry, and advocated the introduction of foreign capital and investments, foreign technology, and foreign talent. The notion of infant industry is based on the formidable competition from forerunners, the novelty of the industry, and the importance of national production capacity.29 List (1856: 297) argued that advanced economies adhere to three strategies: ‘First, to prefer constantly the importation of productive power to that of commodities; second, to maintain and carefully protect the development of productive power; third, to import only raw materials and agricultural products, and to export only manufacturing articles.’ Support and protection instruments include both inducements and restrictions: a favourable tax regime, selecting industries based on their economic dynamism, defence interests, and technological considerations. Industrial policies have been effectively used by, for example, the Meiji Restoration, advanced economies of Western Europe, rapidly industrializing economies of East Asia and China, and late latecomers. Effective industrial policies demonstrate a number of key features. First, a policy pattern that focuses on the transformation of productive capacity, constant upgrading of techno logical capability, and steady increase in exports should be embedded in the country’s specific development path, mobilizing and developing abundant resources with an initial reliance on latent comparative advantages while ultimately developing competitive advantage and technological capabilities (Hirschman 1958; Schwartz 2010). Additionally, productive investment and linkage effects have been recognized as a critical sustained dynamic of thriving industrial policies. From a structuralist and catch-up perspective, the key features are: a strategic focus on diversification of economic activities and industries; constant development of productive capacity, technological, and innovation capabilities; and an exportled industrialization orientation that thrives on international learning and international competitiveness. Second, a sectoral focus on economic and entrepreneurial activities with high value, productivity, and technological content is important. This requires an understanding of the inter-sectoral industry linkages, and those between productive sectors and public investment in infrastructure. This has significant implications for cluster dynamics. Third, the selective targeting of industries and economic activities is based on productive criteria, applying the reciprocity principle and rewarding productive activity and constant learning (performance-driven incentives). 28 For instance, see Hamilton (1934), List (1856), Chang (2003), and Reinert (2010). 29 List (1827: 32) suggests that the old country ‘will keep down a rising manufacturing power . . . A new country is, moreover, the less able to contend against the manufacturing power of the old country’.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
A Literature Review 27 Fourth, an important dimension of industrial policy is the complementary roles of market forces and the state. A deliberate process of development and catch-up cannot be left to market forces and calls for an activist or entrepreneurial or developmental state.30 In contrast to the neoclassical perspective that welcomes the temporary fixing of markets and a facilitative role for the state, an alternative view has been that states should play developmental or entrepreneurial roles, creating and shaping market forces, and strategically building production and innovation capabilities.31 Fifth, empirical evidence shows that structure of the industry, political economy, and linkages are important factors in the unevenness and variations in industrial policies and outcomes, necessitating policy learning to improve policymaking and performance.32 Finally, policies related to clustering, industrial hubs, and agglomeration economies have been a critical part of broader industrial policy frameworks at every stage, as evidenced in the industrialization and successful catch-up of East Asian economies.33 Evidence also suggests that the absence of a strategic approach and dissociation of hubs development is most likely to have a less successful outcome.34
2.4 Empirical Perspectives on Industrial Hubs and Recent Literature In industrial policy literature, economic geography and urban economics tend to have been considered separately from industrialization and technological capabilities. Literature on industrial hubs, meanwhile, has attracted various researchers and has focused on different types of industrial hubs: industrial districts, export-processing zones and special economic 30 See Mazzucato (2013). 31 Even the twentieth-century US government is a typical example of a state that has played a strategic and entrepreneurial role in developing new technologies and new high-tech industries, following the Hamiltonian tradition. According to Mazzucato (2013: 21), ‘Economists willing to admit the State has an important role have often argued so using a specific framework called “market failure”. From this perspective the fact that markets are “imperfect” is seen as the exception, which means that the state has a role to play—but not a very interesting one.’ 32 For instance, see Oqubay (2015) and Cheru et al. (2019). 33 Hirschman (1958) observes that concentration is the basic economic rule, with implications for path dependency such as shaping the clustering of investments to initial growth points, and creating regional imbalances. Hirschman (1958: 183–4) highlights that ‘economic progress does not appear everywhere at the same time and that once it has appeared powerful forces make for a spatial concentration of economic growth around initial starting points . . . there can be little doubt that an economy, to lift itself to higher income levels, must and will first develop within itself one or several regional centres of economic strength . . . Thus, in the geographical sense, growth is necessarily unbalanced.’ Such regional clustering patterns have significant implications for political choices and economic policies, for instance in decisions related to public investment and infrastructure. The challenge for policymakers is to recognize this nature of agglomeration economies and the critical role of economic policies throughout the whole industrialization or economic development process. 34 For instance, many initiatives in developing parks in many African countries have been less successful because of the absence of a comprehensive industrial policy design or lack of a strategic approach to hubs development. Many of the Chinese economic and trade cooperation zones in Africa have had mixed outcomes for the same reason.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
28 Arkebe Oqubay zones, and small industrial clusters. This section aims to map or outline the diversity of industrial hubs, their uneven practices and outcomes, and their relevance to policymaking.
2.4.1 The Industrial District Literature (1985 to 2000s) The literature on industrial districts after the 1980s was inspired by the emergence of post-Fordist flexible production ideas as pioneered by Sabel and Zeitlin (1985), and clusters in northern and central regions of Italy, such as Tuscany, Emilia, and Veneto (‘Third Italy’), that were popularized by Becattini (1986), amongst others.35 These clusters were fashiondriven industries like furniture, clothing, footwear, and leather products. Research on these clusters has highlighted the importance of the existing social fabric and social relations, the dominance of small and medium entrepreneurial firms characterized by cooperation and competition, the bottom-up approach, and the unique role of local governments. However, the literature largely ignores production methods (especially in firms dominated by large innovative conglomerates and multinational corporations) and the application of the concept in various sectors. Little emphasis is given to national-level policies and the role of national government in accelerating industrialization and catch-up. In his studies on the footwear cluster in the Sinos Valley, Brazil, Schmitz (1999) demonstrates how collective efficiency has been gained through managing cooperation and competition. Despite the massive wave of industrial hubs in East Asia and China, literature and research on them was very limited before the early 2000s, when Japanese researchers at IDE-JETRO and JICA undertook a considerable programme of research with scholars internationally.36 This literature is not limited to clusters of small firms or those which developed organically but has also extended to policy-driven or hybrid clusters. These insights have benefited from a more comprehensive and systematic approach to industrial clusters, covering sectors such as automotive, light manufacturing, and high-tech industries, and various countries (with a focus on Asia). Empirical evidence has been linked to theoretical perspectives on agglomeration economies, sectoral dimensions, and policy, with research enriched by the divergence of contributors from interdisciplinary backgrounds and various countries of origin. The ‘economies of sequencing’, or Kuchiki’s ‘flowchart’ approach to industrial clusters, has added a new dimension to how policymaking should promote industrial hubs during the life cycle of industrial clusters. This focuses on the role of ‘anchor firms’ popularized in Japan and other East Asian economies by big conglomerates such as Toyota and Samsung, illustrating how related firms are clustered. It is also relevant to the evolution of global production networks where leading buyers/brands or leading technology firms develop a network of component sub-contractors and second-tier firms. The flowchart approach departs from an analysis of markets, whether domestic or export, and underlines the 35 On European industrial districts and Third Italy, see also Brusco (1982), Wilkinson (1983), Pyke et al. (1990). On the technological dimensions and mechanics of industrial districts see UNCTAD (1994), and Schmitz (1999). On the dominance of small firms of industrial districts, see also Goodman and Bamford (1989), Pyke et al. (1990), Asheim (2000), and Becattini et al. (2009). 36 Akifumi Kuchiki amongst others (see Kuchiki and Tsuji 2004). See also Sonobe and Otsuka (2011) who focused on empirical evidence from industrial clusters of small firms in Asia and Africa.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
A Literature Review 29 importance of facilitation of infrastructure, labour supply, and training facilities for building manufacturing production. While commencement of production and economies of scale are critical during the first phase, the approach highlights the importance of building an ecosystem for innovation. Despite its apparent rigidity, the approach has significant implications for policymaking. The roles of government and policies are critical in this framework, and the cluster concept has been extended to sectors such as agriculture and tourism. Another literature stream focuses on the empirical evidence on special economic zones. This literature views SEZs as a tool to ‘enhance industry competitiveness and attract foreign direct investment’, and defines ‘a geographically delimited area administered by a single body, offering certain incentives (generally duty-free importing and streamlined customs procedures, for instance) to businesses which physically locate in the zone’ (FIAS 2008: 1).37 This literature has primarily focused on labour-intensive industries or light manufacturing, and emphasizes the role of the private sector, the business climate, and economic liberalization. However, it overlooks the productive roles of government and the critical importance of industrial policy frameworks. The empirical perspectives of East Asian economies, and the dynamics of industrial hubs and innovation, receive less attention. This literature has contributed to popularizing special economic zones and presenting empirical evidence in an accessible way, aspects that were previously ignored in industrial hubs literature. A parallel approach by Justin Lin’s New Structural Economics (NSE) seeks to integrate the development of special economic zones into country diagnostics,38 focusing on a developing-country strategy oriented to latent comparative advantage, and arguing against strategies that aim to create and alter comparative advantages. This perspective demonstrates that African countries should industrialize and build light manufacturing by attracting foreign direct investment from China, and shows that they can create special economic zones to develop hubs despite constraints of business climate, institutions, and infrastructure. This radical view is optimistic about the possibility of industrialization in Africa, where many influential figures advocate the reverse. Despite its strength, however, this perspective gives insufficient focus to the dynamics of industrial hubs, the importance of innovation capabilities, building competitive advantages, and how states can play a strategic role beyond that of facilitation.
2.4.2 Lessons from the Practice of Industrial Hubs 2.4.2.1 Concept of industrial hubs There is ambiguity around the concepts used to reflect the empirical evidence in various contexts. Agglomeration economies is a fundamental principle of industrial capitalism and the clustering pattern has been shaped by many firm, sectoral, national, and external factors. Multiple terms have been coined to express agglomeration and clustering in the twentieth century, including some that are country or literature specific. ‘Industrial districts’ was used from the time of the first industrial revolution to the time of Alfred Marshall. In the post-World War II era, the term ‘industrial district’ was widely used for the industrial clusters and agglomeration economies in central and northern Italy, the so-called Third Italy. 37 See also Farole and Akinci (2011) and Zeng (2010).
38 For instance, see Lin (2012).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
30 Arkebe Oqubay This term has been linked to post-Fordist flexible production, the dominance of small firms, and bottom-up patterns, but not in other contexts such as innovation hubs in the West, new hubs associated with exports, or manufacturing clusters in newly industrializing economies (see Saxenian 1996; Porter 1990, 2000; Best 2001; Amsden 1989).39 Various terms reflect local types of hubs. The terms ‘industrial complex’ and ‘export processing’ have been widely used in South Korea and Taiwan, China.40 In Singapore, the notion of ‘industrial parks’ or ‘hubs’ is widespread. In China, pragmatic political considerations prompted the use of a new term, ‘special economic zones’. Industrial zones and economic zones have been referred to in Vietnam and the Philippines. ‘Free zones’ has been a popular term in the Middle East and the Gulf countries. Terms such as science park, technology park, or innovation park are specific to technology activities. The term ‘industrial clusters’ has been associated with small enterprises. The term ‘industrial hubs’ captures various types and forms of industrial clusters. Industrial hubs may be defined as the industrial and spatial agglomeration of firms in the same or related industries, where various support institutions and stakeholders (firms, institutions, and government) dynamically interact, cooperate, and compete for mutual gains in productivity, linkage effects, and innovation, and develop their competitive positioning.41
2.4.2.2 The large wave of industrial hubs in Asia The evolution of Silicon Valley or Boston’s Route 128 has inspired many latecomers to explore ways of developing high-tech industries and innovation hubs. Typical examples are the IT industry in Bangalore, India, the Hsiu Science Park which was the most successful science park in Taipei, the Beijing and Shenzhen Science and Technology Parks in China, innovation hubs in Singapore, and the Malaysia Multimedia Super Corridor, which has had mixed outcomes.42 The greatest expansion and development of industrial hubs has occurred in sectors outside the technology and innovation hubs, especially in both the light and heavy manufacturing industry. A core feature of the pattern of industrial hubs has been divergence and unevenness amongst countries and regions. In almost all East Asian economies and China, the approach to industrial agglomeration has been embedded in their industrial policies, with the states playing strategic roles (see Chapter 27 for a summary review of Asian experiences).43 Three generations of industrial hubs made up the East Asian wave: 1960s–1980s, 1980s–2000s, and post–2000. The first generation included pioneers of industrial hubs 39 According to Goodman and Bamford (1989: 21), industrial districts are ‘essentially a territorial system of small and medium-sized firms’. Piore and Sabel (1984) highlight ‘the central feature of the industrial district in the balance between competition and co-operation among firms’ (see You and Wilkinson 1994: 259). Brusco (1989: 259) states that they exhibit the ‘progressive specialization of all the firms working in the same sector in the same area’. The origins of Piore and Sabel’s (1984) work are on post-Fordism as the emphasis on a shift from low (Taylorism and Fordism) forms of work and industrial organization to high-road industrialization, characterized by flexible specialization targeted at raising participation of workers, rapid market response, and supplier networks for eventual horizontal integration. 40 See Amsden and Chu (2003) on industrial upgrading and roles of industrial hubs. 41 This builds on Porter’s definition. 42 See also Saxenian (1996), Saxenian and Hsu (2005), and Rasiah and Nazeer (2016), amongst others. 43 See also various chapters in Part IV.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
A Literature Review 31 and accelerated industrialization and catch-up in South Korea, Taiwan, Singapore, and Malaysia.44 These countries climbed the ladder starting with light manufacturing, heavy and chemical industries, and innovation and high-tech industries.45 South Korea’s approach to building industrialization was somewhat different, with chaebols playing a leading role, while foreign firms were critical for Singapore. Industrial hubs specialized in particular sectors and were characterized by fast learning and an innovative approach. The state built a productive partnership with the private sector, with industrial hubs serving as the core of industrial policies and as the institutional innovation for late industrialization. China was the prime mover in the second wave, its policies of reform and opening initiated under the leadership of Deng Xiaoping after 1978. The Chinese experience is unique in terms of its enormous economic impact, accounting for about two-thirds of global industrial hubs.46 Beginning with an experimental approach, China gradually developed its industrial hubs, shifting from generic industrial hubs aimed at attracting foreign direct investment (overseas Chinese especially from Hong Kong and Taiwan, China) to targeting global industrial and technological groups (Fortune 500 companies) especially through the economic and technological zones in coastal areas, and targeted science and technology parks. Various mechanisms (such as joint ventures and developing national champions) were used to develop local capability. Geographical concentration was initially focused around the eastern coastal area, gradually shifting to central and western regions. Institutional labour regimes supported the manufacturing hubs, improving labour mobility. Industrial hubs were directly led by central government, although provincial and local governments played pivotal roles. In recent years, experimental free economic zones and new areas have been developed. New learning through experimentation (piloting), learning by doing, and emulation (especially from East Asian and other industrialized economies) demonstrated the depth of policy learning. The third wave, after the late 1990s, which covers initiatives in many South and Southeast Asian countries, such as Vietnam and India, has been influenced by the success of China. In addition to the expansion of parks in Asia, a new wave of proliferation has occurred in the Middle East and Latin America. This shift has been accelerated by the globalization of investment and trade, as seen in Bangladesh and Mexico. Africa has lagged behind in the expansion of industrial hubs and the outcome has been less successful, mirroring the weakness of industrial policies in the continent. The global expansion of industrial hubs in the early twenty-first century was shaped by new forces. First, climate change and the negative externality associated with industrialization and urbanization created pressure to adopt policies that promoted sustainability
44 Mauritius, an African outlier, pursued the development of export economic zones as the key strategy of its export-led industrialization in the 1970s, benefiting from a window of opportunity to attract the apparel industry from Hong Kong. 45 Initially, East Asian export-processing zones evolved as platforms for productive investment creating industrial agglomeration. Gradually, with policies that focus on industrial upgrading and innovation, these industrial agglomerations advanced into dynamic industrial clusters where learning and innovation became increasingly prominent. 46 In China, special economic zones were combined with other policies to develop domestic capabil ities (Rodrick 2012).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
32 Arkebe Oqubay and environmental protection.47 Eco-industrial hubs reflect this positive trend, as does consumer pressure on retailers and brands, although this is not so common. Views that regard ecological protection and economic development as mutually exclusive are not yet dominant. Second, the acceleration of technology has significant implications for hub development, broadening the focus on innovation, open systems, and the upgrading of industrial hubs. The ecology of industrial hubs is also shaped by the potential for smart industrial hubs or smart cities in the future. Third is the globalization of production (GVCs), the increasing importance of value addition, domestic linkages and domestic learning, and the use of industrial hubs as a catalyst for broader transformation. Industrial hubs do not exist in a void and are shaped by politics and political economy, which is featured in many aspects. Industrial hubs embody the broader development path and the political-economic goals served by many of the initiatives. The pattern of agglomeration reflects the political landscape of the country: the testing of reforms and opening in China, the ability of the state to manage reciprocity of productive rents from firms, the relative power of industrialists and labour, the allocation and administration of land, and the broader aims of industrial hubs in terms of employment addressing potential tensions/unrest derived from widespread unemployment or underdevelopment (Amsden 1989; Wade 1990). The relative role of the state to direct and execute policies mirrors its political configuration. Hence, industrial hubs have to be seen within the broader social and political complexity. Despite the transformative effect of industrial hubs as a component of industrial policies more broadly, there are many myths around industrial hubs and the roles of the state and market, which mirror the wider debates amongst various schools of thought around industrial policies. From a transformation viewpoint, industrial hubs should not be viewed as instruments of liberalization. Another widely shared myth is the naive and simplistic view that industrial hubs are ‘miracle medicines’ to facilitate industrialization, which overlooks the various complex dimensions of industrial hubs, from the wider issues associated with skills formation to infrastructure development. Despite empirical evidence from East Asia and other regions, the roles of the state are widely regarded as conditional and irrelevant. In view of the inconsistency of policies and outcomes, policy learning is an effective instrument to improve the contribution and outcomes of industrial hubs. In summary, driven by the pursuit of rapid industrialization and technological catch-up, thriving industrial policy and state activism play an essential role in the outcomes of industrial hubs.
2.5 Industrial Hubs and Economic Development: A Synthesis and Conclusions This section presents a summary of the literature review sections, and a synthesis of industrial hubs and economic development. The author acknowledges the fragmentation of the reviewed literature and the variety of analytical categories stemming from different disciplines, intellectual traditions, and individual authors. 47 See Chapters 23, 24, and 25 on sustainability.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
A Literature Review 33
2.5.1 Summary of Literature Review The chapter has reviewed the historical perspectives on industrial hubs and economic development starting from the early days of the first industrial revolution, drawing on a wide range of literature from Adam Smith and Alfred Marshall to other twentieth-century thinkers. It has also examined issues rooted in structural transformation and economic catch-up, and the connection between cluster dynamics and industrial policy frameworks. It has highlighted the strategic importance of manufacturing, exports, sectoral and intersectoral linkages, and domestic learning. The chapter also maps empirical perspectives and an overview of the policies and practices of industrial hubs behind the uneven global practices and outcomes in various national contexts. This chapter attempts to offer a synthesis of industrial hubs, in the process connecting their role in industrialization with how industrial hubs can synergize industrialization and technological catch-up. What is key in catching up is productive capacity driven by policies that focus on accumulation of production capability, international learning, and a constant focus on technological capability (List 1856; Amsden 1989; Reinert 2010; Best 2001, 2018). The middle-income trap is associated with insufficient focus on technological capability.48 It highlights that agglomeration economies and clustering are the dominant characteristics of productive activities in national economies that were at various stages of capitalist development during the late twentieth and early twenty-first centuries. The evolving new environment shapes the dynamics of agglomeration economies in multiple ways. A central concept of economic catch-up is that catching up is driven by accumulation of productive capability, a notion described by various terms such as ‘production capability’ (Best 2018), ‘technological capability’ (Rosenthal), and ‘productive forces’ (Karl Marx), as well as by Hamilton (1934), and List (1856).49 Amsden (1989) emphasizes that latecomers catch up through learning from abroad and learning by doing, and Kim (1997) underscores the distinction between imitation and innovation, and the necessity of innovation and techno logical learning for economic catch-up.50
2.5.2 Innovation and Technological Capabilities Innovation and technological capabilities have increasingly become the key drivers of agglomeration economies. Multiple forces, such as the advance of digital transformation, 48 For instance, see empirical studies by Paus (2017) and Rasiah and Nazeer (2016) on the middleincome trap in Latin American and Asian economies. Rasiah and Nazeer (2016: 167) highlight that in many Asian countries, ‘Technological upgrading was never an integral part of industrial policy. In contrast, the developmental role of the state, with a strong focus of technological catch-up and science-based education, is what propelled South Korea’s leading firms to the world’s technology frontier.’ 49 On various paths of development and economic catch-up of advanced economies (such as the United Kingdom, Germany, Japan, South Korea), see Chang (2003), Amsden (1989), Kim (1997), Kim and Nelson (2000) amongst others. See also Veblen (1915) on the role of institutions in Germany’s industrialization. 50 See also Schumpeter (1934, 1954) on the distinction of innovation such as radical and incremental innovations. See also Rosenberg (1982) and Mowery and Rosenburg (1989) on technological capabilities, and Nelson and Winter (1982) on developing innovation systems.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
34 Arkebe Oqubay the fusion of technologies, and the accelerated pace of innovation and technological advancement, have brought this about. In contrast to the importance of the manufacturing process during the first industrial revolution, it is innovations in products, production processes and organizations, and the focusing of resources on research and development that have gained the utmost importance in the twenty-first century. The product development phase has been gradually shortened to almost 12 months, with huge implications for in-house innovation capabilities, investment in basic research by governments and the private sector, and stronger industry-learning institutions linkages. Innovation and technological capabil ities have become more important than productivity gains or linkages. Shifts in productivity gains and economies of scale have been accompanied by significant and unforeseen expansion in the second half of the twentieth century, especially after the 1970s. Economic globalization and the internationalization of production have been evidenced by the increase in international trade and the expansion of foreign direct investment. Many leading thinkers have illustrated this transformation as global production networks (GPNs), a global business revolution (GBR), and global value chains (GVCs).51 This shift at a global scale has been characterized by concentration of production networks in some regions (such as the Pacific Rim), and some countries such as the Asian Tiger economies, China, and Mexico. Leading firms in new industries (such as semi-conductors and electronics) have typically emerged within the contours of the landscape.52 The development of linkage effects has gained increased currency in some industries as speed to market becomes the prime competitive factor, shaping industrial hubs to reflect this requirement. This is true in many fashion-driven industries such as apparel and clothing, where advantages in global sourcing of inputs by leading multi-national corporations and the localization of inputs through the vertical integration of production have come to co-exist. This underlines the opportunities for building backward and forward production linkages in multiple sectors in developing countries. It also highlights the importance of economy of speed in combination with the economy of scale and innovation as key drivers of industrial clustering and agglomeration economies. In 2014, Samsung opened its new facility for touch screens near Hanoi, Vietnam. This new cluster involves networks of component suppliers, both first and second tier, employing approximately 50,000 workers. This is linked with the research and development facilities at its headquarters and the supply of components across the Chinese border, despite tensions in the region. 51 For instance, see Dicken (2015), Nolan (2003), UNCTAD (2013), World Bank (2009), and Gereffi (2018). 52 A typical example that shows agglomeration of economies of a different magnitude featuring a firm-level strategy is Foxconn. Foxconn was founded in Taipei in 1974. The company later expanded to mainland China with improved production facilities. While building its main facility in Shenzhen Science and Technology Park, it expanded to over thirty technology and manufacturing industrial parks employing nearly 1.5 million workers, half a million of whom were based in Shenzhen. Foxconn is currently the major manufacturer of not only iPhones for the global market, but also of Amazon’s Kindle and Huawei devices. It has also acquired the Nokia manufacturing facilities in Vietnam and India.52 In 2017, it generated a total revenue of US$150 billion and accounted for 3.9 per cent of China’s exports. It has emerged as the largest global device manufacturer and is gradually transforming itself from an original device manufacturer (ODM) to the partial development of its own branded products. To serve this strategy, it has acquired leading technology firms such as Sharp, Nokia, and Toshiba PC. This strategy is embedded in a regime of unprecedented economy of scale, development of a responsive supply chain, and building a learning ecosystem that is conducive to constant upgrading.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
A Literature Review 35 In summary, industrial hubs have to generate agglomeration economies in a new environment where innovation and technological capabilities have become essential and should serve as the source of domestic learning. It also shows that linkage dynamics and economy of speed become the sources of domestic linkages, in a globalized production network which offers economy of scale and productivity gains.
2.5.3 Strategies of National Economies and Firms It is conceivable to generalize that in almost all industrial hubs and clusters the key features are in full motion. Cooperation and competition are prime features of industrial hubs. Cooperation amongst firms is driven by common interests that contribute to promote the competitiveness of all firms. This may be common infrastructure, shared institutional supports, an intention to improve bargaining position, or collaboration in knowledge and information sharing. There may also be a common vision on issues such as environmental protection and sustainability.53 Cooperation is not limited to firms but also covers other institutions such as knowledge institutions, government regulatory agencies, and private-sector industrial associations. Firms compete within the global competitive context and each firm has to put maximum effort into strengthening its competitive position within and beyond the hub. Another key characteristic is the dense network of institutions that are found in the cluster, which typically comprise universities, accreditation bodies, industrial associations, and so on. Moreover, the production system allows for the development of competitivecapability corporate strategies. In recent periods, the development of industrial hubs has been deployed to facilitate broader strategic aims, which were not common in the twentieth century. Emerging economies have also used hub development as a key vehicle for their internationalization strategy, for example China’s ‘Go Global’ strategy for developing ‘trade and economic cooperation zones’ in Africa and other countries. Industrial hubs are also continually being used to accelerate the outflow of foreign direct investment and to expand trade. However, the outcome has been mixed across different countries depending on the industry and existing policies. Digital transformation has started to impact on hub development, with concepts such as smart industrial parks or smart cities. Urban dynamics and ecosystems have also proved to be sources of urban agglomeration economies, labour supply, construction and (housing) property development, and urban planning; for example, the Dubai service sector—healthcare, education, innovation, logistics, and aviation—and its integrated approach to urbanization and hub development. The scarcity of infrastructure, the weak business environment, and a desire to provide better government services and incentive packages have contributed to this process.
Acknowledgements The author is grateful to Professor Rajah Rasiah and Dr Carlos Oya for their constructive comments, and to Deborah Kefale, Samuel Arkebe, and Binyam Arkebe for inputs to improve the draft, and for their continued support. 53 See Chapters 23, 24, and 25 on sustainability.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
36 Arkebe Oqubay
References Akyüz, Yilmaz (2017) Playing with Fire: Deepened Financial Integration and Changing Vulnerabilities of the Global South. Oxford: Oxford University Press. Amsden, Alice (1989) Asia’s Next Giant: South Korea and Late Industrialization. Oxford: Oxford University Press. Amsden, Alice and Wan-wen Chu (2003) Beyond Late Development: Taiwan’s Upgrading Policies. Cambridge, MA: The MIT Press. Asheim, Bjorn T. (2000) ‘Industrial Districts: The Contributions of Marshall and Beyond’, in Gordona Clark, Maryann Feldman, Meric Gertler (eds) The Oxford Handbook of Economic Geography. Oxford: Oxford University Press, pp. 413–31. Bairoch, Paul (1988) Cities and Economic Development: From the Dawn of History to the Present. Chicago, IL: University of Chicago Press. Becattini, Giacomo (1986) ‘Small Firms and Industrial Districts: The Experience of Italy’, Economia Internazionale 39(2–4): 98–103. Becattini, Giacomo, Marco Bellandi, and Lisa De Propris (eds) (2009) A Handbook of Industrial Districts. Cheltenham: Edward Elgar. Best, Michael (2001) The New Competitive Advantage: The Renewal of American Industry. Oxford: Oxford University Press. Best, Michael (2018) How Growth Really Happens: The Making of Economic Miracles through Production, Governance, and Skills. Princeton, NJ: Princeton University Press. Bottazi, Giulio, Giovanni Dosi, and Giorgio Fagiolo (2005) ‘On Sectoral Specificities in the Geography of Corporate Location’, in Stefano Breschi and Franco Malerba (eds) Clusters, Networks, and Innovation. Oxford: Oxford University Press, pp. 54–79. Breschi, Stefano and Franco Malerba (2005) Clusters, Networks, and Innovation. Oxford: Oxford University Press. Brusco, Sebastiano (1982) ‘The Emilian Model: Productive Decentralisation and Social Integration’, Cambridge Journal of Economics 6: 167–84. Brusco, Sebastiano (1989) ‘A Policy for Industrial Districts’, in Edward Goodman and Julia Bamford (eds) Small Firms and Industrial Districts in Italy. London: Routledge, pp. 259–69. Castells, Manuel (1991) The Informational City: Economic Restructuring and Urban Development. Chichester: John Wiley & Sons. Chandler, Alfred (1990) Scale and Scope: The Dynamics of Industrial Capitalism. Cambridge, MA: Harvard University Press. Chang, Ha-Joon (1994) The Political Economy of Industrial Policy. London: Macmillan. Chang, Ha-Joon (2003) Kicking the Ladder: Developing Strategy in Historical Perspectives. London: Anthem Press. Cheru, Fantu, Christopher Cramer, and Arkebe Oqubay (eds) (2019) The Oxford Handbook of the Ethiopian Economy. Oxford: Oxford University Press. Dicken, Peter (2015) Global Shift: Reshaping the Global Economic Map in the 21st Century. 7th edition. New York: The Guilford Press. Dosi, Giovanni (1984) Technical Change and Industrial Transformation. London: Macmillan. Farole, Thomas and Gokhan Akinci (eds) (2011) Special Economic Zones: Progress, Emerging Challenges, and Future Directions. Washington, DC: World Bank. FIAS (2008) Special Economic Zones: Performance, Lessons Learned, and Implications for Zone Development. Washington, DC: World Bank. Gereffi, Gary (2018) Global Value Chains and Development. Cambridge: Cambridge University Press. Gerschenkron, Alexander (1952) ‘Economic Backwardness in Historical Perspective’, in Bert F. Hoselitz (ed.) The Progress of Underdeveloped Areas. Chicago, IL: Chicago University Press, pp. 3–29. Gerschenkron, Alexander (1962) Economic Backwardness in Historical Perspective. Cambridge, MA: Harvard University Press.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
A Literature Review 37 Glaeser, Edward (2011) Triumph of the City. London: Macmillan. Goodman, Edward and Julia Bamford (eds) (1989) Small Firms and Industrial Districts in Italy. London: Routledge. Hamilton, Alexander (1934) Papers on Public Credit, Commerce and Finance, ed. Samuel McKee. New York: Columbia University Press. Henderson, Vernon (1974) ‘The Sizes and Types of Cities’, American Economic Review 64(4): 640–56. Henderson, Vernon (1988) Urban Development: Theory, Fact, and Illusion. Oxford: Oxford University Press. Hirschman, Albert (1958) The Strategy of Economic Development. New Haven, CT: Yale University Press. Hoover, Edgar (1948) The Location of Economic Activity. New York: McGraw-Hill Book Company. Hymer, Stephen H. (1976) The International Operations of National Firms: A Study of Direct Foreign Investment. Cambridge, MA: The MIT Press. Jacobs, Jane (1969) The Economy of Cities. New York: Vintage Books. Jacobs, Jane (1984) Cities and the Wealth of Nations: Principles of Economic Life. New York: Vintage Books. Kaldor, Nicholas (1967) Strategic Factors in Economic Development. Ithaca, NY: Cornell University Press. Kim, Linsu (1997) Imitation to Innovation: The Dynamics of Korea’s Technological Learning. Cambridge, MA: Harvard University Press. Kim, Linsu and Richard Nelson (2000) Technology, Learning, and Innovation: Experience of Newly Industrializing Economies. Cambridge: Cambridge University Press. Krugman, Paul (1993) Geography and Trade. Cambridge, MA: MIT Press. Kuchiki, Akifumi and Masatsugu Tsuji (2004) Industrial Clusters in Asia: Analyses of their Competition and Cooperation. New York: Palgrave Macmillan. Tokyo: IDE-JETRO. Lee, Keun (2019) The Art of Economic Catch-up: Barriers, Detours, and Leapfrogging in Innovation Systems. Cambridge: Cambridge University Press. Lefebvre, Henri (1996) Writings on Cities. Translated by Eleonore Kofman and Elizabeth Lebas. Oxford: Blackwell Publishing. Lin, Justin Yifu (2012) New Structural Economics: A Framework for Rethinking Development and Policy. Washington, DC: World Bank. List, Friedrich (1827) ‘Outlines of American Political Economy’, in Series of Letters. Letter VII: Reading. 22 July 1827. Philadelphia, PA: Samuel Parker. List, Friedrich (1856) The National System of Political Economy. Volumes I–IV. Memphis, TN: Lippincott. Losch, August (1940) The Economics of Location. Jena: Fischer. English translation, New Haven, CT: Yale University Press. Marshall, Alfred ([1890] 1920) Principles of Economics. London: Macmillan. Mazzucato, Mariana (2013) The Entrepreneurial State: Debunking Public vs Private Sector Myths. London: Anthem Press. Mazzucato, Mariana (2018) The Value of Everything: Making and Taking in the Global Economy. London: Allen Lane. Monga, Célestin and Justin Yifu Lin (2019) The Oxford Handbook of Structural Transformation. Oxford: Oxford University Press. Mowery, David and Nathan Rosenburg (1989) Technology and the Pursuit of Economic Growth. Cambridge: Cambridge University Press. Nelson, Richard (1993) National Innovation Systems. Oxford: Oxford University Press. Nelson, Richard and Sidney Winter (1982) An Evolutionary Theory of Economic Change. Cambridge, MA: The Belknapp Press of Harvard University Press. Nolan, Peter (2003) ‘Industrial Policy in the Early 21st Century: The Challenge of the Global Business Revolution’, in Ha–Joon Chang (ed.) Rethinking Development Economics. London: Anthem Press, pp. 299–321. Ocampo, José A., Cordina Rada, and Lance Taylor (2009) Growth and Policy in Developing Countries: A Structuralist Approach. New York: Columbia University Press.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
38 Arkebe Oqubay Ohlin, Bertil (1933) Interregional and International Trade. Cambridge: Cambridge University Press. Oqubay, Arkebe (2015) Made in Africa: Industrial Policy in Ethiopia. Oxford: Oxford University Press. Oqubay, Arkebe and Kenichi Ohno (2019) How Nations Learn: Technological Learning, Industrial Policy, and Catch-up. Oxford: Oxford University Press. Pasinetti, Luigi (1981) Structural Change and Economic Growth: A Theoretical Essay on the Dynamics of the Wealth of Nations. Cambridge: Cambridge University Press. Pasinetti, Luigi (1993) Structural Economic Dynamics: A Theory of the Economic Consequences of Human Learning. Cambridge: Cambridge University Press. Paus, Eva (2017) ‘Innovation Strategies Matter: Latin America’s Middle-income Trap Meets China and Globalisation’, The Journal of Development Studies, DOI: 10.1080/00220388.2019.1595600. Piore, Michael J. and Charles F. Sabel (1984) The Second Industrial Divide: Possibilities for Prosperity. New York: Basic Books. Porter, Michael (1990) The Competitive Advantage of Nations. New York: Free Press. Porter, Michael (2000) ‘Locations, Clusters, and Company Strategy’, in Gordon Clark, Maryann Feldman, and Meric Gertler (eds) The Oxford Handbook of Economic Geography. Oxford: Oxford University Press, pp. 253–74. Proost, Stef V. and Jacques-François Thisse (2017) ‘What Can Be Learned from Spatial Economics?’. Higher School of Economics Research Paper No. WP BRP 167/EC/2017. Available at SSRN: https:// ssrn.com/abstract=3003873 or http://dx.doi.org/10.2139/ssrn.3003873. Pyke, Frank, Giacomo Becattini, and Werner Sengenberger (1990) Industrial Districts and Inter-Firm Co-operation in Italy. Geneva: International Institute for Labour Studies. Rasiah, Rajah and Nazia Nazeer (2016) ‘Comparing Industrialization in Pakistan and the East Asian Economies’, Lahore Journal of Economics 21(Special Edition): 167–92. Reinert, Erik S. (2010) How Rich Countries Got Rich and How Poor Countries Stay Poor. London: Constable. Rodrick, Dani (2012) The Paradox of Globalization: Why Global Markets, States, and Democracy Can’t Coexist. Oxford: Oxford University Press. Rosenberg, Nathan (1982) Inside the Black Box: Technology and Economics. Cambridge: Cambridge University Press. Sabel, Charles and Jonathan Zeitlin (1985) ‘Historical Alternatives to Mass Production’, Past & Present 108: 133–76. Saxenian, AnnaLee (1996) Regional Advantage: Culture and Competition in Silicon Valley and Route 128. Cambridge, MA: Harvard University Press. Saxenian, AnnaLee and Jinn-Yuh Hsu (2005) ‘The Silicon Valley–Hsinchu Connection: Technical Communities and Industrial Upgrading’, in Stefano Breschi and Franco Malerba (eds) Clusters, Networks, and Innovation. Oxford: Oxford University Press, pp. 235–60. Schmitz, Hubert (1999) ‘Collective Efficiency and Increasing Returns’, Cambridge Journal of Economics 23: 465–83. Schumpeter, Joseph (1934) The Theory of Economic Development. Cambridge, MA: Harvard University Press. Schumpeter, Joseph (1954) Capitalism, Socialism, and Democracy, 3rd edition. New York: Harper & Row. Schwartz, Herman (2010) States versus Markets: The Emergence of a Global Economy. New York: Palgrave. Smith, Adam ([1776] 1976) An Inquiry into the Nature and Causes of the Wealth of Nations. Volumes I and II. London: W. Strahan and T. Cadell. Sonobe, Tetsushi and Keijiro Otsuka (2011) Cluster-Based Industrial Development: A Comparative Study of Asia and Africa. Basingstoke: Palgrave Macmillan. Stiglitz, Joseph (2018) Speech at UN-WIDER, Nariobit. Thirlwall, Anthony (2013) Economic Growth in an Open Developing Economy: The Role of Structure and Demand. Cheltenham: Edward Elgar.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
A Literature Review 39 UNCTAD (1994) ‘Technological Dynamism in Industrial Districts: An Alternative Approach to Industrialization in Developing Countries?’. Papers and Synthesis of Discussions of a Symposium on Industrial Districts and Technology. Geneva: United Nations. UNCTAD (2013) Global Value Chains: Investment and Trade for Development. World Investment Report. Geneva: United Nations. UNCTAD (2019) World Investment Report, WIR 2019, Chapter IV Special Economic Zones. Geneva: UNCTAD. Veblen, Thorstein (1915) Imperial Germany and the Industrial Revolution. London: Macmillan. Von Thünen, Johann ([1863] 2009) The Isolated State in Relation to Agriculture and Political Economy. Translated by Ulrich Van Suntum. New York: Palgrave Macmillan. Wade, Robert (1990) Governing the Market: Economic Theory and the Role of Government in East Asian Industrialization. Princeton, NJ: Princeton University Press. Wilkinson, F. (1983) ‘Productive Systems’, Cambridge Journal of Economics 7: 413–29. World Bank (2009) Reshaping Economic Geography. WDR Report 2009. Washington, DC: World Bank. You, Jon-Il and Frank Wilkinson (1994) ‘Competition and Co-operation: Toward Understanding Industrial Districts’, Review of Political Economy 6(3): 259–78. Young, Allyn (1928) ‘Increasing Returns and Economic Progress’, The Economic Journal 38: 527–42. Zeng, Douglas Z. (ed.) (2010) Building Engines for Growth and Competitiveness in China: Experience with Special Economic Zones and Industrial Clusters. Washington, DC: World Bank.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
chapter 3
Heterodox A pproach e s to I n dustr i a l Policy a n d the Im plications for I n dustr i a l H u bs Christopher Cramer and Fiona Tregenna
3.1 Introduction In this chapter we present an overview of the main economic ideas underpinning the case for industrial policy and we try to show how these relate specifically to thinking about, and policy approaches to, industrial hubs. The chapter sets the scene for subsequent chapters of the Handbook. We put these ideas within a historical context. Given that industrial policy has lately become newly fashionable, attracting support across a broader range of economists and political parties than had been the case at least for decades, we also put the ideas within the context of debates and differences of assumption and method in economics. One theme running through the discussion is the difficulty that economic analysis and theory has often had in capturing realities of production. First, despite the obvious fact of increasing returns in many economic activities, many economists preferred to assume they did not exist. Second, economists work with a standard simple distinction amongst sectors, and this is especially important for those economists arguing for the special significance of manufacturing. But changes in global production have meant that service, manufacturing, and agricultural (as well as broader natural resource) activities increasingly blend into each other. This is important for analysis—for example, of how important and how real premature deindustrialization is in developing countries—and for policy design. Both these points are relevant to industrial hubs. First, dynamic returns, poorly addressed in mainstream economics, often unfold with particular intensity through spatial proximity and agglomeration. Second, if there is a logic to supporting the development of industrial hubs, then that logic may apply across a broader range of economic activities than those
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Heterodox Approaches to Industrial Policy and the Implications 41 traditionally (and increasingly narrowly) conceived within the category of manufacturing industry. The chapter emphasizes structuralist economic approaches to industrial policy, and it draws on these to set out some of the main arguments for why industrial hubs and zones may play a particularly dynamic role in structural change (see Chapter 2 for a broader set of perspectives on industrial hubs and agglomeration). But we also highlight that the benefits of hubs do not accrue automatically and that, for example, even if they generate jobs and foreign exchange, they do not necessarily stimulate wider structural change.
3.2 Industrial Policy in History and Theory Rarely has industrial policy been implemented simply as the application of theory. Rather, the true history of industrial policy has involved pragmatic support of particular economic activities in pursuit of wider goals, often propelled by nationalist concerns with regional rivalry, global competition, and ‘catching up’. An example is the use of industrial policy in the United Kingdom after 1979. The United Kingdom, especially during the Thatcher years, is often regarded as a country where the government, adopting neoliberal economic ideas and policies, eschewed industrial policy, with the result that deindustrialization has been deeper and more rapid in the United Kingdom than in most other advanced capitalist economies. There were many policies that affected industry, not least the assault on organized labour and the ‘big push’ for privatization. There were special initiatives designed, for example, to encourage Japanese investors to commit to the auto industry. But none of these resembled what most people have in mind as industrial policy, except perhaps for one significant policy intervention. Margaret Thatcher, resisting advice arguing that better value for money could be had by importing, strongly backed government support to ensure that major new military hardware platforms ‘will be made in Britain’.1 One result was that the largest single manufacturing employer in the United Kingdom for decades afterwards continued to be British Aerospace, the leading arms manufacturer in the country.2 But there are of course other examples of nationalist pragmatism driving industrial policy. Alexander Hamilton, in the United States, famously argued in favour of trade protection to support domestic US manufacturing expansion, pointing out that the United Kingdom had done just that until it no longer suited its leading interests. Even Friedrich List, often cited as one of the key early economists advocating infant industry protectionist policies, derived much of his thinking from his observations of the ‘American system’ in the United States and the protectionist policies advocated and implemented in France (Henderson 1982). And industrial policy in South Korea and Taiwan was driven largely by political compulsion: as Doner et al. (2005) argue, the combination of internal 1 Personal communication from a senior defence procurement advisor to Mrs. Thatcher. 2 See https://www.baesystems.com/en-uk/what-we-do-rgb/what-we-do-in-the-uk-rgb/uk-manufacturing.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
42 Christopher Cramer and Fiona Tregenna and external threats was fundamental to the adoption of industrial policies and support for emerging national champions. It is nonetheless also important to stress the role of a non-mainstream tradition of economics in building the case for industrial policy, and in setting out what the appropriate objectives and instruments of industrial policy should be. The analytical equipment developed within neoclassical economics has often been used in industrial and other policies. But the case for industrial policy overall, as a key to promoting faster sustained economic growth and structural change, rests fundamentally on non-orthodox economic arguments; the next section focuses specifically on structuralist arguments. And at the heart of these arguments is the far greater attention given outside the economics mainstream to increasing returns, and the implication of this for sectoral policy. There has been a great deal of ambivalence about increasing returns in economics, much of which has had to do with a tension between observing reality and developing a particular method of economic analysis. Adam Smith (1994) clearly noted the significance of increasing returns, early in The Wealth of Nations, but his own discussion veered away from their implications. Kaldor used to quip that economics went wrong after Book 1, Chapter 4 of The Wealth of Nations, when Smith dropped the assumption of increasing returns. Something similar occurs in Marshall’s Principles of Economics (Thirlwall 2002: 7). Best (2018) points out how the notion of increasing returns shrinks in significance through successive editions of the Principles, while the emphasis on the Law of Diminishing Returns grows stronger. But Marshall was too attached to observing the real world to give up on the idea and even later editions still capture the power of increasing returns in modern economies: ‘while the part which nature plays in production shows a tendency to diminishing return, the part which man plays shows a tendency to increasing return’ (Marshall 1920: 196). The idea remained uncomfortable and it was increasingly sidelined in marginalist, neoclassical economics. As Best argues, perhaps the chief reason was that increasing returns were too difficult to model in an analytical framework in which not only must everything tend to market equilibrium but also the models must be mathematically tractable: the equations had to resolve. Only with the advances (or as many would have it, the partial recovery of insights) of endogenous growth theory, did mainstream economists manage at least partially (as, for example, in Krugman’s trade models) formally to capture the significance of increasing returns.3 As we argue below, despite an almost grudging and limited folding into some orthodox economic analysis of increasing returns, the policy implications have tended to be too much to bear: economists noting the theoretical possibility and even the historical significance of increasing returns have, like nervous horses, refused the jump to the field of drawing clear policy recommendations for, especially, lower-income economies. Outside the main canon of economics, increasing returns have played a greater role. One of the seminal developments was Young’s (1928) attention to the distinction between static increasing returns, largely captured simply within a firm, and dynamic increasing returns. The latter evolved through ‘intimate connections’ amongst firms in a form of feedback loop that propelled the progressive specialization of individual firms, and differentiation amongst firms, in a cumulative process. Clearly, this has important implications for the dynamic potential of hubs, zones, parks, and other forms of organized productive agglomeration, as
3 Kurz and Salvadori (1998).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Heterodox Approaches to Industrial Policy and the Implications 43 discussed below. Increasing returns and ‘cumulative causation’, or an economic understanding less fixated on equilibrium, also played a central role in later work by Kaldor and o thers.4 Kaldor’s growth laws (see section 3.3 below) turn on increasing returns. And one of the key sources of increasing returns in manufacturing is ‘learning by doing’, a process closely related to Marshall’s own discussion of manufacturing firms. Marshall (1920: 196) defined the law of increasing returns thus: ‘An increase of labour and capital leads generally to improved organization, which increases the efficiency of the work of labour and capital’. And Marshall—like Smith before him and Hirschman and many others later—identified particular scope for the logic of increasing returns in particular activities, mainly in manufacturing and transport: ‘in most of the more delicate branches of manufacturing, where the cost of raw material counts for little, and in most of the modern transport industries the law of increasing return acts almost unopposed’ (Marshall 1920). Kaldor emphasized perhaps more clearly than other economists before him the central significance of manufacturing industry in the growth process. Combined with a less doctrinal commitment to comparative advantage than was held amongst orthodox economists, this produced the case for industrial policy. While this applies in all economies, it does so perhaps especially in lower-income economies where promotion of manufacturing might be the key to catching up. That was then the spur for the spate of work on industrial policy in, above all, East Asia, which challenged the then influential idea that East Asian economies had achieved unpre cedented rates of growth and structural change on the basis of laissez-faire economic management. Amsden (1989), Wade (1990), Woo-Cumings (1999), Chang (2003), and others effectively put paid to such ideas. Later work, more palatable to orthodox economists perhaps, reinforced their insights. Thus Rodrik (2012) showed that there is no ‘automatic convergence’ generally amongst economies, but that there is automatic convergence within manufacturing. Lane (2017) also reinforced arguments that state intervention in manufacturing was decisive in generating South Korean growth. Krugman (1987) too had seen the economic case for intervention based on increasing returns but had argued strongly that largely political reasons meant such intervention was unlikely to succeed. In short, the line of analysis in economics that gives greatest emphasis to the special characteristics of manufacturing industry in growth and development is that which gives greatest attention to increasing returns and then identifies a greater scope for these in manufacturing than in other sectors. That then forms the basis of a case for industrial policy, which typically also entails confounding the ‘law’ of comparative advantage, especially when understood in a static sense. Those advancing the case for industrial policy have generally relied on a different method ological approach to the mainstream as well as building on different assumptions. ‘Natura non facit saltum’, goes the epigraph to Marshall’s Principles; this stands not just for a ‘framebelief ’ that economic change advances by small and gradual steps, inspired by the influence of Darwinian evolutionary theory, but also for an analytical method, inspired in Marshall by Cournot’s development of differential calculus.5 Marshall captured and represented, of course, a broader fundamental shift in economics, the marginalist revolution that produced a lasting form of economic analysis. And this vision of economics as a particular kind of 4 On Kaldor’s critique of equilibrium economics see King (2009: 162–6). 5 Natura non facit saltum is a Latin phrase meaning nature does not advance in leaps and bounds.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
44 Christopher Cramer and Fiona Tregenna science, aping natural sciences and codified in the behavioural axioms of rational-choice utility maximization, solidified into an increasingly mathematical and deductive framework through the twentieth century. ‘Rigour’ came to be regarded exclusively in terms of mathematical tractability (Backhouse 2010; Lawson 1997). But natura, or at least the economy, does appear sometimes to facit saltum, to advance by leaps and bounds, as well as by smaller steps. Non-linearity at times trumps Marshall’s ‘Principle of Continuity’. Industrialization in Japan, South Korea, and Taiwan was not simply a matter of applying marginal gains, like the advantage secured by the Sky cycling team.6 That is why these experiences draw such imagery as ‘miracles’ and ‘technological leaps’ or, in one recent contribution, the ‘moonshot’ approach and the ‘leapfrog approach’. For instance, Ethiopia’s dramatic episode of economic growth from the early 2000s, though of course the ground was prepared in some ways by institutional changes in the previous century, was in many ways a sharp break with prior economic history in the country (Cheru et al. 2019). Many technical and organizational innovations in industry (and in wider economies) have come about through the terrible saltus of warfare. And so on. Most of the advances in our understanding of industrial policy have drawn upon analytical methods better adapted to capture insights into saltus as well as gradual change.7 And these have included more inductive methods. Though they may well search for statistical patterns these analyses are more likely to eschew the thinking by averages that dominates most economics. Cherif and Hasanov (2019: 6) nicely quote a Nobel laureate physicist arguing: ‘Much of the real world is controlled as much by the tails of distributions as [by] means or averages . . . by the exceptional not the common place; by the catastrophe, not the steady drip . . . we need to free ourselves from “average” thinking’. The key to rigour is organized attentiveness rather than a single specific form of observation or modelling, such social scientists would argue. And cross-country growth regressions based on standard growth models appear unsuited to such close observation. Hausmann et al. (2005), for example, find that most episodes of growth acceleration have not been preceded or accompanied by improvements in the determinants included in standard linear growth models—the models cannot predict the growth acceleration. Best (2018) traces a genealogy of a more inductive approach, systematically deriving principles from close observation of firms, clusters, and policies, from Babbage in the nineteenth century, through Penrose in the midtwentieth, to Schumpeter and Jacobs, a line we might extend to Amsden and others in the late twentieth and early twenty-first century. The distinction between theories derived from abstract axioms and principles, on the one hand and, on the other, policies derived from practical, typically political compulsions was beautifully captured in the contrast between Adam Smith’s Wealth of Nations and Alexander Hamilton’s Report on the Subject of Manufactures. McNamara (1998) highlights the key difference between Smith and Hamilton as methodological, citing Hamilton’s argument that Smith’s way of theorizing produced propositions that were ‘geometrically true’ but ‘practically false’. 6 The Sky cycling team and the British Olympic cycling team (they shared a head coach) famously dominated international cycling and claimed that their model was based on ‘marginal gains’. See Cherif and Hasanov (2019) on the contrasts between moonshot, leapfrog and snail’s pace industrialization strategies; and the debate in Lin and Chang (2009). 7 Against Marshall’s Darwinian principle of continuity there is the biologist Stephen Jay Gould’s notion of ‘punctuated equilibrium’, which has also been adopted in social sciences.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Heterodox Approaches to Industrial Policy and the Implications 45 These distinctions—of method, assumption, and analysis—matter. In recent years, industrial policy has become fashionable amongst mainstream economists, rather than remaining the preserve of heterodox economists. The idea of industrial policy has even been revived in the heartland of neoliberal economic policies, where governments—including the United Kingdom’s—for decades disparaged industrial policy (at least officially).8 When even IMF authors champion industrial policy there may appear to be cause for relief and the hope of progress (Cherif and Hasunov 2019). But when economists of all stripes profess to support the case for industrial policy, they do not necessarily mean the same thing, which produces confusion and may mislead policy officials. One economist’s ‘facilitative state’ in pursuit of industrialization (Lin and Chang 2009) is another economist’s ‘reciprocal control mechanism’ industrial policy (Amsden 2001). One economist’s ‘horizontal’ industrial policy is another’s ‘vertical’ interventions. And one economist’s analysis of market failures that may preclude industrialization is another economist’s violation of market principles in support, over decades, of small and medium enterprises (e.g. in South Korea; see Amsden 2013). One view might be that the neoclassical engagement with industrial policy and the longer-standing non-mainstream tradition on this are simply two different ascents up one mountain to a single peak from where there is a clear view of the role of industrial policy in structural change, sustained growth, and development. Another view, however, is that they remain very different paths, haring off in different directions altogether and with different destinations. For, first, despite limited efforts to engage with increasing returns, mainstream views remain ambivalent about their importance. Second, critics of these mainstream views argue that industrial policy requires more than tweaks to correct for market failure or to ‘facilitate’ exploitation of comparative advantage. Rather, they argue that effective industrial policy involves creating and shaping markets (Mazzucato 2011); deviating potentially quite far from comparative advantage (Chang 2003); positively violating free-market principles in pursuit of productive capabilities and competitive advantage; and coordinating a ‘capability triad’ of business model, production capabilities, and skill formation (Best 2018) to enable dynamic increasing returns to flourish and for industrial policy to take place. While there may be some nominal convergence that at least acknowledges the need for industrial policy of some kind, beneath this apparent convergence remain fundamental differences around the role, scope, and means of industrial policy.
3.3 Overview of Structuralist Development Economics and Structural Transformation Structuralism as a cohesive school of thought in development economics can be traced to the ideas developed at the Economic Commission for Latin America and the Caribbean (ECLAC/CEPAL) from the early 1950s on (Blankenburg et al. 2008). 8 See https://www.ft.com/content/b51df920-4db5-11e6-8172-e39ecd3b86fc#axzz4GG5zg8NN and also https://www.ft.com/content/bb242140-ef2b-11e8-89c8-d36339d835c0.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
46 Christopher Cramer and Fiona Tregenna A structuralist approach draws attention to the product specificity of growth, in which manufacturing has a special role to play. Industrialization is considered a necessary condition for developing countries to be able to sustain high rates of growth and thereby to catch up with developed countries (Blankenburg et al. 2008). Hence, structuralist and Kaldorian approaches agree on the role of manufacturing as an engine of economic growth (see Kaldor 1978, 1980; also Thirlwall 1983 and Verdoorn 1993). According to Kaldor’s three laws, a faster rate of output growth in manufacturing is associated (in a fundamental causal sense) with faster aggregate growth; productivity growth in manufacturing is endogenous to the growth of manufacturing output; and aggregate productivity growth is positively related to the growth of manufacturing output and employment. From a structuralist perspective, structural change is fundamental to the processes of growth and development, and to developing countries catching up with advanced economies. Structuralists argue for active policy interventions to promote changes in the sectoral composition of output and employment if developing countries are to sustain higher rates of economic growth. In developing countries in particular, it is thus important to shift towards high-productivity activities. What matters is not only the levels of productivity at a point in time, but also the scope for cumulative productivity increases that build on one another in a virtuous circle. Below we suggest that changes in global capitalist production may mean that the focus needs to shift away from traditionally conceived sectors and towards specific types of activity that have particular effects linked to increasing returns, learning by doing, growth, and structural change, but still based on the idea of product-specific characteristics of these activities. In structuralist approaches, the manufacturing sector has historically been considered to have a special role as the primary engine of growth (see, for example, Alcorta et al. 2013; Hirschman 1958; Prebisch 1950, 1963; Szirmai 2012; Tregenna 2009, 2013; Weiss and Jalilian 2016). The interrelated ‘special properties’ of manufacturing discussed below are associated with learning by doing, increasing returns to scale, inter-sectoral linkages, technological progress, and exports. First, there is thought to be superior scope for learning by doing in manufacturing. The importance of learning by doing lies not only at the level of an individual worker, whose productivity is increased by repeated performance of the same task, but also at the levels of the firm and sector, and the labour force as a whole. The accumulation and dissemination across the workforce of productivity-enhancing skills is central to the analysis. This also involves managerial skills and marketing know-how, negotiations skill and the planning of production and technology, just one of the areas where manufacturing and service-type activities blend into one another. Second is the opportunity (directly linked to learning by doing) to realize increasing returns to scale in manufacturing. Of particular importance here is the dynamic nature of these increasing returns, as distinct from static economies of scale. Manufacturing is regarded as having a relatively high scope for dynamic increasing returns, thus there is a special relationship between output growth and productivity growth in manufacturing (as per Kaldor’s laws). Third, backward and forward linkages have been generally considered to be stronger for manufacturing than for other sectors. These are direct relationships of physical supply and demand between activities and sectors. Strong intersectoral linkages give manufacturing the capacity to pull along the rest of the economy, through positive externalities of manufacturing
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Heterodox Approaches to Industrial Policy and the Implications 47 growth and the deepening of economies of scale in the domestic economy (below in section 3.5 we give the example of linkages between industrial floriculture, and transport services and infrastructure). The fourth key aspect associated with the specificity of manufacturing for growth is its role in advancing research and development (R&D), innovation, and technological advancement, both in manufacturing and more widely. Technological progress diffuses from manufacturing to the rest of the economy—both embodied and disembodied technology, both directly and through spillover effects. The arguments developed in this chapter also suggest that this kind of enhanced technological diffusion has become increasingly significant also within some parts of agriculture, for example, in the Brazilian timber and second-generation biofuel sectors. Similarly, the boundaries between manufacturing and services are especially blurred in information and communications technology activities, which are of particular importance in innovation and technological development. Finally, manufacturing is regarded as important for its contribution to exports and the balance of payments. Related to import elasticities (both income and price), manufacturing is considered critical to alleviating balance-of-payments constraints that can impose a ‘stop-go’ pattern on developing countries’ growth. This is important in supporting sustained high growth rates, particularly in the absence of a strong primary commodity export sector with stable and favourable terms of trade. We return to these dimensions of the specificity of manufacturing, in relation to the role of hubs in industrialization and growth.
3.4 The Challenge of Premature Deindustrialization Industrial policy has returned to front of stage in both advanced and lower-income economies (see above, section 3.2). But another development may undermine the case for industrial policy. This is the phenomenon of premature deindustrialization: the trend in which many developing economies have experienced a shrinkage in the share of manufacturing value added (MVA) in GDP and employment, at lower levels of income per capita and at lower shares of manufacturing than such declines had ever been experienced before. Such a trend, especially if it were an ineluctable function of the structure and technological characteristics of modern global capitalism, threatens any idea that manufacturing may still be an engine of growth and development. For example, the evidence suggests that MVA as a share of GDP has been falling on average across both developed and less developed economies (Figure 3.1). Worse still, ‘deindustrialization shows up most clearly and in its strongest form in employment’ (Rodrik 2016: 1), and most of the manufacturing employment loss in deindustrializing economies takes place in low-skilled jobs. More generally, premature deindustrial ization is often thought to mean that the gains from manufacturing as an engine of growth and catching up are effectively no longer available to developing economies. Rodrik concludes that African economies will have to get what growth they can from the expansion of services, without strong employment-generating effects, and that they will have to wait patiently for skill levels to rise.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
48 Christopher Cramer and Fiona Tregenna Nonetheless, other evidence undermines this pessimism: deindustrialization has not proceeded at the same rate in all regions of the world. The pace and depth of deindus trialization in OECD economies has varied depending on policy. Thus, for example, UK deindustrialization has been notoriously pronounced (deeper and faster), compared to that in Germany or other OECD economies. By the same token, there is surely scope for coherent policy to make a difference in lower-income countries, to slow down, put off, or reverse premature deindustrialization. Thus in Ethiopia, there has in fact been a rise in the share of industry in GDP and in employment, albeit from an exceptionally low base (Oqubay 2015, 2019). Manufacturing employment has also grown: from 1.1 million in 1999 to almost 1.9 million by 2013 (Martins 2018). And the wider evidence confirms that premature deindustrialization is not uniform or inevitable. Countries—especially in East Asia—that adopted deliberate strategies and have succeeded in securing rapid growth through manufacturing show how important policy is to whether or not a country succumbs to premature deindustrialization (Haraguchi et al. 2017). Indeed, aggregate (as opposed to average) MVA in GDP has fallen in developed economies but has risen in developing countries as an aggregate group (Figure 3.2). Lavopa and Szirmai (2015) also show that, when measured in constant prices, globally the share of manufacturing has remained roughly constant over time. Clearly, there is huge variation across developing countries. That would seem to suggest scope for policy to affect the pace of industrialization/deindustrialization in lowerincome countries. Further, there is other research finding that manufacturing is still an engine of growth and development (Haraguchi et al. 2017). A number of recent studies support the argument that Kaldor-type effects continue to be relevant: that is, that manufacturing output growth is linked to overall economic growth, and more closely than are other sectors; and that manufacturing output growth generates productivity growth within (and beyond) manufacturing. Szirmai and Verspagen (2015) argue that manufacturing is indeed still a driver of
20 18 16 14
Average of MVA/GDP (developed)
2016
2014
2012
2010
2008
2006
2004
2002
2000
1998
1996
1994
1992
1988
1990
1986
1984
1982
1980
1978
1976
1974
10
1972
12 1970
Share of MVA GDP (%)
22
Average of MVA/GDP (developing)
Figure 3.1 Average manufacturing value added as a share of GDP, developed and developing countries (1970–2017) Source: UN Statistics Division (2019), Income Classification: World Bank GNI per capita, Atlas methodology (2019).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Heterodox Approaches to Industrial Policy and the Implications 49
20 18 16 14
Aggregate MVA/GDP (developed)
2018
2016
2014
2012
2010
2008
2006
2004
2002
2000
1998
1996
1994
1992
1990
1988
1986
1984
1982
1980
1978
1976
1974
10
1972
12 1970
Share of MVA GDP (%)
22
Aggregate MVA/GDP (developing)
Figure 3.2 Aggregate manufacturing value added as a share of GDP, developed and developing countries (1970–2017) Source: UN Statistics Division (2019), Income Classification: World Bank GNI per capita, Atlas methodology (2019).
growth though this effect has weakened over time and probably depends on decent levels of education. Marconi et al. (2016) similarly confirm Kaldor’s growth laws and especially in low- and middle-income countries. Su and Yao (2017) find that manufacturing drives growth in services rather than the other way around. However, we argue that there is another, perhaps more important, reason to be sceptical about the sweeping claims of premature deindustrialization and the pessimism that goes with them. This we take up in the next section: what constitutes the ‘industrial’ is in reality changing in ways that international statistical classification has not accommodated.
3.5 From Assembly Plants to Plant Assembly: Classification and Structural Change Sectoral classifications shape what we understand about patterns of industrial output, productivity, and employment. This is especially true for structuralist and other economists who have argued that there are special characteristics of manufacturing that give it a particularly strong role in accelerating structural change. All classification systems impose artificial breaks on continuous realities, yet they are necessary to order such realities in the interests of analysis, to ‘tame the wild profusion of existing things’ (Foucault 1973). But some classification systems are more helpful than others and the simple sectoral division of economic activities has become increasingly ill-suited to its purpose as the characteristics of production have changed. The distinctions between sectors—the categorial borders—have become more porous, even as our classifications (for example, through ISIC or SITC) have insisted on manning the same sentry posts dividing and sub-dividing within sectoral categories. Marshall was very alert to this danger (1920: 8). And more recently, others have suggested, it ‘would be useful to refine the standard three-sector focus’ (Herrendorf
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
50 Christopher Cramer and Fiona Tregenna et al. 2014: 929) of the literature on growth, production, and structural change. One important implication for development policy is that the ‘classic focus on factories can be misleading’ (Baldwin and Lopez-Gonzalez 2015: 26). There is growing heterogeneity in the characteristics of activities within sectors. Broadly, sectors may continue to be characterized by important common denominators that are relevant for growth. However, sectors encompass very different types of activities, with differences in scope for cumulative productivity increases and for pulling along the rest of the economy. For instance, certain types of services are increasingly tradable, some non-manufacturing sub-sectors can have stronger intersectoral linkages than some manufacturing sub-sectors, learning by doing can be important in a range of activities, and so on. While analysis at the sectoral level remains meaningful, the degree of heterogeneity within sectors, the fuzziness in categorical boundaries between sectors, and the close integration between sectors, all mean that not only sector specificity but product and activity specificity matter. In cases of deindustrialization or other changes in the sectoral structure of an economy, it is essential to look deeper than changes in overall sectoral categories, such as manufacturing and services, to understand which activities within sectors are growing or in decline, along with the characteristics of these activities (see Tregenna 2015). Industrial policies (including those supporting hubs) need to support whatever activities can generate dynamic returns. Certain aspects of this problem have attracted growing attention in recent years. In particular, the fuzzy distinctions and imbricated relationships between services and manufacturing have generated a research literature. Neither data nor policies have caught up with this development. For example: U.S. labor data—and many policies that are informed by these data—do not fully recognize the ties between these services and manufacturing . . . The conceptual foundation of [the North American Industrial Classification System (NAICS)] typology rests on the distinction that extractive industries (such as raw material recovery), manufacturing industries (such as fabricated metal or automotive manufacturing), and service industries (such as b usiness and professional services) rely on essentially different production processes and establishments in each of these three categories have similar processes. (Whitefoot et al. 2015: 3)
NAICS data ignores product value chains and as a result excludes from its definition of manufacturing those services—even in a different building adjacent to a manufacturing building on the same site—necessary for the production and distribution of final physical goods. Servicification, at one end of the sectoral classification system, is increasingly well understood (Timmer et al. 2014; Koopman et al. 2010; Baldwin and Lopez-Gonzalez 2015). Less attention has been paid to the ‘industrialization of freshness’ at the other end. Although the phenomenon is not remotely new, it has gathered pace in recent years, increasingly confounding a categorical distinction between agriculture and manufacturing/industry (or indeed, between agriculture and services). By a reasonable definition of what constitutes ‘industrial’ production, agriculture globally and increasingly in developing countries is industrial in nature. We take as useful Young’s (1928) definition of the industrial: what matters is less whether or not production takes place in factories than what forms of industrial
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Heterodox Approaches to Industrial Policy and the Implications 51 organization are at play and how ‘roundabout’ is production, the extent to which an ‘intricate nexus has inserted itself between the producer of raw materials and the consumer of the final product’ (Young 1928: 528). As Katz (see Chapter 41) shows, natural resource-based commodities and foodstuffs involve highly location-specific production processes and requirements, one factor propelling an increasingly science-based production in these activities. Beyond location and product specificity, another factor driving the industrialization of freshness is the growth and stringency of global phyto-sanitary standards. This was one of the key forces driving sophisticated upgrading in Uruguayan beef exports. Uruguay now has real-time traceability capability for 100 per cent of its immense cattle population, with micro-chips inserted at birth into calves’ ears (Gereffi 2018: 356–7). While the evolution of processed agricultural products has proceeded apace, including the development of ultra-processed foods (Monteiro et al. 2018), what is less obvious is the increasing sophistication of production of what at first sight are unprocessed products (potted pelargoniums, fresh oranges, blueberries, avocados, tomatoes, etc.). It has become increasingly difficult to distinguish—in terms of intricacy, roundaboutness, or complexity, or the ‘lengthy linkage of production phases’ (Gala et al. 2017: 2)—between ‘processed’ agricultural commodities and apparently unprocessed or ‘fresh’ commodities. The boundaries between Wood and Mayer’s (1998) ‘processed primary’ and ‘unprocessed’ primary products or between Mendez-Ramos and Paustian’s (2017) primary and manufactured agribusiness, for example, have been eroded (where does the processing of a Uruguayan head of cattle begin?). And it becomes less easy to have confidence in Wood’s (2017: 9) suggestion that ‘processed primary exports are . . . clearly more skill-intensive than unprocessed primary exports’. It is this increasingly complex, roundabout production that characterizes the ‘industrialization of freshness’ (Cramer et al. 2018; Cramer and Sender 2019). Getting ‘ripe and ready-to-eat’ avocados, punnets of fresh blueberries or tomatoes, or a young plant pot onto a shelf in a nursery or greengrocer’s in, say, the United Kingdom, typically involves a highly intricate nexus, a chain of linkages. Blueberries, for example, are not pulped or juiced, not skinned, pasteurized, or de-husked. But they are knowledge intense. They are the fruit of highly technical genetic plant stock R&D and of tacit knowledge about specific localized agro-climatic conditions. They rest on a bed of sophisticated capabilities in running highly capitalized firms, sourcing inputs, negotiating government incentive schemes and constraints, and supervising and training a labour force. They require investment in tracing and monitoring IT software and hardware, partly in order to ensure that demanding phytosanitary standards are met. Skills are needed to make subtle adjustments to computer management of drip fertigation, lighting, and humidity. Profitability turns on marketing strategies that connect blueberry consumption to ‘super-food’ health benefits. Many very similar features, as well as complicated packaging and logistics arrangements linked to the just-in-time demands of global trade in fresh agricultural and floricultural goods, are characteristic too of production and trade in oranges, avocados, cut roses, etc. Some products—for example, poinsettias produced in Ethiopia—undergo a regime of sensory deprivation, pest control, reverse osmosis of water, temperature manipulation, and misting designed to maximize yields. A carton of orange juice may at first blush seem more ‘complex’ and knowledge intense than a fresh orange. Yet a South African orange producer, for example, will explain that the
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
52 Christopher Cramer and Fiona Tregenna truth is the reverse. Carton technology is now relatively standardized (and easy to import) and generally rather poor-quality oranges are pulped into the juice cartons. What is much more difficult is to generate a consistent supply of very high-quality oranges, unblemished, easy to peel, without too much thick pith, and perfectly juicy. And it is the production of a consistent supply of high-quality inputs even to juice and pulp processing (mango or passion fruit, for example) that makes it complex. These examples challenge the traditional view that there is little scope for innovation and technological progress in agricultural and extractive activities. They also call into question the old argument that there are far fewer linkages in agriculture than in manufacturing. The linkages involved in the cut rose global trade or the fresh herb business, for example, are legion. They include fertigation equipment, software development, worker uniform manufacture, reverse osmosis machinery, cardboard packaging, cold-storage rooms, suitably enabled trucks, and, of critical importance, logistics facilities and transport infrastructure. Having a state-owned airline in Ethiopia, for example, helped to make viable the promotion of the floriculture sector in the 2000s (a form of backward linkage that was far from automatic, but that government policy realized). And the rapid initial growth and consolidation of the sector then fed back into further (forward linkage) advances at Ethiopian Airlines, with the development of cold-storage facilities and a huge expansion of the freight terminal. We argue that expanding the terrain over which industrial policy may apply—in support of accelerated gains in productivity, employment, foreign exchange earnings, complexity, linkages, and learning by doing—may help policy officials identify and respond faster to the range of constraints that undermine the further expansion of the kinds of activity we have characterized as subject to the industrialization of freshness. For the proliferation of linkages is a potential feature of this new industrialization: it is not automatic. Nevertheless, seeing the significance of these shifting boundaries of classification is only part of the policy issue. For most, and perhaps all, of the constraints that typically hold back growth in such sectors in countries like Ethiopia or South Africa are not purely technical or exclusively economic; they are fundamentally social and political. This is the case, for example, with under-investment in water and irrigation infrastructure, or in transport infrastructure or energy (Cramer and Sender 2015; Cramer et al. 2018).
3.6 A Structuralist Approach to Hubs The issues set out above all have a spatial dimension. Various theories of industrial location have been put forward in the literature, seeking to explain firms’ decisions about where to locate, from Ricardian trade theory (Ricardo 1821), through Marshall’s (1920) work on agglomeration, to Krugman’s (1979, 1993) work on trade theory and economic geography, and beyond (see Chapter 2). Furthermore, there are well-established bodies of literature on the benefits of clustering amongst manufacturing firms (see, for example, Gordon and McCann 2000; Huang et al. 2008; Ketels 2013; Porter 1998; and Schmitz and Nadvi 1999), as well as on the potential advantages of industrial development zones, special economic zones, and other forms of economic zones (see, for example, Aggarwal 2010; Belloc and di Maio 2011; Stein 2011; Warr and Menon 2016).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Heterodox Approaches to Industrial Policy and the Implications 53 Of particular interest here is the role of hubs in fostering the clustering and agglomeration of firms, and the ways in which this can strengthen the role of manufacturing as an engine of growth. As discussed earlier, what is characteristic about the Kaldorian and structuralist conceptions of increasing returns to scale in manufacturing is the dynamic nature of these economies of scale. From this perspective, it is important to understand the potentially dynamic nature of the benefits of agglomeration through hubs, in supporting cumulative productivity increases and in catalysing growth. Many of the dynamic returns Young (1928) identified come about through ‘intimate connections’ amongst firms that multiply through close physical proximity. Hubs can potentially deepen some of the benefits of industrialization by strengthening the role of manufacturing in pulling along other sectors. We have reviewed the special properties associated with manufacturing from a structuralist perspective. The spatial concentration of firms through hubs can heighten the realization of these special properties, thereby not only promoting industrialization but also enhancing the growth-pulling property of that industrialization. This is of particular importance in developing countries with weak manufacturing sectors, whether manufacturing is as yet undeveloped or has declined through premature deindustrialization. Where manufacturing firms are spatially dispersed, especially in countries with relatively small manufacturing sectors, the potential benefits of manufacturing are less likely to be realized fully. Obstacles faced by manufacturing firms, especially young firms and firms in countries with nascent manufacturing sectors, typically include lack of access to finance, lack of access to appropriate technologies, small product markets, and lack of physical infrastructure and skills (Huang et al. 2008; Long and Zhang 2011; McCormick 1999). Clustering and agglomeration through hubs can potentially mitigate these constraints, thus aiding the growth of manufacturing firms and supporting industrialization. The role of hubs in the ‘thickening’ or ‘densification’ of both physical and non-physical infrastructure is thus especially important in low-income countries in the early stages of industrialization. A structuralist perspective considers manufacturing to have superior scope for cumulative productivity increases. We have argued above that some other activities outside the traditional boundaries of the manufacturing sector share similar properties. The evidence suggests that hubs can not only achieve levels of productivity above national averages, but can contribute positively to economy-wide productivity by raising productivity outside of the hubs through various forms of linkages.9 For each of the ‘special properties’ of manufacturing discussed earlier, we now discuss how hubs can provide a fertile ground for the actualization of these properties, thus strengthening the role of manufacturing as an engine of growth and contributing to industrialization. These are all merely potential gains, which may or may not accrue in practice. First, the proximity in hubs of manufacturing firms (and firms in related or similar activities) can enhance learning by doing and learning more broadly, at both the firm and sector levels, thus strengthening productive capabilities. Lack of knowledge can be a 9 Kingombe and te Velde (2016) find that EPZs raised labour productivity in Kenyan manufacturing by 20 per cent between 1990 and 2007. Burki and Khan (2013) show that localization economies had a positive impact on the productivity of manufacturing firms in Pakistan, especially those in the textiles and leather industry, where firms were learning from each other.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
54 Christopher Cramer and Fiona Tregenna constraint on emerging manufacturers. This relates not only to technological competencies, but also a range of managerial competencies and productive capabilities. Amongst firms in hubs, workers can act as conduits for the transfer of knowledge to producers as they communicate with workers in other firms or switch from one firm to another (Widodo et al. 2015). Agglomeration of firms in hubs can enable employee mobility between firms as well as collaboration in solving common problems, encouraging networks of information sharing and knowledge transfer that can potentially have positive spin-offs beyond hubs as well. Similarly, when managers or other staff in hub firms subsequently move into positions outside the hubs, this can facilitate skills transfer. Inkpen and Pien (2006) find evidence of knowledge transfer in the China–Singapore Suzhou Industrial Park (SIP). In addition, foreign firms attracted to hubs may promote human capital formation by providing on-the-job training and improving supervisory and managerial skills through learning by doing (Tang 2015). Some firms in EPZs and SEZs such as Lat Krabang (Thailand), Kaohsiung (Taiwan), Masan (Republic of Korea), and Shenzhen (China) provided on-thejob training to operators and overseas training at the parent company for employees at managerial level (Kusago and Tzannatos 1998). Skills linkages amongst firms within hubs can foster learning and skills development amongst hub firms, while skills linkages between hub firms and the rest of the domestic economy can support skills development more broadly. These skills linkages can thus enable improvements in productivity, competitiveness, and industrial upgrading. However, the realization of these effects, both within hubs and between hubs outside firms, is contingent on the nature of the hubs as well as on how hub and non-hub firms are linked. Hubs based on low-skill, low-productivity, low value-added activities are unlikely to require or develop many skills, or productive capabilities more broadly. Similarly, where hubs are delinked from the rest of an economy, there is unlikely to be much transfer of skills, knowledge, or capabilities in ways that can facilitate wider industrial upgrading. Second, hubs may be able to contribute to the realization of dynamic, increasing returns to scale (see above, section 3.3). The agglomeration of firms, especially of those in the same sector, can potentially mitigate some of the problems faced by small firms operating in isolation. This is especially the case in low-income developing countries that have a small industrial base. Clustering and agglomeration through hubs may assist firms in achieving economies of scale beyond their own size. For example, hubs can assist firms (especially small ones) in mitigating the constraints associated with size and help them realize something akin to broader economies of scale. This can be through mechanisms such as bulk purchasing of inputs, knowledge sharing, bulk tendering, skilled labour pooling, common market outlets, risk reduction, and various forms of joint action. Nadvi (2016) argues that clustering can help small firms overcome their size limitations by gaining improved market access through cooperation, as well as specialized markets for labour and other inputs, thus lowering costs for individual firms. The evidence suggests that firms in clusters can engage in cooperation and collective action, leading to competitive advantages through R&D and the upgrading of technological capabilities and skills, and the establishment of marketing consortia (Nadvi 2016). Through clustering and agglomeration in hubs, firms may also realize greater differentiation and specialization, feeding into dynamic economies of scale. An integrated production process can be split into smaller parts, allowing firms to focus on a distinct stage of production
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Heterodox Approaches to Industrial Policy and the Implications 55 that does not require large capital fixed investment (Long and Zhang 2011; Schmitz and Nadvi 1999). There is evidence of specialization and differentiation of firms in some industrializing clusters (McCormick 1999). In the case of the Wenzhou region in China, an area initially lacking the basis for economic development, a high degree of specialization emerged in the industrial cluster of footwear firms, transforming it into an important industrial region (Huang et al. 2008). Mukkala (2004) finds that agglomeration economies, especially localization and specialization, positively influenced manufacturing productivity in Finland. Aggarwal (2011) argues that by addressing constraints that had prevented firms from reaping the benefits of economies of scale, such as rigidities in capital and labour markets and unavailability of land, SEZs have assisted in the emergence and growth of some large firms such as the Jamnagar SEZ refinery (Aggarwal 2011). Agarwalla (2011) finds that agglomeration economies, particularly urbanization economies, had a non-linear relationship with productivity because at lower levels of urbanization, support services that help reduce costs of production were underdeveloped. Third, hubs can potentially contribute to the strengthening of intersectoral linkages, both amongst manufacturing firms within hubs and possibly with the rest of the domestic economy. Intersectoral linkages are amongst the key ways in which manufacturing can pull along the rest of the economy, which would be a positive externality of manufacturing growth. Where there are strong backward and forward linkages between manufacturing firms in hubs and non-manufacturing firms either in the hubs or outside, this can stimulate demand and build productive capabilities. This can also contribute to capturing more of a global value chain domestically. There is mixed empirical evidence around the strength of linkages between firms within and outside hubs. For example, in the case of Madagascar, Cling and Letilly (2001) find there to have been strong linkages between EPZ firms and firms outside the zones, with EPZ firms sourcing supplies from external firms. But even after thirty years of EPZs in Dominica, less than 0.0001 per cent of material inputs to EPZ production were sourced from the domestic economy (Engman et al. 2007). There is considerable heterogeneity in the strength of backward linkages, including amongst different sectors in the same zone.10 In addition to the overall degree of economic integration between the hub and the rest of the domestic economy, other factors that appear to positively influence the strength of backward linkages include both characteristics of hub firms—local ownership, domestic sales, number of years in operation (Jenkins and Arce 2015)—and country-level characteristics such as population size (Schrank 2001). Sector and the nature of products also matter, with low-volume, high-value inputs in sectors such as electronics being more likely to be imported from other countries than sourced through domestic backward linkages, relative to low-value, high-volume and especially perishable inputs, such as in food processing (Jenkins 2005; Jenkins and Arce 2015). More widely, a hub operating as an enclave, in which
10 Spinanger (1984) finds that in the case of the Penang EPZ, domestic purchases of capital goods from the rest of Malaysia grew from 12 per cent in 1973 to 29 per cent in 1979 while domestic purchases of raw materials and components grew from less than 1 per cent to 3 per cent over the same period. According to Warr (1989), domestic raw material purchases by EPZ firms increased from US$0.07 million in 1972 to US$8.5 million in 1982 in Masan EPZ in South Korea and local purchases of capital goods in Jakarta EPZ in Indonesia grew from a zero base in 1977 to US$300,000 in 1981.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
56 Christopher Cramer and Fiona Tregenna inputs are largely imported into the hub, would have little stimulating effect through the intersectoral linkages channel. Fourth, hubs can deepen the growth-pulling effects of manufacturing through the technological progress channel. If appropriately designed and managed, hubs may foster R&D, innovation, and technological advancement, both in manufacturing and more widely. The flow of information, knowledge, and know-how within hubs can lower the costs of production, enable upgrading, and strengthen productive capabilities amongst firms in hubs. Moreover, through technological and other linkages with firms outside the hubs, technological spillovers and technology transfer can contribute to technological progress in the rest of the economy. There is extensive evidence for the effects of clustering in promoting innovation—see, for instance, Bathelt et al. (2004) and Malmberg and Maskell (2010). Nadvi (2016) shows that the agglomeration of suppliers and producers operating in the same sector can enhance innovation and knowledge spillovers. Zeng (2011) provides evidence that Chinese SEZs contributed to technological advancement and upgrading in China. Hsu et al. (2013) argue that, by making it easier for firms to access capital resources, infrastructure (both physical and technological), and high-quality human capital, clusters tend to promote horizontal competition and cooperation amongst firms, leading to innovations and improvements in technology. The tendency of clusters to foster technological upgrading through R&D is also supported by Fu and Gao (2007) who found that in 2002 HIDZs accounted for 24.4 per cent and 35.1 per cent of China’s total expenditure on R&D in 2002 and 2006, respectively. Finally, exports are one of the channels through which manufacturing can generally act as an engine of growth, and hubs can strengthen this contribution. Hubs, and in particular zones, typically have an explicit export-promotion orientation. EPZs in particular are generally delimited areas separated from the rest of the domestic economy— both physically and by regulation—whose products are solely permitted to be exported. More broadly, while industrial hubs need not be limited to exports, a focus on exports is generally seen as an important objective, and this is reflected in the design and operations of these zones. There is extensive empirical evidence for the contribution of zones to increasing exports, both directly and by raising exports by firms outside hubs (see Engman et al. 2007; Leinbach 1982; Warr 1989; Fu and Gao 2007; Aggarwal 2011; Angko 2014). Johansson and Nilsson (1997) show that Malaysian EPZs raised exports not only directly in the zones, but also in the rest of the Malaysian economy. In Mauritius, EPZs’ share in total exports rose dramatically from 3 per cent in 1971 to 53 per cent in 1986, with impressive growth in EPZ exports also observed in Costa Rica and Bangladesh (Engman et al. 2007). In China, Economic and Technology Development Zones (ETDZs) and the Hightech Industrial Development Zones (HIDZs) each accounted for a third of high-tech product exports (Fu and Gao 2007). Furthermore, zones can contribute to export diversification (particularly from traditional to non-traditional exports) and upgrading within global value chains. In terms of the potential of hubs to realize dynamic gains and catalyse structural transformation, perhaps more important than the volume of exports is their composition. If used strategically, hubs can potentially provide a base for building the productive capabilities needed to move into more sophisticated, higher value-added products.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Heterodox Approaches to Industrial Policy and the Implications 57
3.7 Hubs and Structural Transformation: Difficulties and Policies Through the channels discussed in the previous section, hubs can potentially deepen the growth-pulling properties of manufacturing. Taking into account not only sector specificity but also product and activity specificity, as discussed earlier, it is important that hubs retain, but also go beyond, a focus on manufacturing. For hubs to be effective tools of structural transformation, not only traditional manufacturing activities but also other activities with scope for cumulative productivity increases and growth pulling need to be targeted for hubs (and in industrial policy more broadly). For instance, in the ‘industrialization of freshness’, speed of access to market is of the essence, and hubs can shorten the time from production to export. Furthermore, hubs can play a role in strengthening links between manufacturing, quasi-manufacturing, and non-manufacturing. The potential advantages of hubs for industrialization and growth are by no means automatic. At the micro level, the potential for cooperation between firms in hubs and for the realization of economies of scale, technological spillovers, and other potential benefits does not necessarily occur in practice.11 At the macro level, even if hubs do contribute in a static sense to employment creation and growth in output—which are of course important gains in their own right for developing countries—they do not necessarily contribute meaningfully to structural transformation (see Chapter 18 on SEZs and structural transformation; Chapter 37 on Vietnam; Chapter 42 on Latin America). A risk for broader growth and development is that hubs function as enclave economies, effectively delinked from the rest of an economy. For instance, maquiladoras operate essentially as foreign enclaves within a domestic economy. Components that have been manufactured (and even partially assembled) elsewhere are imported into the zone, assembled (in some cases just the final stages of assembly) and then exported (Palma 2005). The output and jobs of these zones are static gains that count positively in Mexico’s economic statistics. Yet more fundamentally, there is limited value addition, skills transfer, building of productive capabilities, contribution to technological or other upgrading, innovation, or pulling along of firms outside the zones through linkages. In short, there is little real contribution to structural transformation, diversification, industrialization, or sustainable growth. Related to this, hubs whose appeal to investors centres primarily around low wages, weak labour standards, poor working conditions, and restricted labour and union rights are more likely to attract low skills and low value-added production, which do not contribute to upgrading and structural change. While maquiladoras can be considered an extreme case, they demonstrate that the potential benefits of hubs certainly do not always materialize; these benefits are contingent and conditional. This highlights the importance of planning hubs in such a way that they 11 For example, McCormick’s (1999) study of six enterprise clusters in African countries found no or weak intermediate input links, technological spillovers, and bilateral or multilateral linkages in the Eastlands garment cluster and the Kamukunji metalwork cluster in Kenya; overall, she finds weak labour market pooling, weak intermediate input effects, and weak or limited technological spillovers.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
58 Christopher Cramer and Fiona Tregenna are linked with rest of the domestic economy through multiple channels—supply linkages, technological linkages, skills linkages, etc. Industrial policy is of crucial importance in this regard, not only for the hubs themselves, but also for the nature of their articulation with the rest of an economy. One of the policy implications associated with this is that positive support measures are more likely to support the integration of the hubs with the rest of the economy and to maximize the dynamic gains from hubs than are negative support measures in the form of exemptions from what is generally required of non-hub firms. These exemptions could be financial (such as tax breaks) or non-financial (such as exemptions from the labour legislation that applies outside hubs). These sorts of negative support measures may be simpler to implement than positive ones. However, positive support measures can be more effectively deployed to achieve particular policy objectives, when designed with appropriate conditionalities and carefully implemented. These policy objectives might include the building of productive capabilities, the diversification and upgrading of production, and the strengthening of various types of linkages (production, skills, knowledge, etc.) between hubs and non-hub firms. Realizing the potential benefits of hubs in practice is conditional on various factors. These include: the character of these hubs (how they are conceptualized, designed, and run, including what support measures are provided to firms locating in the zones); the extent to which they give rise to net new production rather than just subsidizing production that would otherwise have occurred anyway; broader industrial policies; how hubs and the firms within them are articulated with the rest of the domestic economy; the nature of a country’s trade regime and what sort of regional trade bloc it is located within; and so on. Some of these aspects are discussed in greater depth in other contributions in this volume. Of particular interest here is how the character of hubs as either comparative advantagedefying or comparative advantage-conforming affects the extent to which they can contribute to industrialization and growth. Indeed, this is a broader issue in debates on the role and practice of industrial policy (see, for example, Lin and Chang 2009). In their design and implementation, hubs could merely do ‘more of the same’, and yield static gains: attracting investment based on a country’s existing comparative advantage and broad sectoral structure. Alternatively, hubs can be explicitly oriented towards defying a country’s existing comparative advantage, targeting more sophisticated or higher value-added activities in which a country aspires to become competitive. These are what we consider the potential dynamic gains from hubs, where hubs have the potential of acting as levers or catalysts for broader structural transformation. This kind of criterion for allocating resources to support hub development may be especially important where a government is keen to develop agro-parks. In such cases (Ethiopia, for example), it is important to find ways to maximize the scope for linkages, productivity gains, export revenues, and dynamic returns associated with the ‘industrialization of freshness’ rather than merely to try to pool activities supporting traditional lower-value agricultural output. In this sort of approach, hubs are a tool of a broader industrial strategy that promotes upgrading through both financial and non-financial support measures. This is consistent with structuralist thinking, which stresses the dynamic nature of comparative advantage, the need for continual upgrading of the production structure for developing countries to catch up, and the need for active policy interventions to make these happen. While there is
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Heterodox Approaches to Industrial Policy and the Implications 59 a strong case for industrial policy in general to be comparative advantage defying, the case is arguably even stronger with respect to policy on hubs. Leapfrogging in comparative advantage can be achieved more easily in hubs than in the wider economy. This is due to the economies of scale that can be realized through hubs, and the industrial policy measures that can be designed with the specific purpose of supporting upgrading (product, process, and function) and rapid progression in productive capabilities. This underscores the importance of a country’s policy on hubs being integrated with broader industrial and other policies. If a country has a hubs promotion policy that is isolated from other economic and developmental policies, it is likely that the hubs themselves will also develop as enclaves and will fail to pull along wider industrialization and growth. Policies need not only promote hubs, but to actively promote their linkages with the rest of an economy, including linkages through supply chains, knowledge and skills transfers, and so on. The success of hubs must be measured not only in the narrow terms of their own performance in isolation, but more importantly in their broader growth-pulling role. This, in turn, requires support from a range of policies in addition to bold industrial policies, including technology and innovation policies, labour market and skills development policies, trade policies, spatial and regional policies, and supportive and appropriate macroeconomic policies.
References Agarwalla, Astha (2011) ‘Agglomeration Economies and Productivity Growth in India’. Working Paper No. 2011-01-08. Indian Institute of Management, Ahmedabad. Aggarwal, Aradhna (2010) ‘Economic Impacts of SEZs: Theoretical Approaches and Analysis of Newly Notified SEZs in India’. MPRA Paper No. 20902, Department of Business Economics, University of Delhi, India. Aggarwal, Aradhna (2011) ‘Promoting Agglomeration Economies and Industrial Clustering through SEZs: Evidence from India’, Journal of International Commerce, Economics and Policy 2(2): 201–27. Alcorta, Ludovico, Nobuya Haraguchi, and Gorazd Rezonja (2013) ‘Industrial Structural Change, Growth Patterns, and Industrial Policy’, in Joseph E. Stiglitz, Justin Yifu Lin, and Ebrahim Patel (eds) The Industrial Policy Revolution II. International Economic Association Series. London: Palgrave Macmillan, pp. 457–91. Alonso-Villar, Olga, José-Maria Chamorro-Rivas, and Xulia González-Cerdeira (2007) ‘Agglomeration Economies in Manufacturing Industries: The Case of Spain’, Applied Economics 36(18): 2103–16. Amsden, Alice (1989) Asia’s Next Giant: South Korea and Late Industrialization. Oxford: Oxford University Press. Amsden, Alice (2001) The Rise of ‘the Rest’: Challenges to the West from Late-industrializing Economies. Oxford: Oxford University Press. Amsden, Alice (2013) ‘Securing the Home Market: A New Approach to Korean Development’. Research Papers. UNRISD, Geneva. Angko, William (2014) ‘Analysis of the Performance of Export Processing Zones in Ghana’, Journal of Business Administration and Education 5(1). António, Nelson Santos and Shaozhuang Ma (2015) ‘China’s Special Economic Zones in Africa: Context, Motivations and Progress’, Euro Asia Journal of Management 25(1/2): 79–103. Backhouse, Roger (2010) The Puzzle of Modern Economics. Cambridge: Cambridge University Press. Baldwin, Richard and Javier Lopez-Gonzalez (2015) ‘Supply-Chain Trade: A Portrait of Global Patterns and Several Testable Hypotheses’, The World Economy 38(11): 1682–721.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
60 Christopher Cramer and Fiona Tregenna Bathelt, Harald, Anders Malmberg, and Peter Maskell (2004) ‘Clusters and Knowledge: Local Buzz, Global Pipelines and the Process of Knowledge Creation’, Progress in Human Geography 28(1): 31–56. Belloc, Marianna and Michele Di Maio (2011) ‘Survey of the Literature on Successful Strategies and Practices for Export Promotion by Developing Countries’. International Growth Centre Working Paper No. 11/0248. London School of Economics and Political Science, London. Best, Michael (2018) How Growth Really Happens: The Making of Economic Miracles through Production, Governance, and Skills. Princeton, NJ: Princeton University Press. Blankenburg, Stephanie, Gabriel Palma, and Fiona Tregenna (2008) ‘Structuralism’, in Lawrence Blume and Steven Durlauf (eds) The New Palgrave: A Dictionary of Economics, 2nd edition Basingstoke: Palgrave Macmillan, pp. 69–74. Burki, Abid Aman and Mushtaq A. Khan (2013) ‘Agglomeration Economies and their Effects on Technical Inefficiency of Manufacturing Firms: Evidence from Pakistan’. IGC Working Paper. International Growth Centre, London. Chang, Ha-Joon (2003) Kicking away the Ladder: Development Strategy in Historical Perspective. London: Anthem. Cherif, Reda and Fuad Hasanov (2019) ‘The Return of the Policy That Shall Not Be Named: Principles of Industrial Policy’. IMF. Available at https://www.imf.org/en/Publications/WP/Issues/2019/03/26/ The-Return-of-the-Policy-That-Shall-Not-Be-Named-Principles-of-Industrial-Policy-46710. Cheru, Fantu, Christopher Cramer, and Arkebe Oqubay (eds) (2019) The Oxford Handbook of the Ethiopian Economy. Oxford: Oxford University Press. Cling, Jean-Pierre and Gaëlle Letilly (2001) ‘Export Processing Zones: A Threatened Instrument for Global Economy Insertion?’. DIAL Working Paper DT/2001/17. Cramer, Christopher and John Sender (2015) ‘Agro-Processing, Wage Employment and Export Revenue: Opportunities for Strategic Intervention’. TIPS Working Paper for the Department of Trade and Industry, London. Cramer, Christopher and John Sender (2019) ‘Oranges Are Not Only Fruit: The Industrialization of Freshness and the Quality of Growth’, in Ravi Kanbur, Akbar Noman, and Joseph Stiglitz (eds) The Quality of Growth in Africa. New York: Columbia University Press, pp. 209–33. Cramer, Christopher, Jonathan Di John, and John Sender (2018) Poinsettia Assembly and Selling Emotion: High Value Agricultural Exports in Ethiopia. Paris: Agence Française du Développement. Doner, Richard Fredrick, Bryan K. Ritchie, and Dan Slater (2005) ‘Systemic Vulnerability and the Origins of Developmental States: Northeast and Southeast Asia in Comparative Perspective’, International Organization 59(2): 327–61. Engman, Michael, Osamu Onodera, and Enrico Pinali (2007) ‘Export Processing Zones: Past and Future Role in Trade and Development’. OECD Trade Policy Papers No. 53. OECD Publishing, Paris. Foucault, Michel (1973) The Order of Things: An Archaeology of the Human Sciences. New York: Vintage. Fu, Xiaolan and Yuning Gao (2007) Export Processing Zones in China: A Survey. Geneva: International Labour Organization. Gala, Paulo, Jhean Camargo, Guilherme Magacho, and Igor Rocha (2017) ‘Sophisticated Jobs Matter for Economic Complexity: An Empirical Analysis Based on Input–Output Matrices and Employment Data’, Structural Change and Economic Dynamics 45(C): 1–8. Gereffi, Gary (2018) Global Value Chains and Development: Redefining the Contours of 21st-century Capitalism. Cambridge: Cambridge University Press. Gordon, Ian Richard and McCann, Philip (2000) ‘Industrial Clusters: Complexes, Agglomeration and/or Social Networks?’, Urban Studies 37(3): 513–32. Haraguchi, Nobuya, Charles Fang, Chin Cheng, and Eveline Smeets (2017) ‘The Importance of Manufacturing in Economic Development: Has This Changed?’, World Development 93: 293–315. Hausmann, Ricardo, Lant Pritchett, and Dani Rodrik (2005) ‘Growth Accelerations’, Journal of Economic Growth 10(4): 303–29. Henderson, William Otto (1982) ‘Friedrich List and the French Protectionists’, Zeitschrift Für Die Gesamte Staatswissenschaft/Journal of Institutional and Theoretical Economics 138(2): 262–75.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Heterodox Approaches to Industrial Policy and the Implications 61 Herrendorf, Berthold, Richard Rogerson, and Ákos Valentinyi (2014) ‘Growth and Structural Transformation’, in Philippe Aghion and Steven N. Durlauf (eds) Handbook of Economic Growth, Vol. 2: 855–941. Elsevier, https://doi.org/10.1016/B978-0-444-53540-5.00006-9. Hirschman, Albert (1958) The Strategy of Economic Development. New Haven, CT: Yale University Press. Hsu, Maw-Shin, Yung Lung Lai, and Feng Jyh Lin (2013) ‘Effects of Industry Clusters on Company Competitiveness: Special Economic Zones in Taiwan’, Review of Pacific Basin Financial Markets and Policies 16(3): 1–28. Huang, Zuhui, Xiaobo Zhang, and Yunwei Zhu (2008) ‘The Role of Clustering in Rural Industrialization: A Case Study of the Footwear Industry in Wenzhou’, China Economic Review 19(3): 409–20. Inkpen, Andrew C. and Wang Pien (2006) ‘An Examination of Collaboration and Knowledge Transfer: China–Singapore Suzhou Industrial Park’, Journal of Management Studies 43(4): 779–811. Jenkins, Mauricio (2005) ‘Economic and Social Effects of Export Processing Zones in Costa Rica’. ILO Working Paper No. 97. ILO, Geneva. Jenkins, Mauricio and Ronald Arce (2015) ‘Do Backward Linkages in Export Processing Zones Increase Dynamically? Firm-level Evidence from Costa Rica’, Journal of Business Research 69(2): 400–9. Johansson, Helena and Lars Nilsson (1997) ‘Export Processing Zones as Catalysts’, World Development 25(12): 2115–28. Kaldor, Nicholas (1978) Further Essays on Economic Theory. London: Duckworth. Kaldor, Nicholas (1980) Essays on Economic Stability and Growth, 2nd edition. London: Duckworth. Ketels, Christian (2013) ‘Recent Research on Competitiveness and Clusters: What Are the Implications for Regional Policy?’, Cambridge Journal of Regions, Economy and Society 6(2): 269–84. King, John (2009) Nicholas Kaldor. Great Thinking in Economics Series. London: Palgrave Macmillan. Kingombe, Christian and Dirk W. te Velde (2016) ‘The Role of Special Economic Zones in Manufacturing Development in Sub-Saharan Africa: Structural Transformation and Employment Creation’, in John Weiss and Michael Tribe (eds) Routledge Handbook of Industry and Development. Abingdon: Routledge, pp. 240–58. Koopman, Robert, William Powers, Zhi Wang, and Shang-Jin Wei (2010) Give Credit Where Credit Is Due: Tracing Value Added in Global Production Chains. Cambridge, MA: National Bureau of Economic Research. Krugman, Paul (1979) ‘Increasing Returns, Monopolistic Competition, and International Trade’, Journal of International Economics 9: 469–79. Krugman, Paul (1987) ‘Is Free Trade Passé?’, Economic Perspective, 1(2): 131–44. Krugman, Paul (1993) ‘Increasing Returns and Economic Geography’, Journal of Political Economy 99: 183–99. Kurz, Heinz Dieter and Neri Salvadori (1998) ‘The “New” Growth Theory: Old Wine in New Goatskins’, in Fabrizio Coricelli, Massimo di Matteo, and Frank Hahn (eds) New Theories in Growth and Development. London: Palgrave Macmillan, pp. 63–94. Kusago, Takayoshi and Zafiris Tzannatos (1998) Export Processing Zones: A Review in Need of Update. Social Protection Group, Human Development Network. Washington, DC: World Bank. Lane, Nathan (2017) ‘Manufacturing Revolutions: Industrial Policy and Networks in South Korea’. Working Paper. Available at http://www.eh.net/eha/wp-content/uploads/2017/08/Lane.pdf. Lavopa, Alejandro and Szirmai, Adam (2015) ‘Industrialisation in Time and Space’. UNU-MERIT Working Paper Series No. 2015–039. UNU-WIDER, Helsinki. Lawson, Tony (1997) Economics and Reality. London: Routledge. Leinbach, Thomas R. (1982) ‘Industrial Strategy in Malaysia: The Role of Export Processing Zones’, GeoJournal 6(5): 459–68. Lin, Justin Yifu and Ha-Joon Chang (2009) ‘Should Industrial Policy in Developing Countries Conform to Comparative Advantage or Defy It? A Debate between Justin Lin and Ha-Joon Chang’, Development Policy Review 27(5): 483–502. Long, Cheryl and Xiaobo Zhang (2011) ‘Cluster-based Industrialization in China: Financing and Performance’, Journal of International Economics 84(1): 112–23.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
62 Christopher Cramer and Fiona Tregenna Malmberg, Anders and Peter Maskell (2010) ‘An Evolutionary Approach to Localized Learning and Spatial Clustering’, in Ron Boschma and Ron Martin (eds) The Handbook of Evolutionary Economic Geography. Cheltenham: Edward Elgar, pp. 391–405. Marconi, Nelson, Cristina Froes Reis de Borja, and Eliane Cristina de Araújo (2016) ‘Manufacturing and Economic Development: The Actuality of Kaldor’s First and Second Laws’, Structural Change and Economic Dynamics 37: 75–89. Marshall, Alfred (1920) Principles of Economics, 8th edition. London: Macmillan. Martins, Pedro M. G. (2018) ‘Structural Change in Ethiopia’, Development Policy Review 36: 183–200. Mazzucato, Mariana (2011) ‘The Entrepreneurial State’, Soundings 49(49): 131–42. McCormick, Dorothy (1999) ‘African Enterprise Clusters and Industrialization: Theory and Reality’, World Development 27(9): 1531–51. McNamara, Peter (1998) Political Economy and Statesmanship: Smith, Hamilton, and the Foundation of the Commercial Republic. DeKalb, IL: Northern Illinois University Press. Mendez-Ramos, Fabian and Nina Paustian (2017) ‘Measurement and Patterns of World Agribusiness Trade’. Research Policy Brief. Global Knowledge & Research Hub in Malaysia. World Bank, Washington, DC. Monteiro, Carlos Augusto, Geoffrey Cannon, Jean-Claude Moubarac, Renata Bertazzi Levy, Maria Laura da Costa Louzada, and Patrícia Constante Jaime (2018) ‘The UN Decade of Nutrition, the NOVA Food Classification and the Trouble with Ultra-Processing’, Public Health Nutrition 21(1): 5–17. Mukkala, Kirsi (2004) ‘Agglomeration Economies in the Finnish Manufacturing Sector’, Applied Economics 36(21): 2419–27. Oqubay, Arkebe (2015) Made in Africa. Oxford: Oxford University Press. Oqubay, Arkebe (2019) ‘Structure and Performance of the Ethiopian Manufacturing Sector’, in Fantu Cheru, Christopher Cramer, and Arkebe Oqubay (eds) The Oxford Handbook of the Ethiopian Economy. Oxford: Oxford University Press. Palma, Gabriel (2005) ‘The Seven Main “stylized facts” of the Mexican Economy Since Trade Liberalization and NAFTA’, Industrial and Corporate Change 14(6): 941–91. Porter, Michael (1998) ‘Clusters and the New Economics of Competition’, Harvard Business Review 76(6): 77–90. Prebisch, Raul (1950) ‘The Economic Development of Latin America and its Principal Problems’, Economic Bulletin for Latin America, No. 7, United Nations Department of Economic Affairs, New York, NY. Prebisch, Raul (1963) Towards a Dynamic Development Policy for Latin America. New York: United Nations. Ricardo, David (1821) The Principles of Political Economy (1963 edition). Irwin: Homewood. Rodrik, Dani (2012) ‘Unconditional Convergence in Manufacturing’, Quarterly Journal of Economics 128(1): 165–204. Rodrik, Dani (2016) ‘Premature Deindustrialization’, Journal of Economic Growth 21: 1–33. Schmitz, Hubert and Khalid Nadvi (1999) ‘Clustering and Industrialization: Introduction’, World Development 27(9): 1503–14. Schrank, Andrew (2001) ‘Export Processing Zones: Free Market Islands or Bridges to Structural Transformation?’, Development Policy Review 19(2): 223–42. Smith, Adam (1994) The Wealth of Nations. New York: Random House. Spinanger, Dean (1984) ‘Objectives and Impact of Economic Activity Zones—Some Evidence from Asia’, Weltwirtschaftliches Archiv 120(1): 64–89. Stein, Howard (2011) ‘Africa, Industrial Policy, and Export Processing Zones: Lessons from Asia’, in Akbar Noman, Kwesi Botchwey, Howard Stein, and Joseph E. Stiglitz (eds) Good Growth and Governance in Africa: Rethinking Development Strategies. Oxford: Oxford University Press, pp. 322–44. Su, Dan and Yang Yao (2017) ‘Manufacturing as the Key Engine of Economic Growth for Middle-income Economies’, Journal of the Asia Pacific Economy 22(1): 44–70.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Heterodox Approaches to Industrial Policy and the Implications 63 Szirmai, Adam (2012) ‘Industrialisation as an Engine of Growth in Developing Countries, 1950–2005’, Structural Change and Economic Dynamics 23(4): 406–20. Szirmai, Adam and Bart Verspagen (2015) ‘Manufacturing and Economic Growth in Developing Countries, 1950–2005’, Social Change and Economic Dynamics 34: 46–59. Tang, Vanessa T. (2015) ‘Does Learning by Importing, Self-selection of Markets and Financial Innovation Matter for Exporting Firms in Special Economic Zones?’, Investment Management and Financial Innovations 12(1): 198–206. Thirlwall, Anthony Philip (1983) ‘A Plain Man’s Guide to Kaldor’s Growth Laws’, Journal of PostKeynesian Economics 5(3): 345–58. Thirlwall, Anthony Philip (2002) The Nature of Economic Growth: An Alternative Framework for Understanding the Performance of Nations. London: Edward Elgar. Timmer, Marcel Peter, Abdul Azeez Erumban, Bart Los, Robert Stehrer, and Gaaitzen Johannes de Vries (2014) ‘Slicing up Global Value Chains’, Journal of Economic Perspectives 28(2): 99–118. Tregenna, Fiona (2009) ‘Characterising Deindustrialisation: An Analysis of Changes in Manufacturing Employment and Output Internationally’, Cambridge Journal of Economics 33(3): 433–66. Tregenna, Fiona (2013) ‘Manufacturing Productivity, Deindustrialization and Reindustrialization’, in Adam Szirmai, Wim Naudé, and Ludovico Alcorta (eds) Pathways to Industrialization in the Twenty-first Century: New Challenges and Emerging Paradigms. Oxford: Oxford University Press, pp. 76–101. Tregenna, Fiona (2015) ‘Deindustrialisation, Structural Change and Sustainable Economic Growth’. MERIT Working Paper No. 032. United Nations University—Maastricht Economic and Social Research Institute on Innovation and Technology (MERIT), Maastricht, the Netherlands. Verdoorn, Petrus Johannes (1993) ‘On the Factors Determining the Growth of Labor Productivity’, in Luigi Pasinetti (ed.) Italian Economic Papers, Vol. II. Oxford: Oxford University Press. Wade, Robert (1990) Governing the Market: Economic Theory and the Role of Government in East Asian Industrialization. Princeton, NJ: Princeton University Press. Warr, Peter and Jayant Menon (2016) ‘Cambodia’s Special Economic Zones’, Journal of Southeast Asian Economics 33(3): 273–90. Warr, Peter George (1989) ‘Export Processing Zones: The Economics of Enclave Manufacturing’, Research Observer 4(1): 65–88. Weiss, John and Hossein Jalilian (2016) ‘Manufacturing as an Engine of Growth’, in John Weiss and Michael Tribe (eds) Routledge Handbook of Industry and Development. Abingdon: Routledge, pp. 26–37. Whitefoot, Kate S., Walter D. Valdivia, and Gina C. Adam (2015) Innovation and Manufacturing Labor: A Value-Chain Perspective. Centre for Technology Innovation. Washington, DC: Brookings. Widodo, Wahyu, Ruhul Salim, and Harry Bloch (2015) ‘The Effects of Agglomeration Economies on Technical Efficiency of Manufacturing Firms: Evidence from Indonesia’, Applied Economics 47(31): 3258–75. Woo-Cumings, Meredith (1999) The Developmental State. New York: Cornell University Press. Wood, Adrian (2017) ‘Variation in Structural Change around the World, 1985–2015’. WIDER Working Paper. UNU-WIDER, Helsinki. Wood, Adrian and Jörg Mayer (1998) ‘Africa’s Export Structure in a Comparative Perspective’, SSRN Electronic Journal. Available at https://doi.org/10.2139/ssrn.141202. Young, Allyn (1928) ‘Increasing Returns and Economic Progress’, Economic Journal 38(152): 527–42. Zeng, Douglas Zhihua (2011) ‘How Do Special Economic Zones and Industrial Clusters Drive China’s Rapid Development?’. Policy Research Working Paper No. 5583. World Bank, Washington, DC.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
chapter 4
The Economics of I n novation behi n d Cluster Dy na mic Proce sse s Michael H. Best
4.1 Introduction Industrial districts and clusters of skill-related companies that enjoy the agglomeration benefits of co-location are present across the whole spectrum of the world’s regions and countries from the least to the most technologically advanced. Cluster dynamic processes, however, are at the core of rapid growth and transformational experiences and of the high-tech regions often described as innovation hubs. In this chapter, I review major scholarly contributions to a capabilities and innovation perspective, to an economics of industrial change and transformation. The purpose is to better understand the dynamic processes, organizers, and drivers of transformational experiences. I conclude with the claim that infrastructural agencies are mediating institutions by which central and local governments together implement industrial policies. I begin by presenting an expositional schema of cluster dynamic processes.
4.2 Cluster Dynamics in a Circular Flow Schema Economic ‘miracles’ involve the creation or reorganization of firms and sectors on an industrial scale. Figure 4.1 illustrates the core processes of cluster dynamics. On the right is the entrepreneurial firm. Think of a firm as producing both products and knowledge in the pursuit of a distinctive capability to survive if not thrive in the marketplace. In this process, the firm comes under market pressure to focus on its core capability and network for
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
The Economics of Innovation behind Cluster Dynamic Processes 65 Industrial District specialization and speciation dynamics
Inter-Firm Networks open-systems dynamics
Entrepreneurial Firms internal dynamics New Firms technological diversification
Figure 4.1 Cluster growth dynamics Source: Best (2001).
complementary capabilities. This creates opportunities for other firms, perhaps a new firm, to partner with the original firm by focusing on a complementary capability. The resulting inter-firm network is shown on the left. An entrepreneurial firm can also create a technodiversification niche opportunity for a new entrant to pursue rejected capability development choices, as illustrated in the bottom sphere. A proliferation of partnering enterprises in the form of open-system networks creates opportunities for new and repositioning enterprises to specialize within pre-existing networks without having to invest in the whole range of activities that go into the production of most products. The proliferation of nodes cutting across networks creates new opportunities for new product development and technology management across the system of specialist and networked enterprises, illustrated in the top sphere. This enables the emergence of yet more entrepreneurial firms, as shown in the right sphere, creating yet another round of focusing and networking, followed by networking and focusing, and in the process cumulatively and collectively advancing the region’s distinctive skills and production capabilities.
4.3 Marshall and Penrose: Linking Intra-Firm and Inter-Firm Growth Dynamics We start with Alfred Marshall; not the caricatured Marshall of mainstream economic geography, but the real Marshall, an astute observer and theorist of economic progress. Marshall was followed by Edith Penrose and what became known as the learning theory of the business enterprise. For Marshall, economic progress is explained in terms of knowledge and organization or, more precisely, increases in the complexity of organizations and in sources or creation of new knowledge. Marshall drew on Adam Smith’s axiom that advances in productivity are a consequence of increasing specialization of labour combined with more ‘intimate connections’ integrating increasingly differentiated activities (Marshall [1890] 1920: 241).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
66 Michael H. Best Marshall identified three specific forms of organizational sources of productivity improvement: inside the firm, amongst various businesses in the same ‘trade’ or sector, and various trades relative to one another (Marshall [1890] 1920: 138–9).1 The firm, to paraphrase Edith Penrose, is a system for the production of goods and new knowledge-based resources. For Brian Loasby, Penrose independently originated a dialectical process of resource creation internal to the firm that was invented by Marshall: ‘The transformation of novelty into routine releases cognitive resources for the imagination of further novelty, which, if successful, becomes embodied in further routines.’ This knowledge-increasing dialectic, in Loasby’s interpretation, is the basis of Marshall’s theory of economic progress; but it ‘requires various forms of organization to realize its potential’ (Loasby 2009: 81). Penrose extended Marshall in important ways. Her theory of the growth of the firm locates the novelty/routine dialectic within a second, encompassing dialectic between a company’s ‘core competence’, or unique capability, and market opportunity. Refinements in a firm’s distinctive capabilities enhance its capacity to imagine, identify, and seize emerging market opportunities and, in the process, offer insights for a new round of capability refinement. Success at exploiting emergent market opportunities requires designing, building, and/or at least tweaking existing productive capabilities—a learning process that both increases the firm’s knowledge base and, once completed, releases design and development resources to be directed to identifying new market opportunities. Furthermore, capabilities for Penrose are about collective cognition;2 they cannot be reduced to individual expertise. Teams of people working together can create and act on knowledge beyond the capacities of individuals working alone. As Penrose often stated, capabilities are activities you cannot do alone or at once. They take teamwork and time.
4.4 Kuznets, Penrose, and Sector Transitions Since Penrose did not articulate or seek a theory of the growth of sectors, regional clusters, or national economies, our challenge is to identify an aggregate economic growth perspective complementary to that of Penrose’s enterprise growth dynamics. Simon Kuznets, whose Six Lectures on Economic Growth were published in 1959, was awarded the third Nobel Prize 1 The quote in full brings out a fourth factor which will become important in section 4.8. 2 A 2002 Santa Fe Institute Collective Cognition workshop characterized the idea as follows: Many forms of individual cognition are enhanced by communication and collaboration with other intelligent agents. We propose to call this collective cognition, by analogy with the well known concept of collective action. People (and other intelligent agents) often ‘think better’ in groups and sometimes think in ways which would be simply impossible for isolated individuals. Perhaps the most spectacular and important instance of collective cognition is modern science. An array of formal organizations and informal social institutions also can be considered means of collective cognition. For instance, Hayek famously argued that competitive markets effectively calculate an adaptive allocation of resources that could not be calculated by any individual market-participant. http://www.santafe.edu/~dynlearn/colcog/ (accessed 15 July 2008).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
The Economics of Innovation behind Cluster Dynamic Processes 67 in economics in 1971 ‘for his empirically founded interpretation of economic growth which has led to new and deepened insight into the economic and social structure and process of development’. Although they overlapped only briefly, Kuznets and Penrose were both at Johns Hopkins University in the 1950s (A. Penrose 2017). Kuznets’ national economy growth frame focused attention on new industries replacing old industries: [A] sustained high rate of growth depends upon a continuous emergence of new inventions and innovations, providing the bases for new industries whose high rates of growth compensate for the inevitable slowing down in the rate of investment and innovation, and upon the economic effects of both, which retard the rates of growth of the older industries.
But just as Penrose did not extend the theory of the growth of the firm to sectors, Kuznets’ theory of growth does not go inside the firm.3 Nevertheless, the growth dynamics of the two perspectives, although focusing on different domains, share common themes and one important institution. The institutionalization of technological research is a theme common to both growth perspectives. For Kuznets growth and prosperity in the long run were about ongoing sectoral trans formation over time driven by the marriage of science and industry. Science was the creative force and transformation was about the emergence of new more technologically advanced industrial sectors. Penrose focused on the institutionalization of the research department within the large corporation, on the technology capability, and on new product development processes. In the long run the profitability, survival, and growth of a firm does not depend so much on the efficiency with which it is able to organize the production of even a widely diversified range of products as on the ability of the firm to establish one or more wide and relatively impregnable ‘bases’ from which it can adapt and extend its operations in an uncertain, changing, and competitive world. It is not the scale of production nor even, within limits, the size of the firm, that are the important considerations, but rather the nature of the basic position that it is able to establish for itself (Penrose 1959: 137). Penrose wrote: The industrial research laboratory . . . is the logical response of the individual firm to the challenge inherent in the Schumpeterian ‘process of creative destruction’. The normal activities of large firms cannot be discussed without reference to it. (Penrose 1959: 112)4
The implicit assumption of conventional macroeconomics is that aggregate growth can be conceptualized without controlling for what goes on inside the production system. 3 Like Penrose, for Kuznets new product development was important, but unlike Penrose only as an outcome, not as an enterprise production capability: ‘A country’s economic growth may be defined as a long-term rise in capacity to supply increasingly diverse economic goods to its population, this growing capacity is based on advancing technology and the institutional and ideological adjustments that it demands’ (opening sentence of the Nobel Prize Lecture by Simon Kuznets, 1971). 4 SMEs lack the scale and revenues to fund a research lab and thereby develop a pipeline of new products: ‘for the small firm the use of its resources for “research” in general is as likely to be wasteful as it is profitable unless the firm has specific and reasonably original ideas upon which it is working’ (Penrose 1959: 115).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
68 Michael H. Best Macroeconomic theorists, Kuznets included, do not address the institutional mechanisms by which new technologies are developed, integrated into production processes, or adopted and modified on the shop floor to achieve their full potential. This is a major weakness: without a concept of the Penrosian distinctive capability development and market opportunity dynamic, Kuznets’ and Schumpeter’s conceptual frameworks fail to explain the rapid growth of new sectors and fall prey to a linear, science-push theory of technological and industrial innovation. At the same time, Penrose’s theory of the growth of the firm is a theory of enterprise innovation, not a theory of industrial innovation. The innovative firm is the pivotal link in an industrial innovation chain that takes us into the terrain of inter-firm relations, extrafirm infrastructures, and policy frameworks.
4.5 Veblen and Young: Cumulative Increasing Returns The challenge of linking the enterprise capability dynamic of Penrose and the sectoral transition dynamic of Kuznets draws our attention to a conceptual space between the two. A third, intermediate innovation dynamic involving systemic interactions that generate dynamic increasing returns was originally formulated by Allyn Young. Young’s dynamic increasing returns, combined with the innovation dynamics of Penrose and Kuznets, is the key to a theoretical understanding of a sector as constituted by increasingly differentiated enterprises and, consequently, to an understanding of the real-world dynamics of new rapidly growing sectors. Young’s Increasing Returns and Increases in Prosperity, published in 1928, outlines a cumulative innovation dynamic involving the interplay between the pursuit of distinctive capability by individual enterprises and a process of increasing differentiation amongst enterprises. In fact, Thorstein Veblen, the most prominent figure in the American institutional school of economics, coined the term ‘cumulative causation’ at the turn of the century (Hodgson 2004: 152).5 But it was with Young’s characterization of ‘increasing returns’ that the concept took on the modern systems theory connotation of non-linear processes of positive feedback.6 He wrote: Every important advance in the organisation of production, regardless of whether it is based upon anything which, in a narrow or technical sense, would be called a new ‘invention’, or involves a fresh application of the fruits of scientific progress to industry, alters the conditions of industrial activity and initiates responses elsewhere in the industrial structure which in turn have a further unsettling effect. Thus change becomes progressive and propagates itself in a cumulative way. 5 ‘Any evolutionary science . . . is a theory of a process, of an unfolding sequence . . . of cumulative causation’ (Veblen 1898: 375–8). The institutional school was prominent in American economics journals and research methods until the postwar emergence and triumph of general equilibrium theory and mathematical modelling. 6 Lachlin Currie, Roosevelt’s personal economic adviser, had been a student of Young’s at Harvard.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
The Economics of Innovation behind Cluster Dynamic Processes 69 Young rejects linear cause and effect with a systemic understanding of innovation processes: The mechanism of increasing returns is not to be discerned adequately by observing the effects of variations in the size of an individual firm or of a particular industry, for the progressive division and specialisation of industries is an essential part of the process by which increasing returns are realised. What is required is that industrial operations be seen as an interrelated whole. Second, the securing of increasing returns depends upon the progressive division of labour, and the principal economies of the division of labour, in its modern forms, are the economies which are to be had by using labour in roundabout or indirect ways. (Young 1928: 539)
Dynamics of firm and industry, specialization and differentiation, drive innovation: With the extension of the division of labour among industries the representative firm, like the industry of which it is a part, loses its identity. Its internal economies dissolve into the internal and external economies of the more highly specialised undertakings which are its successors, and are supplemented by new economies. (Young 1928: 538)
Young explicitly, and with great respect, extends Smith’s driver of prosperity from a focus on increasing division of labour within a production unit to increasing differentiation of labour both within and amongst increasingly specialized enterprises. Smith integrated his logic of increasing division of labour within a production process to a discovery process by which innovations in production methods emerged. Young adds system feedback or cumulative effects to Smith’s account. Here we have a concept of inter-firm process by which market and entrepreneurial opportunities are created. Penrose’s individual enterprise capability/opportunity dynamic is extended into a dynamic process integral to the ongoing process of increasing division of labour across sectors and industry as a whole. It is a theory of industrial organization that accounts for the propagation of organizational increasing returns. The process of increasing division of labour within the firm is one that fosters the discovery of new methods, but Smith’s extent of the market that limits the division of labour is integrated by Young into a process of increasing differentiation of enterprises and increasing market opportunities by interactive or feedback effects. Focusing on core capabilities and networking for complementary capabilities is an opportunity creation process for partner companies, and their successful response by specializing increases the market for the originating firm. The self-organizing driver of economic progress is an open-system business model in which firms focus and network, and networks, in turn, expand opportunities to focus and innovate. The macro growth terrain of Kuznets and the micro growth dynamic of Penrose are inter connected by Young’s concept of cumulative interactions generating increasing returns. As firms and nations develop more advanced production capabilities their competitive advantage shifts to more technologically advanced production activities and products. Here Young’s doctrine of increasing returns complements Kuznets’ claim that the marriage of production and science drives the transition to technologically more advanced sectors.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
70 Michael H. Best This brings into the transformational frame system integration as a fundamental rinciple of production and organization. The principle of system integration is manifest in p the organizational capability of firms, individually and networked, to foster rapid technological change. Entrepreneurial firms drive innovation and sectoral transitions but they do not do so alone, but as members of networked groups of collectively innovating enterprises, often within regionally based innovation systems. The application of the principle of system integration to business organization means integrating an ongoing technology management capability into a production system. The effect is a network or cluster of entrepreneurial firms in which design is decentralized within the enterprise and diffused amongst networked enterprises. It is a business model ideally suited to product-led strategies and technological innovation.
4.6 Jane Jacobs and Marshall’s Large Cities Marshallian industrial districts and Jane Jacobs’ city growth dynamics can be interpreted as complementary sub-systems. Marshall’s concept of industrial district did not preclude the idea of the co-location of various sectors and mutual interactions contributing to the progress of each. In fact, he saw at least one important advantage: ‘A district which is dependent chiefly on one industry is liable to extreme depression . . . this evil . . . is in a great measure avoided by those large towns or large industrial districts in which several distinct industries are strongly developed’ (Marshall [1890] 1920: 273). But the advantages of a composite industrial district with multiple sectors extend beyond offsetting depression. Variety and diversity contribute to economic progress: ‘The advantages of variety of employment are combined with those of localized industries in some of our manufacturing towns, and this is a chief cause of their continued growth’ (Marshall 1890 [1920]: 226). The idea of a large or ‘composite’ industrial district of multiple interrelated sectors deepens the city growth dynamics of Jane Jacobs. It calls attention to what Jacobs described as epigenetic as distinct from preformative growth. Preformative growth assumes all firms expanding at the same pace, producing more of the same goods, sectors expanding doing the same production activities, and cities producing more of the same mix of products. Epigenetic growth is a dynamic, innovative process of increasingly differentiated but interactive enterprises embedded in capability-enhancing industrial ecosystems. Jacobs examined cities historically, giving special attention to cases of rapid growth to discover patterns of complex interactions in their most pronounced forms. She conceptualized cities as the engines of economic advancement, providing markets, jobs, capital, and technology for themselves, the regions around them, and other cities as well. Jacobs’ city growth dynamics are expressed as the rate of addition of new goods and services. Sustained city growth is simultaneously a process of increasing differentiation of skills and an experimental process of new product development and sector evolution. In her words: ‘Existing divisions of labor multiply into more divisions of labor by grace of intervening added activities that yield up more sums of work to be divided’ (Jacobs 1969: 58).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
The Economics of Innovation behind Cluster Dynamic Processes 71 The ‘intervening added activities’ are described in terms of new ‘work’ combined with multiple trials and errors linking the old to the new divisions of labour. The increasing differentiation in skills increases the opportunities for innovation and for sustained city growth. Jacobs offers a rich language such as ‘symbiotic nests of suppliers and their markets’ (Jacobs 1984: 76) and ‘lateral interrelationships’ to characterize dynamic, mutual adjustment processes amongst firms and sectors. In an observation on the sources of growth of ‘creative cities’ in general, Jacobs wrote: Tightly packed bunches of symbiotic enterprises . . . have always been the strengths and wonders of creative cities . . . the huge collections of little firms, the symbiosis, the improvisations, the ease of breakaways, the flexibility, the economies, efficiencies, and adaptability—these are precisely the assets that, among other things, have always made successful and significant importreplacing a process realizable only in cities and their nearby hinterlands. (1984: 40–1).
Perhaps Jacobs would not have objected to replacing the term ‘assets’ with ‘capabilities’. Firms cumulatively and collectively developed distinctive regional capabilities, which in turn created a knowledge resource base in deep craft skills for the next generation of innovative activities. What is evident in economic miracles is the scale of change and growth. It is not just one firm; it is thousands of firms and it is not just one sector but dozens of sectors and not just one city but multiple cities. Jacobs’ focus on city growth dynamics takes the concept of industrial district to a scale at least an order of magnitude beyond the small-city examples in Marshall, and the literature on the ‘Third Italy’ to the scale of transformative experiences in large national economies. A few brief comments on China are instructive.
4.7 Industrial Districts and Rapid Growth in China In a study of the sudden emergence of industrial districts in China, Jici Wang and Lixia Mei document a striking parallel to the industrial dynamic processes involving SMEs and city growth dynamics, but on a massively larger scale, that scholars found during the Italian ‘miracle’ years (2009: 598–612). Mao’s experiment in collectivization and communal organization of production was officially abolished in 1982 and replaced by a new ‘market socialist’ production system under Deng Xiaoping (1978–92). Ezra Vogel quotes comments by Deng from 1987: ‘In the rural reform our greatest success—and it is one we had by no means anticipated—has been the emergence of a large number of enterprises run by villages and townships. They were like a new force that just came into being spontaneously’ (Vogel 2011: 445). The ‘new force’ in the form of town and village enterprises (TVEs) was triggered by communes ‘contracting responsibility down to households’ in response to widespread starvation in the late 1970s. By 1982, 98 per cent of rural households had both national plan contracts and economic freedom. The result was a massive increase in the productivity of land as agricultural households undertook intensive garden farming and markets emerged in village towns.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
72 Michael H. Best The reforms created a vast surplus labour force which deluged into cities (600). In the process, household production units and commune workshops transitioned into TVEs (neither state-owned enterprises nor private enterprises) and small firms proliferated ‘like bamboo’. Household production empowerment triggered city–region growth dynamics (Jacobs acknowledges Ronald Dore’s (1980) description of the process in Japan’s transformation (1984: 47ff.)). The output of TVEs rose from 48 to 1800 billion yuan between 1978 and 1992 and small private enterprises were legalized. After 1987, the restriction to seven of the number of employees an owner (capitalist?) could employ was abolished. Thus began city–region growth dynamics and the creation of industrial districts on a scale beyond what anyone could have imagined. Wang and Mei identified some 536 industrial districts with 1,679,803 SMEs, employing over 45 million people. The industrial districts specialized in many of the same sectors as in Italy: textiles and apparel, footwear, furniture, home electrical appliances, toys, motorcycles, all basically labour intensive at the lower end of the production capability spectrum. But thereafter China’s and Italy’s industrial districts underwent different developmental trajectories in several important ways with important implications for how we think about industrial growth and industrial transformation. First, China’s rapid growth evolved in the age of global production networks (GPNs) and massive expansion of foreign investment. On the one hand, the scale of the large industrial districts, combined with low wages, was attractive to MNEs as outsourcing targets, and on the other GPNs were attractive partners for marketing, technology acquisition, and managerial learning (Gereffi 1999; Gereffi et al. 2005; Schmitz and Knorringa 2000; Ernst and Kim 2002). China’s opening to the West involved bold leadership. It meant gambling that a latecomer strategy could be implemented in which the positive productive spillover effects of integrating China’s nascent production system into GPNs owned and managed by Western MNEs would, in the end, dominate the crowding-out effects on China’s indigenous manufacturing enterprises (Wang and Mei 2009: 609).7 So far, with qualifications, it seems to have worked. Second, China’s policy framework under Deng was strategic in a global context. China did not aspire to function indefinitely as an offshore manufacturing re-export platform for foreign-headquartered enterprises. China’s policy framework was informed by Deng’s visits both to Japan’s economic model of world-class manufacturing and indigenous enterprises, and to Singapore’s strategic approach to leveraging FDI to access and assimilate advanced management skills and technologies. Deng saw Japan’s and Singapore’s policy frameworks as alternative pathways to organize industry transitions, from labour-intensive, through skill-intensive, capital-intensive, and technology-intensive to knowledge-intensive activ ities and sectors.8 Their success came not from following a generic institutional blueprint but from crafting a strategy informed by the unique context of each country. Both used latecomer advantage to access America’s vast science and technology infrastructure. 7 Penrose cautioned governments to maintain a posture of ‘circumspect vigilance’, with respect to the risks to structural autonomy of the integration of productive activities within MNEs given the ‘economic and political power and operations of the great private bureaucracies of international business’ (Penrose 1992: 250; Blundel 2014). 8 Singapore’s Economic Development Board created a series of focused production and business development agencies designed to learn about and foster indigenous innovation, albeit in the context of a tiny nation without natural resources or a low-wage labour force. See Chapter 35.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
The Economics of Innovation behind Cluster Dynamic Processes 73 The key, in part, was to take full advantage of America’s underinvestment in manufacturing capabilities (Best 2018). Third, Marshallian industrial districts may be an important stage in the emergence of networked groups of SMEs enjoying both Marshallian and Jacobian externalities. But sustainable growth depends upon transitioning to industrial districts within industrial ecosystems in which both local and national governments work together to craft and undertake policy frameworks that combine centralized strategic policy planning at the national level with decentralized authority and accountability at the local level. The Chinese multi-layered economic governance framework, while not unique, illustrates the limits of the industrial district model, in which local governments operate largely without the benefit of complementary intermediate and national layers of government. The Italian industrial districts enjoyed a structural competitive advantage contra regions in which enterprises pursued autarchic, vertical integration strategies to compete alone in the global economy. The opensystem business model that throve in the Italian industrial districts in the 1970s and 1980s enabled firms to pursue a strategy of focus on a core capability and partner/network for complementary capabilities.9 But the Italian central government did not seize the oppor tunity of city–region growth dynamics to build an enabling national policy framework that supported sectoral evolution within regional industrial ecosystems. China, however, has to all appearances done so to some extent. The result is the rapid growth of a whole series of ‘second-tier’ cities with governments that organized cluster dynamic growth processes.10
4.8 Policy Frameworks: The Capability Triad and Extra-firm Infrastructures Transformative growth experiences are many. But they do not just happen. Extraordinary leaders are important in responding to daunting challenges by crafting appropriate strategic policy frameworks and organizing the modus operandi of their implementation at both government and enterprise levels. Only governments can organize change programmes across the economy and on the scale required to overcome low-productivity structural inertia. But the power to drive capability development and industrial change does not lie in the legitimacy of government to organize change agendas. Ultimately, it is business enterprises that implement change programmes, transition up the production capability spectrum, and make industrial innovation happen. But they do not do so alone. Transformational growth takes us beyond the terrain of the economics of equilibrium and optimal resource allocation, although economics is critical to it. But it is an economics 9 For example, the structural advantage of the ‘Third Italy’ or Emilian Romagna open-system business model rapidly overwhelmed the North London furniture district with the entry of the United Kingdom into the European Economic Community (Best 1990: chs 7–8). 10 West Germany’s post-war economic miracle illustrates the importance of multi layer government in sustaining regional industrial ecosystems (Best 2018: ch. 5). Germany achieved this in the form of ‘col lective property’ and local government that integrates strategic planning with local responsibility and powers—similar to M-form or multi-divisional enterprise. The Italian industrial districts were not leveraged into sustained development.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
74 Michael H. Best
Skill Formation
Business Model
Production Capabilities
Figure 4.2 Capability triad of competitive advantage Source: Best (2018).
of capability development and mutual adjustment processes in which changes in business, production, and skill formation are inextricably bound together. The three domains are not separable, additive components of growth, but mutually interdependent sub-systems of a single developmental process. None of the three can contribute to growth independently of mutual adjustment processes involving all three. I call this the capability triad (see Figure 4.2).11 In rapid growth success stories economic governance is informed by a productioncentric awareness of where a region or country’s enterprises and their production capabilities fit within the global economy. The policy challenge is to design and enact a unified set of extra-firm specialist infrastructures that foster capability development change programmes within and across a population of companies, driving an order-of-magnitude advance in performance standards. New product development and technology management are examples of production capabilities. Science and technology, vocational education, and development finance are examples of extra-firm infrastructures.12 Once again, Marshall’s focus on organization and knowledge was prescient. Marshall referred to ‘external economies which arise out of the collective organization of the district as a whole’ (Marshall [1890] 1920: xii). An example is a region’s extra-firm infrastructures that individual companies can draw upon to their own internal benefit. At the same time, extra-firm infrastructures are capability development levers by which governments can construct and implement industrial policy frameworks. An examination of extra-firm infrastructures as an economic category takes us into the domain of economic governance 11 The policy challenge is to foster a set of extra-firm infrastructures by which change programmes in all three domains are in sync with one another. This targets the inter-relationships rather than each elem ent within a system; it also means that for a design change in any one element to be successful, the other elements need to be reconfigured to take full advantage. W. Edwards Deming is credited by the Japanese with introducing this fundamental organizational principle which informed their post-war rapid growth experience (Best 1990: 158–61). 12 Charles Babbage articulated the role of science and technology infrastructure as an instrument of industrial innovation (1832; Best 2018: 94–102). Marshall would not have been surprised about the role of the state in the early stages of the development of modern industries and emergent industrial ecosystems. The full quote abbreviated above is: ‘Organization aids knowledge; it has many forms, e.g. that of a single business, that of various businesses in the same trade, that of various trades relatively to one another, and that of the State providing security for all and help for many’ (Marshall [1890] 1920: 138–9).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
The Economics of Innovation behind Cluster Dynamic Processes 75 and the respective roles of central, provincial, and local government. It casts industrial policy in a new light. The organizational principle involves the same combination of centralization of strategic planning at the top and decentralized powers and responsibility by which 1920s American big business established industrial leadership in the form of the multi-divisional enterprise and which was adopted by the American government’s World War II policy framework (Best 2018). Central government has the legitimacy to pass regulatory laws and to establish strategic priorities and budgetary powers to finance public goods and services, including material and immaterial infrastructures. The central government level is where economic policy frameworks are articulated, sectoral transitions are envisioned and planned, the missions of department agencies are tasked and coordinated, the requisite business and production capabilities development goals are formulated, and the infrastructure agendas required to implement the change programmes are negotiated and unified. But it is only city and provincial governments that have the intimate knowledge of local economic conditions that are required to craft and implement a development strategy. Local governments have the power to convene in the same room the actors required to turn strategies into realities. This includes representatives from businesses, unions, banks, educational institutions, tax authorities, and infrastructural agencies. It involves policymaking powers to coordinate, unify, and shape the activities of the infrastructural agencies. Each layer of government depends upon the other to implement policy frameworks. The concept of industrial ecosystem is a modern expression of Marshall’s ‘collective organization of the district as a whole’: capability development infrastructures are external to individual firms but internal and constitutive of a region’s industrial ecosystem, not separate from the region’s economy. They are a way to think of a region’s specialized and differentiated enterprises and associated extra-firm infrastructures and agencies as a system of actual and potential inter-firm connections and networks, formal and informal. In hightech regional innovation systems such as Greater Boston, the sheer density of potential connections constitutes a regionally unique resource for fostering industrial transitions and addressing technological and business development challenges by any firm embedded within the industrial ecosystem (Best 2018).13 Marshall has the first and last word: The trade of one individual with another is mainly of private concern: while the causes which enable large quantities of anything to be made of foreign trade at a profit, generally lie deep down in resources and faculties that are not wholly individual, but are in great part the col lective property of a nation as a whole. A country’s foreign trade is something more than a number of dealings between individuals at home and abroad; it is the outcome of relations in which the industries that belong to her, that are part of her life, and embody much of her character, stand to the industries of other countries. (Marshall 1919: 4)
13 The complex whole comprises, in the words of Loasby, ‘different kinds of connections, constituting different forms of organization and many variations within each form, [which] encourage different combinations of evolved specialist knowledge and skills and direct the development of particular kinds of new knowledge and skills’ (Loasby 2009: 79–80).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
76 Michael H. Best
References Babbage, Charles (1832; reprint 2005) On the Economy of Machinery and Manufactures. Cirencester: Echo Library. Best, Michael (1990) The New Competition. Cambridge, MA: Harvard University Press. Best, Michael (2001) The New Competitive Advantage. Oxford: Oxford University Press. Best, Michael (2018) How Growth Really Happens: The Making of Economic Miracles through Production, Governance, and Skills. Princeton, NJ: Princeton University Press. Blundel, Richard (2014) ‘ “Circumspect Vigilance”: Reconsidering the Implications of Edith Penrose’s Single Argument for Sustainable Economic Development’, in Edith Penrose Centenary Conference, 14–15 November, SOAS, University of London. Dore, Ronald (1980) Shinohata: A Portrait of a Japanese Village. New York: Pantheon Books. Ernst, Dieter and Linsu Kim (2002) ‘Global Production Networks, Knowledge Diffusion, and Local Capability Formation’, Research Policy 31(8–9): 1417–29. Gereffi, Gary (1999) ‘International Trade and Industrial Upgrading in the Apparel Commodity Chain’, Journal of International Economics 48(1): 37–70. Gereffi, Gary, John Humphrey, and Timothy Sturgeon (2005) ‘The Governance of Global Value Chains’, Review of International Political Economy 12(1): 78–104. Hodgson, Geoffrey (2004) The Evolution of Institutional Economics. London: Routledge. Jacobs, Jane (1969) The Economy of Cities. New York: Random House. Jacobs, Jane (1984) Cities and the Wealth of Nations: Principles of Economic Life. New York: Random House. Kuznets, Simon (1959) Six Lectures on Economic Growth. New York: The Free Press. Kuznets, Simon (1971) ‘Modern Economic Growth: Findings and Reflections’, Nobel Prize Lecture, 11 December. Loasby, Brian (2009) ‘Industrial Districts in Marshall’s Economics’, in Giacomo Becattini, Marco Bellandi, and Lisa De Propris (eds) The Handbook of Industrial Districts. Cheltenham: Edward Elgar, pp. 78–89. Marshall, Alfred (1919) Industry and Trade. London: Macmillan. Marshall, Alfred ([1890] 1920) Principles of Economics, 8th edition. London and New York: Macmillan. Penrose, Angela (2017) No Ordinary Woman: The Life of Edith Penrose. Oxford: Oxford University Press. Penrose, Edith (1959) The Theory of the Growth of the Firm, 1st edition. Oxford: Basil Blackwell and New York: John Wiley and Sons. Penrose, Edith (1992) ‘Economic Liberalization: Openness and Integration—But What Kind?’, Development Policy Review 10: 237–54. Schmitz, Hubert and Peter Knorringa (2000) ‘Learning from Global Buyers’, Journal of Development Studies 37(2): 177–205. Veblen, Thorstein (1898) ‘Why Is Economics Not an Evolutionary Science?’, Quarterly Journal of Economics 12(3): 373–97. Vogel, Ezra (2011) Deng Xiaoping and the Transformation of China. Cambridge, MA: Harvard University Press. Wang, Jici and Lixia Mei (2009) ‘Trajectories and Prospects of Industrial Districts in China’, in Giacomo Becattini, Marco Bellandi, and Lisa De Propris (eds) The Handbook of Industrial Districts. Cheltenham: Edward Elgar, pp. 598–612. Young, Allyn (1928) ‘Increasing Returns and Economic Progress’, The Economic Journal 38(152): 527–42.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
chapter 5
L oca l Ecosystems a n d Soci a l Con ditions of I n novati v e En ter pr ise Antonio Andreoni and William Lazonick
5.1 Theoretical Approaches to the Firm–Locality Nexus Despite increasing globalization of economic activity, it remains commonplace to associate industrial development, or lack thereof, with particular national models. The adoption of the nation as the unit of analysis may be justified because of distinctive govern mental strategies to foster industrial development. But the success of those strategies ultim ately depends on investments in productive capabilities undertaken by business enterprises that prosper in particular localities within nation states. In some cases, a particular locality becomes a ‘regional hub’ of complementary productive activities that enable the business enterprises in that locality to become world-class competitors, driving the economic devel opment of the nation in which it is located. In the history of modern capitalism three regional hubs of innovative activity stand out as exemplars of localized economic development. They are: a) the Lancashire cotton textile district which in the last half of the nineteenth century enabled Britain to become the ‘workshop of the world’ b) the globally competitive towns and cities specializing in a variety of light industries, especially in the Emilia Romagna regional district, that, as the ‘Third Italy’, brought economic modernity to that nation in the decades after World War II c) the area in California south of San Francisco, centred on Stanford University, that, as ‘Silicon Valley’, made the United States the world leader in the microelectronics and Internet revolutions of the last decades of the twentieth century. In this chapter, we provide a perspective on the social conditions related to business strategy, organizational learning, and investment finance that enabled these three iconic
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
78 Antonio Andreoni and William Lazonick districts to generate industrial products that, in their time, achieved dominant positions on global markets and drove national economic development. Business strategy entails decision-making to invest in productive capabilities that, by transforming technologies and accessing markets, can gain sustained competitive advantage. Organizational learning reflects the social processes through which large numbers of people interacting in a hier archical and functional division of labour acquire the productive capabilities to transform technologies and access markets. Investment finance provides participants in this division of labour with cash and/or credit to sustain organizational-learning processes until they can generate the competitive products that can bring financial returns. The basic unit of analysis for examining business strategy, organizational learning, and investment finance is ‘the firm’, a legal entity that, through its internal governance structure, exercises control over its investment strategy, which includes investment in the productive capabilities of its labour force. An industrial district is made up of a large number of geo graphically co-located and socially embedded firms with related productive capabilities. Given the strategic autonomy of the firm, the question then is how the productive capabil ities and economic performance of the firms located in an industrial district may be affected by virtue of their close proximity to one another. This question points to the existence of a ‘firm–locality nexus’ that can support, or possibly undermine, the productivity of the firm and the local region in which it is based. From a more conventional point of view, the firm–locality nexus has been understood in terms of the advantages of geographical agglomeration or external economies that co-location gives to firms. Productivity and cost advantages may be associated with the presence of physical infrastructure, labour supplies, intermediate inputs, and technology spillovers. This perspective contributes to our understanding of why firms decide to locate in a certain place based on productivity and cost advantages that are only accessed by being based in that locality. From a structural and evolutionary perspective, at the local level, the firm–locality nexus can be understood as the set of institutions and processes linking co-located firms in their development trajectory (Richardson 1972; Andreoni 2014, 2018). This dynamic perspective reveals why the co-location of firms and their evolving social-embeddedness in a particular locality may (i) lead firms to develop a closely complementary even if dissimilar set of pro ductive capabilities, both within and across sectors; (ii) transform the innovative trajectory of one firm in a co-evolving innovative trajectory of a district; and (iii) cause the business strategy of one firm to be influenced by the business strategy of others, both in terms of enabling and constraining them. These dynamic interdependencies are at the core of the local industrial district and can be driven by a plurality of firms operating at different stages of production. In some cases, a plurality of different-sized firms can develop increasing returns out of their evolving division of labour and drive innovation within a multi-hierarchy industry organization. In other cases, large firms operating as lead system integrators drive the increasing returns and innovation dynamics of a district by leveraging their internal economies of scale. Given the different ways in which firms located in an industrial district affect each other’s productive-capabilities development and economic performance over time, different approaches can be adopted in the analysis of the firm–locality nexus. These approaches can be traced back to the seminal theoretical contributions on industrial districts, industry organisation, and business enterprise growth, as well as the set of iconic historical cases of
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Local Ecosystems and Social Conditions of Innovative Enterprise 79 industrial districts which have informed these theories. While these theoretical approaches have been mainly inspired by cases of early industrializers, some of them have also fertilized development thinking and informed the analysis of industrial-district development amongst late and late-late industrializers. One approach to disentangling the firm–locality nexus comes from the work of Alfred Marshall (1842–1924) who centred his Principles of Economics, first published in 1890, on the distinction between internal and external economies of scale (Marshall [1920] 1998; Young 1928; Andreoni and Scazzieri 2014). Making an argument that ‘success breeds success’, Marshall recognized that through superior management and unique improvements in its productive capabilities, a small firm could grow large and possibly dominate its industry by reaping internal economies of scale. In effect, he was arguing that such an innovative firm incurred high fixed costs of investment in unique physical and human capabilities that enabled it to generate a high-quality product that was superior to that of its competitors. Then, by virtue of this higher-quality product, the firm could capture a large extent of the market that enabled the firm to transform its high fixed costs into low unit costs, reaping internal econ omies of scale. Such a firm would contribute to the prosperity of the locality in which it emerged, but its very growth would probably lead it to expand its operations well beyond that locality. Indeed, during the three decades in which Marshall revised Principles of Economics, many multinational corporations emerged, with their home bases in nations such as the United States, United Kingdom, Germany, and Japan (Whitaker 1975; Kerstenetzky 2010). Based on his study of Britain’s late nineteenth-century economic development, including the case of Lancashire cotton textiles, however, Marshall analysed a different type of indus trial dynamics, achieved through external economies of scale. In the growth of the Lancashire cotton textile industry, specific towns became known for a proliferation of firms specializing in the spinning of yarn and weaving of cloth of different types, ranging from coarse to very fine. Each firm had limited internal economies of scale, but benefited from external economies of scale at the town and county levels. At the town level, these economies came from the growth of large supplies of skilled work ers, trained on the job, that enabled the entry of new firms with productive capabilities on a par with already existing firms. As Marshall ([1920] 1998: 350) put it in Principles of Economics: When an industry has thus chosen a locality for itself, it is likely to stay there long: so great are the advantages which people following the same skilled trade get from near neighbour hood to one another. The mysteries of the trade become no mysteries; but are as it were in the air, and children learn many of them unconsciously. Good work is rightly appreciated, inven tions and improvements in machinery, in processes and the general organization of the business have their merits promptly discussed: if one man starts a new idea, it is taken up by others and combined with suggestions of their own; and thus it becomes the source of further new ideas.
In addition, as we shall see, specialized physical infrastructure such as the existence of the cotton market in Liverpool, the yarn and cloth markets in Manchester, and a railroad sys tem connecting these cities with various towns in Lancashire created external economies of scale. As the number of firms in Lancashire increased and the industry grew, the more complete utilization of this infrastructure lowered the unit costs of travel and transactions for all of Lancashire’s constituent manufacturing firms while giving the region an infrastructural
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
80 Antonio Andreoni and William Lazonick advantage that regions in other parts of the world found impossible to emulate (Lazonick 1983; Mass and Lazonick 1990). Over the years, in some regions such as Emilia Romagna, this infrastructural support became even more targeted, leading to the development of voca tional education and training colleges, as well as networks of technology intermediate insti tutions supporting small and medium enterprises absorbing, developing, and scaling up their technological innovations (Andreoni et al. 2017). A second approach to understanding the emergence and evolution of internationally competitive industrial districts is indicated in a sentence with which Marshall concludes his ‘mysteries of the trade become no mysteries’ statement, quoted above. Marshall ([1920] 1998: 350) writes: ‘And presently subsidiary trades grow up in the neighbourhood, supply ing it with implements and materials, organizing its traffic, and in many ways conducing to the economy of its material.’ That is, the district’s growth creates productive supply-chain linkages that, driving its diversification and innovative industrial renewal dynamics, aug ment the district’s competitive advantage. The reappraisal of this Marshallian intuition of an organic social and productive process underpinning the emergence and development of industrial districts was mainly due to two Italian scholars, Giacomo Becattini and Sebastiano Brusco, in the 1970s. The seminal idea of a different model of production—i.e. the industrial district—started emerging in the 1970s and acquired momentum in the 1980s, driven by the fact that a number of Italian regions had stood out internationally, showing increasing com petitiveness in a plurality of industries, ranging from traditional light industry areas focus ing on textiles like Prato (Tuscany) to medium-high-tech ones like ceramics in Sassuolo and machinery in Bologna. Later in the 1990s, the industrial district idea partially merged with a germane—though different—literature on ‘clusters’ introduced by Michael Porter and several other contributions applying the ‘Third Italy’ model to the analysis of other regions around the world presenting similar features and dynamics (Sabel 1982; Best 1990; Porter 1990; Salais and Storper 1992; Lazonick 1993; Camagni 1994).1 Becattini with a focus on the Tuscany region and Brusco with a focus on the Emilia Romagna region set out to disentangle the complex set of social conditions underpinning the successful performance of several industrial districts (Becattini 1978; Brusco 1975, 1982). At the core of this innovative model of production there was, in the words of Becattini, a ‘complex and tangled web of external economies and diseconomies, of joint and associated costs, of historical and cultural vestiges, which envelops both interfirm and inter-personal relationships [and tends] towards the multi-sectorial’ (Becattini 1989: 132). The social con ditions of innovative enterprise were, thus, to be understood not simply at the firm level, but also looking at the system of multiple interdependencies linking firms embedded in a given industrial district. Within this framework, the size of the firm was reassessed in light of the economies of scale that the industrial district was able to generate as an organized unit of production. Moreover, in contrast with the emphasis on internal economies of scale, and given the dominant presence of small and medium enterprises (SMEs) in the Italian dis tricts, early contributions tended to emphasize how, by becoming specialized producers and contractors while retaining flexibility and resilience, even small firms can operate in a highly productive manner and be indeed innovative. The way in which the socio-institutional context and regional industrial policy conferred competitive advantage on firms in these districts was also highlighted, especially in the case of Emilia Romagna. 1 See also Chapters 4 and 10.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Local Ecosystems and Social Conditions of Innovative Enterprise 81 Over the years, however, several contributions have reassessed the performance of these SMEs in light of changes in the composition of firms in the industrial districts, and their relationships with large companies. In fact, as we will see in the analysis of the Emilia Romagna model, at the core of the industrial district success there was the emergence of a symbiotic relationship of complementarity between large firms and their supply chains. In fact, on the one hand, the emergence of SMEs with highly specialized productive capabil ities attracted national and international large firms to locate operations in these districts in the 1970s and 1980s; on the other hand, these large firms, deliberately spinning out workers to create specialized contractors, nurtured their local supply chains over the 1980s and 1990s. The relationship between large firms—especially international ones—and local SMEs is critical for understanding today’s dynamics of agglomeration across late and late-late indus trializing countries and regions. Albert Hirschman (1916–2012) was the first to point out the way in which forward and backward linkages along a vertical ‘value chain’ (to use a postHirschmanian term) as well as along horizontal value chains linking companies in a regional hub of activities can drive the process of economic development (Hirschman 1958, 1977). Increasing domestic value-added in production depends critically on the extent to which the attraction of an international company in a developing country is linked to the emer gence of a local chain of suppliers which can over time integrate both vertically to access international buyers and horizontally to access the supply of intermediate products in the regional hub and final domestic products. A third approach to understanding the emergence of an industrial district builds on Marshall’s concept of internal economies of scale. The growth of one or more firms to be very large can drive the district’s growth. In what by 1971 became known as ‘Silicon Valley’, that role was originally played by an ‘Old Economy’ company, Hewlett-Packard, founded in 1939, but what made the district the global centre of the microeconomics revolution was the launching of a number of ‘New Economy’ companies, including Intel (founded in 1968), Advanced Micro Devices (1969), Oracle (1977), Apple (1977), Sun Microsystems (1982), and Cisco Systems (1984) that ultimately grew to employ tens of thousands. Key insights into the social conditions that enabled the growth of these firms can be found in the work of Edith Penrose (1914–96). In her 1959 book, The Theory of the Growth of the Firm, Penrose conceptualizes the mod ern corporate enterprise as an organization that administers a collection of human and physical resources (Penrose 1959; see also Chandler 1962). People contribute labour services to the firm not merely as individuals, but as members of teams who engage in learning about how to make best use of the firm’s productive capabilities—including their own. This learning is organizational; it cannot be done all alone, and hence is collective, and it cannot be done all at once, and hence is cumulative (Best 1990: 125). At any point in time, this col lective and cumulative learning endows the firm with experience that gives it productive opportunities unavailable to other firms, even in the same industry, that have not accumu lated the same experience. The accumulation of innovative experience enables the firm to overcome the ‘managerial limit’ that in the neoclassical theory of the optimizing firm causes the onset of increasing costs and constrains the growth of the firm. The innovating firm can transfer and reshape its existing productive capabilities to take advantage of new market opportunities. Each move into a new product market enables the firm to utilize productive capabilities, and especially human capabilities, that had been accumulated through the pro cesses of organizational learning in generating its previous, now mature, products. These
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
82 Antonio Andreoni and William Lazonick unused productive capabilities, along with some of the profits that they previously generated, can provide foundations for the further growth of the firm, accompanied by in-house complementary investments in new product development or the acquisition of other firms that have already developed complementary productive capabilities. Of critical importance in the approaches of seminal economists such as Marshall, Becattini, Brusco, Hirschman, and Penrose is that they integrated theory and history, deriv ing theoretical logic from historical fact, and then used their theoretical constructs as ana lytical tools to explore historical evolution brought up to the present. As a result, they permit us as researchers of the dynamics of industrial development to analyse in a coherent manner the operation and performance of industrial hubs in capitalist development, includ ing the three iconic districts on which we focus in the body of this chapter. We will close by articulating key insights from these analyses for understanding the evolving role of indus trial hubs in economic development more generally.
5.2 Lancashire Cotton and the ‘Workshop of the World’ The cotton textile industry, producing both yarn and cloth, was at the centre of the world’s first Industrial Revolution of the late eighteenth and early nineteenth centuries.2 The local ization of the British cotton textile industry in the county of Lancashire was reinforced as its port city Liverpool received shipments of cotton from around the world, and particularly from the United States, India, and Egypt. The first building that housed the Liverpool Cotton Exchange opened in 1808. The growth of the Lancashire industry induced yarn and cloth merchants from around the world to open offices in Manchester, Lancashire’s largest city, where they could place orders in person with manufacturers of yarn and cloth for export to the countries from which they came. The first modern building for the exchange of yarn and cloth opened in Manchester in 1809, and underwent a major expansion in 1849, taking the name the Manchester Royal Exchange after a visit by Queen Victoria in 1851. The existence of the Manchester exchange gave small textile manufacturing companies access to orders for their products on a weekly basis, and the Liverpool exchange enabled them to purchase cotton of the desired staple and grade needed to fill those orders. Business travel to these cities was greatly facilitated by the opening of the Liverpool and Manchester Railway in 1830, with extension of branches of the railway system to the numerous towns in Lancashire in which the spinning of yarn and the weaving of cloth were carried out. Thus, by mid-nineteenth century, the fundamental infrastructure for the operation of the Lancashire industrial district was solidly in place. Until the widespread diffusion of the steam-driven power loom in the 1840s, spinning factories, centred in southeast Lancashire, sent yarn to rural households, generally located in northwest Lancashire, to be woven into cloth on hand looms in what was known as the ‘putting-out system’. At the same time, much of the factory-produced yarn was sold directly to buyers on domestic and global markets. With the advent of the power loom, prior to the 2 This section draws on Farnie (1979), Lazonick (1979, 1981a, 1981b, 1983), Mass and Lazonick (1990).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Local Ecosystems and Social Conditions of Innovative Enterprise 83 mid-nineteenth century, the tendency was for an existing spinning firm to add power-loom weaving as a vertically integrated operation. But the availability of a yarn supply on the Manchester exchange and a weaving labour supply formerly engaged in the putting-out system resulted in the proliferation of vertically specialized weaving mills in northwest Lancashire along with the increasing specialization of textile factories in southeast Lancashire in spinning only. As a proportion of employees and capacity, the integrated firms declined in relative importance over the last half of the nineteenth and the first half of the twentieth century. As the Lancashire industry became more vertically specialized, it also became more hori zontally fragmented because the district-wide infrastructure that supported well-developed input and output markets reduced the amount of financial resources and the range of man agerial knowledge required to participate in the industry. In addition, the growth of the manufacturing industry in the nineteenth century gave rise to the world’s leading textilemachinery industry in Lancashire, with eight major companies producing full ranges of spinning and weaving equipment. Indeed, in periodic booms, the machinery companies were active in financing the construction of new spinning mills, which entailed far more complex and expensive manufacturing processes than weaving. In contrast, in expansions, under what was known as the ‘room and power’ system, many new weaving firms were set up by renting space in a larger weaving mill and purchasing access to the building’s steam power. As Alfred Marshall recognized with his statement ‘the mysteries of the trade become no mysteries; but are as it were in the air, and children learn many of them unconsciously’, the growth of an industrial district such as Lancashire benefited from transfer of skills related to specific industrial processes on-the-job from older, experienced workers to younger, inexperienced workers. In the cotton spinning industry, the most important machine tech nology from the 1830s was the ‘self-acting mule’, on which senior mule spinners known as ‘minders’, and invariably male, were paid piece rates, out of which they paid time rates to junior workers known as ‘piecers’ whom they hired and who through on-the-job experi ence could acquire the skills to eventually, if male, become minders. By the last half of the nineteenth century, this training system produced an oversupply of potential minders, even in boom periods. The minders joined unions to safeguard their employment conditions and ensure that they received a fair share of productivity gains. Their collective bargains over the relation between work and pay were printed in complex wage lists, which by the last decades of the nineteenth century had acquired virtually the force of law. These collective agreements encouraged minders to cooperate in getting high levels of productivity out of the mule-spinning machines, which, through effort-saving technical improvements, enabled an increase over the last half of the nineteenth century from about 300 spindles per minder to as many as 2,800 spindles per minder, with the same complement of piecers (Lazonick 1990). In the process, the minders were able to share in the productivity gains while enabling the Lancashire cotton textile industry to remain competitive on world markets. In the two dec ades before World War I, however, the limits of sharing the gains from effort-saving techno logical change were reached in both mule-spinning and power-loom weaving. Many Lancashire employers, when buying cotton in Liverpool, sought to cut costs by purchasing cotton of somewhat shorter staple and somewhat lower grade. In the manufacturing pro cesses, workers experienced this inferior cotton input as ‘bad spinning’ or ‘bad weaving’
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
84 Antonio Andreoni and William Lazonick because it compelled them to expend more effort in the labour process to generate a given level of productivity and hence earn a given weekly wage. Especially in the coarse spinning industry concentrated in the town of Oldham and the coarse weaving industry in the town of Blackburn—which were facing increasing low-wage competition from India and Japan— the practice of using inferior cotton to cut costs became increasingly widespread in the years before World War I, as manifested by recurrent industrial actions by the cotton-textile unions against cost-cutting employers. Lancashire’s vertically specialized and horizontally fragmented industry structure and the district’s ready availability of skilled labour led Alfred Marshall to observe in his 1919 book Industry and Trade that the Lancashire cotton textile industry was ‘perhaps the best present instance of concentrated organisation mainly automatic’ (Marshall 1920: 600–1). Although the Lancashire industry had experienced rapid growth in the decade prior to World War I, by the time Marshall published these words the industry was entering a period of long-run decline. The vertically specialized and horizontally fragmented structure of yarn and cloth manufacturing in Lancashire posed formidable barriers to the district’s adoption of advances in high-throughput machine technologies—ring spinning and auto matic weaving looms—emanating mainly from the United States, but also from Japan. From 1924 when the Toyoda automatic loom became available, Japan in particular began to outcompete Lancashire on foreign markets. The loss of Lancashire’s competitiveness in the 1920s did not result in the sudden dis appearance of spinning factories using self-acting mules and weaving mills using power looms. Rather, these firms continued to operate using the old technologies, which had been built to last, with cuts in wages and profits to remain competitive on global markets. As John Maynard Keynes expressed the problem in 1928: There [is] probably no hall in Manchester large enough to hold all the directors of the cotton companies; they [run] into thousands. One of the first things should be to dismiss the vast majority of these people, but the persons to whom this proposal would have to be made would be precisely those directors. (Lazonick 1983: 21)
Indeed, in the late 1920s, through the creation of the Lancashire Cotton Corporation (LCC), the Bank of England sought to gain strategic control over as many as 200 cotton spinning firms and 20 million spindles to re-equip the most efficient factories and force the ineffi cient ones to shut down. John Ryan, the Bank-appointed director of LCC argued: I do not think that we can leave it to the individuals. Vested interests will always resist a movement which tends to remove them. And vested interests can easily hold a position which is acceptable to themselves but yet is acting as a parasite on an industry and slowly driving it to death. (Lazonick 1983: 215–16)
In the event, by 1930 ninety-six spinning firms with 9.3 million spindles (one-fifth of Lancashire’s total) had been coerced, by threat of termination of credit, to join LCC, but in 1932 its senior executives resigned when the board of directors decided that mill managers should be given more autonomy in decision-making. Unable to rationalize its capacity and modernize its factories, the Lancashire cotton textile industry, which had once made Britain
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Local Ecosystems and Social Conditions of Innovative Enterprise 85 the ‘workshop of the world’, became integral to that nation’s relative economic decline (Elbaum and Lazonick 1986).
5.3 The ‘Third Italy’ and the Emilia Romagna Industrial Ecosystem At the outset of the 1950s, Italian industry was characterized by a dualistic structure, centred around large firms (many of which were built through industrial policy as part of the Instituto per la Ricostruzione Industriale—IRI) and very small productive units (less than eleven employees), which together accounted for more than 57 per cent of the total (Bianchi 2013; Brusco and Paba 2014). Throughout the 1950s and 1960s, however, the employment share of the very small firms decreased dramatically, going from one-third in 1951 to one-fifth in 1971. With the increasing divide between the North and South of Italy, a number of very small firms in the South disappeared while in the North a significant por tion of small firms started increasing their scale and operating in the international market. In the following two decades, the number of small firms (with less than fifty employees) witnessed significant growth, going from 42 per cent of employment in manufacturing in 1971 to reach almost 60 per cent in 1991. A significant number of these newly graduated SMEs were located in industrial districts in the Centre and North-East regions of the coun try, including regions like Tuscany and Emilia Romagna. They came to be called the ‘Third Italy’, in contrast with the North-West, whose industrial take-off started in the early twen tieth century and was led by larger firms like Fiat and Olivetti, and the backward South (Brusco and Paba 2014).3 Throughout these four decades, the number of industrial districts mapped out by ISTAT went from 149 districts employing a total of 360,000 workers, to 238 districts employing 1.7 million workers. Twenty years later, in 2011, the last industrial census identified 611 Labour Market Areas, of which 141 qualified as industrial districts. The latter absorb 24.5 per cent of total Italian employment and 37.9 per cent of manufacturing employment, with 27 per cent of these industrial districts specializing in mechanical industries, another 23 per cent in textiles and clothing, and 17 per cent in household goods (ISTAT 2015). Although these aggregate statistics obscure the increasing heterogeneity and complex sectoral and location boundaries of industrial districts, they point to an exceptional industrial reality which has driven the transformation of some of the most industrialized regions in the world. Industrial districts came to be associated with the Emilia Romagna region in particular, because of the work of Sebastiano Brusco and his formulation of the ‘Emilian model’, but also because industrial districts transformed this region more than others, as revealed by several parameters. With a population of 4.4 million and 425,000 enterprises (of which 50,000 are manufacturing firms), Emilia Romagna in 2017 reached a regional GDP of 150 billion euro and value in export of 60 billion euro. It is also the top region in Italy for number of patents registered in the European Patent Office. Advanced mechanical engineering is 3 See also Chapter 10.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
86 Antonio Andreoni and William Lazonick the key sector of the regional economy, with more than 28,000 companies exporting around two-thirds of their output and making up a similar amount of the regional export turnover. Distributed across several industrial districts, these mechanical engineering companies operate in several sectors such as food processing, agricultural machinery, automotive, medical device, packaging machinery, automation components and digital solutions, con struction, and ceramic tiles (Emilia Romagna Region Website 2019). At the roots of the ‘Emilian model’ of an industrial district, which transformed the regional economy from an agricultural into an advanced industrial one in a few decades, Brusco identified a number of social conditions (Brusco 1982). First he recognized how SMEs that often co-located and concentrated in a bounded and monocultured locality had managed to establish collaborative relationships which had led to low degrees of vertical integration of production. Within these districts, only a limited number of SMEs were directly focused on final goods, while the great majority operated as part of chains of sub contractors involving more or less specialized firms. Thanks to co-development of closely complementary and dissimilar sets of productive capabilities, firms were able to both specialize and diversify, while learning to export. Moreover, the symbiotic relationships developing within industrial districts facilitated the graduation of several small companies to medium status. Between 1981 and 1996, the number of very small firms diminished by nearly one-third and their workforce by nearly one-fifth, while small and medium firms increased in both absolute and relative terms in the region. Second, Brusco pointed to the social embeddedness of these economic relationships amongst co-located firms which equipped each firm and the industrial district as a whole with significant readiness to change, flexibility, and resilience. For example, the successful management of labour relationships and other conflicting claims was made possible by social coalitions of interests, supported by political mediation and provision of ‘real ser vices’, more than distribution of financial incentives. Public investments in technical schools and training institutes, technology intermediate institutes focusing on SMEs, universities focusing on engineering and industrial research, and infrastructures were distinctive fea tures of the Emilian model (Andreoni 2016; Andreoni et al. 2017). In fact, a certain type of politics was both the outcome and the driver of the proliferation of entrepreneurs’ associations, unions, regional and cooperative banks, and the same SMEs. The Italian Communist Party (PCI) in Emilia Romagna provided real services to support the creation of small firms. These services included business development; organizational capabilities, management and systems development; technological scaling up, and markets development. The local government also promoted the development of several intra-firm associations, which helped coordinate SMEs efforts in productive capabilities development. Starting from the 1980s and increasingly throughout the 1990s and 2000s, however, the social conditions of the Emilian model evolved, and the region with its industrial districts witnessed significant processes of industrial restructuring and innovative industrial renewal (Becattini et al. 2009; Dei Ottati 2018; Andreoni 2018). We can identify three main pro cesses of transformation which reshaped the social conditions of innovative enterprise in the regional industrial districts. These are internationalization, consolidation, and h ierarchic restructuring. The internationalization of industrial districts was led by an increasing export orienta tion of SMEs, but also the attraction of international companies to the region. For example, in the medical device industry leading international companies like Baxter, Gambro, Fresenius,
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Local Ecosystems and Social Conditions of Innovative Enterprise 87 and B. Braun were attracted to the region by the presence of a chain of subcontractors with a unique mix of productive capabilities in co-injection moulding and production of medical-device disposables, as well as in fluidic systems design and manufacturing (e.g. pumps, valves, automation solutions). The appearance of these international companies transformed—and in some cases disrupted—some of the social relationships in which the industrial districts were traditionally embedded. In some cases, they introduced new ‘glocal’ linkages which redesigned the geographical boundaries of the industrial districts. For example, by in-sourcing intermediate goods from other regions, they led to the extinction of some local firms and the transformation of the local chains of supply into a global value chain. In other cases, international companies encouraged the spin-off of activities to several specialist contractors which initially worked exclusively for their home company, but later for some of their competitors as well. In some cases, these specialist contractors even managed to develop productive capabilities at the interface of different sectors such as medical device and automotive and became major drivers of technological pollination. Such is the case in the medical device industrial district of Mirandola where forms of indirect cooperation mediated by specialist contractors were found (Andreoni and O’Sullivan 2014). While international companies played an important role in transforming the Emilian model, industrial consolidation was critically driven by local firms through mergers, take overs, and the setting-up of new business groups controlling several industrial districts’ activities within a de facto verticalized model of production (Russo and Whitford 2009). The most successful local firms set up new subsidiaries and embarked on a number of take overs of both local and international companies. This was partially driven by the need to acquire productive capabilities which were not manageable in the traditional contracting relationship within the industrial district, or which were not available locally. In the packaging machinery industry (Andreoni et al. 2017), for example, the leading company in the food segment—IMA—started preparing and supporting its technology transition towards the pharma segment first by increasing its production capacity in packing, blistering, and dosing technologies. Later, IMA acquired a number of companies mainly located within the region (CMS, Zanasi, Farmatic, Farmomac, PM System, and Cestind Centro Studi Industriali) and in 1990 merged these companies within one unique organizational d ivision. In the packaging machine industry, by the end of the 1990s the four biggest groups—IMA, GD, SACMI, and Marchesini—accounted for 43 per cent of the district’s output. Between 1995 and 2005, IMA went through a new phase of M&A, involving both regional and inter national companies specialized in washing and sterilization (Libra), capsules (Kilian, DE, and GS Coating) and packaging (Swiftpack, U.S.). The internationalization and consolidation across several sectors in the region also led to an increasing hierarchic restructuring of the industrial district network and its transforma tion into an ‘industrial ecosystem’ of firms organized along several sectoral value chains (Andreoni 2018). The horizontal cooperative structure of small firms at the centre of Brusco’s Emilian model was replaced by a multi-tiered structure of suppliers organized around system integrator firms, both national like IMA in packaging machinery or inter national like Baxter in medical devices. System integrators play a key orchestrating role in the ecosystem and represent a major knowledge node in terms of technology and markets. Despite the new centrality acquired by these larger firms, across the region, a significant number of specialist contractors along the local chains of suppliers retain significant productive and innovative capabilities.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
88 Antonio Andreoni and William Lazonick However, their development has become much more linked to the business strategy of the lead system-integrating firms. This hierarchic restructuring has thus introduced more dif ferentiation across traditional industrial districts, both in terms of organizational structure and models of production, but also in terms of competitive and innovation dynamics and strategic coordination of productive-capability development efforts. These development efforts have resulted in five distinctive capability domains character izing the Emilia Romagna industrial ecosystem: (i) bio, food, and agro-technologies; (ii) advanced materials; (iii) mechanical systems and automation; (iv) ICT and embedded systems; and (v) biopharma and medical technologies. These capability domains cut across traditionally defined sectors and go beyond the production and technology bases of individual companies and traditionally defined industrial districts. From a historical per spective, ‘mechanical systems and automation’ (in particular machinery and mechanical components) and ‘advanced materials’ (in particular plastics and ceramics) are the most distinctive capability domains of the Emilian industrial ecosystem, alongside ‘bio, food, and agro-technologies’. These capability domains have found applications in both advanced manufacturing and more traditional sectoral value chains, including food and agro-processing. The development of the two capability domains pertaining to ‘ICT and embedded systems’ and ‘biopharma and medical technologies’ emerged in the late 1980s. The development of these latter capability domains have had pervasive effects across all sectoral value chains, in particular by establishing close complementarities with the ‘mechanical systems and auto mation’ capability domain (Andreoni 2018).
5.4 ‘Silicon Valley’ as the Epicentre of a Global Technological Revolution During the 1960s, there were so many semiconductor chip start-ups in the vicinity of Palo Alto, California that in 1971 a local journalist dubbed the area ‘Silicon Valley’.4 As it turned out, the most important of those start-ups was Intel, founded in 1968 by Gordon Moore and Robert Noyce. With recent PhDs—Moore in chemistry from CalTech and Noyce in physics from MIT—in 1956 the two men had come to Palo Alto to work in a business-sector research lab run by William Shockley, co-inventor of the transistor at Bell Labs. Shockley proved to be a difficult boss, and in 1957 Moore, Noyce, and six other scientists and engineers—whom Shockley called ‘the traitorous eight’—left his lab but remained in the Palo Alto area to form Fairchild Semiconductor. In 1959 Noyce invented the integrated circuit, and in 1965 Moore articulated the famous ‘law’ bearing his name which predicted that, through research and development, the computing power of an integrated circuit would double every eighteen months. During the first decade of its existence, Fairchild led in implementing Moore’s Law, developing and producing ‘memory chips’ almost entirely for military purposes. 4 This section draws on Kenney (2000), Berlin (2001), Lécuyer (2006), Lazonick (2009). See also Chapter 12.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Local Ecosystems and Social Conditions of Innovative Enterprise 89 Many scientists and engineers left Fairchild during the 1960s to form start-ups nearby. By the time Moore and Noyce left Fairchild to found Intel in 1968, the cost of a memory chip had declined so much that it became possible to use this revolutionary technology for com mercial purposes, and Intel was founded for this purpose. In 1971, an Intel engineer invented the microprocessor, or a ‘computer on a chip’, and in 1974 the company opened its first microprocessor fabrication facility in Portland, Oregon (a location 700 miles north of Silicon Valley, with from the late 1980s the largest concentration of Intel employees world wide). Memory chips remained Intel’s most important commercial product until, by the mid-1980s, Japanese semiconductor companies took over that product segment while Intel’s microprocessor sales took off with the enormous success of the IBM PC and its clones. Intel became the world’s leading semiconductor manufacturer in 1991, passing Motorola and Texas Instruments, both long-established ‘Old Economy’ companies, located outside Silicon Valley, that had themselves been pioneers in the microelectronics revolution. At a 1993 conference at Harvard Business School on the decline of Old Economy corpor ate research labs, Moore, then Intel chairman, clearly articulated the relation between research done in Old Economy companies such as Motorola and Texas Instruments and product development done in New Economy start-ups such as Intel: Running with the ideas that big companies can only lope along with has come to be the acknowledged role of the spin-off, or start-up. Note, however, that it is important to distinguish here between exploitation and creation. It is often said that start-ups are better at creating new things. They are not; they are better at exploiting them. Successful start-ups almost always begin with an idea that has ripened in the research organization of a large company. Lose the large companies, or research organizations of large companies, and start-ups disappear. (Moore 1996: 171)
One such established company located at the heart of what became Silicon Valley was Hewlett-Packard (HP), founded in Palo Alto in 1939 by two Stanford engineering graduates whose professor had been Frederick Terman, Silicon Valley’s visionary. Characteristic of the ‘Old Economy business model’ (OEBM), HP promised its employees a career-with-onecompany (CWOC), with a company-funded defined-benefit pension in retirement. In a best-selling book, The HP Way, in 1995 co-founder David Packard explained the import ance of CWOC for the company’s decades of innovation that built on its collective and cumulative learning in electronics devices (Packard 1995). Although it is doubtful that Packard had read Penrose’s The Theory of the Growth of the Firm, HP’s history exemplified her thesis of enterprise expansion by deploying the productive capabilities embodied in unused labour services to diversify into new related lines of business. Nevertheless, HP became integrally involved in the computer revolution in a way that fundamentally changed how the company employed people, as it adopted the ‘New Economy business model’ (NEBM) that had come to characterize the more recent Silicon Valley companies that, like Intel, had started up in the 1960s and later. In its 1984 Annual Report, HP announced that, in the new era of the personal computer, it would henceforth develop its printers using an open systems architecture. Under the proprietary systems architecture characteristic of OEBM, a company like HP valued engineers and managers with decades of experience who could serve as ‘system integrators’. Under the open systems architecture characteristic of NEBM, a company valued younger workers with the latest computer-related skills who could add value to an industry-wide platform.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
90 Antonio Andreoni and William Lazonick As during the 1990s printers dominated HP’s revenues, there was pressure on the com pany to jettison the HP Way. During the first half of the 1990s, it sold its manufacturing operations to contract manufacturers. In 1999 it spun off the original HP electronics busi nesses as Agilent, and further cemented its commitment to NEBM when it acquired PC-maker Compaq in 2002—bringing HP in line with the business model that, over the course of the 1980s and 1990s, had come to dominate in Silicon Valley. By the first half of the 2000s, HP had become known as a ‘hire-and fire’ company as part of its full-fledged commitment to ‘maximizing shareholder value’ by doing massive cash distributions to shareholders in the form of dividends and stock buybacks. As a proportion of profits, HP paid out 15 per cent as dividends and 41 per cent as stock buybacks in 1986–95; 30 per cent and 92 per cent in 1996–2005; and 20 per cent and 134 per cent in 2006–15. (In November 2015, HP spun off its software and services businesses as Hewlett Packard Enterprise.) A key characteristic of NEBM is a vastly enlarged importance of the stock market in the operation of the company. For an issuing company, there are five possible functions of the stock market: control, cash, creation, compensation, and combination. Under OEBM, the main function of a stock market listing on the New York Stock Exchange (NYSE) was to separate share ownership from managerial control over corporate resource allocation, thus permit ting professional managers to run companies. It is a myth that the main function of the stock market was to raise cash for corporate investment; for a company to be listed on NYSE it had to have a track record of profitability and substantial market capitalization. That was not the case with the NASDAQ electronic exchange, launched in 1971. Its main function was new firm creation: by enabling a young company to do an initial public offering (IPO) within three to five years from its founding, NASDAQ induces private equity to invest in start-ups by enabling this ‘venture capital’ to do a quick ‘exit’ from its investment in the start-up. Silicon Valley companies such as Intel, Apple, Oracle, Cisco, and Google (now called Alphabet) are all listed on NASDAQ. And with NASDAQ in place, Silicon Valley quickly became by far the world’s most important industrial district for venture-backed start-ups, with a focus on the information-and-communication (ICT) and biotech industries. Besides inducing venture-capital investment in start-ups, the stock market also has a compensation function under NEBM. Start-ups used stock options and stock awards to induce professional, technical, and administrative personnel to leave secure employment at Old Economy companies such as HP, IBM, Motorola, and Texas Instruments to take up inherently insecure employment at a new venture. If and when the start-up could do an IPO on NASDAQ or be acquired by an already existing publicly listed corporation, the individual’s stock-based pay could be worth millions of dollars. As Silicon Valley companies such as those mentioned above grew to employ tens of thousands of people, stock-based pay con tinued to provide a substantial proportion of total compensation, not just for senior execu tives but also for a broad base of employees extending far down the corporate hierarchy. In addition to the creation and compensation functions of the stock market under NEBM, it has also had a combination function as companies have used their shares to acquire other companies for the sake of rapid growth. As Cisco Systems grew from $70 million in revenues and 254 employees in 1990 when it did its IPO to dominate the enter prise network equipment market a decade later with $18.9 billion in revenues and 34,000 employees, it used its stock to pay for sixty acquisitions valued at $32.5 billion, of which 98 per cent was paid in Cisco shares. During this period, Cisco became known for its success in integrating personnel from these acquisitions into its learning organization (Carpenter et al. 2003).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Local Ecosystems and Social Conditions of Innovative Enterprise 91 The most successful of the Silicon Valley companies have raised some cash for their treasuries at the IPO. But these amounts have been miniscule when compared with the funds reinvested from profits, as these companies have eschewed paying dividends and doing buybacks. For example, the only time in its history that Apple secured cash by selling shares on NASDAQ was $97 million in its IPO in 1980. Reflecting the rise of the ideology in the United States from the 1980s that a company should be run to ‘maximize shareholder value’, however, from 1985 to 1997, Apple almost drove itself into bankruptcy by distributing cash to shareholders when it should have been investing in productive capabilities. After Apple founder Steve Jobs, who had been pushed out of the company in 1985, came back to lead the company in 1997, he reinvested its profits to develop iTunes, iPod, iPhone, and iPad, making the company one of the most successful in history when he passed away in 2011 at the age of 56 (Lazonick et al. 2013). Since Jobs’ death, Apple has turned from innovation to financialization, as since 2012 through the second quarter of 2019, the company distributed $271 billion in buybacks (87 per cent of profits) and $79 billion in dividends (another 25 per cent of profits). In the process, Apple has ceased to be an innovative company. In the name of ‘maximizing share holder value’, Apple labels this massive distribution of corporate cash its ‘Capital Return Program’. But beyond $97 million at its 1980 IPO, public shareholders have not invested in Apple’s productive capabilities. So how can the company ‘return’ this cash to them? (Lazonick et al. 2016; Lazonick 2019a). Similarly, Cisco Systems, which epitomized a Silicon Valley high-growth company in the 1990s, has since 2001 turned from innovation to financialization. From 2002 through 2019, Cisco spent $139 billion on buybacks (113 per cent of profits) and $36 billion on dividends (29 per cent of profits). By reinvesting its profits in sophisticated communication-technology products, Cisco could have built on its leading position in enterprise networking to become a dominant player in communication infrastructure. In short, Cisco could have been Huawei Technologies, based in Shenzhen, China, the current world leader in communica tion infrastructure technology, including 5G. Founded in 1987, three years after Cisco, Huawei is not tempted to give its profits away to public shareholders who merely buy and sell outstanding shares because the company is not listed on a stock market. Rather, it is 100 per cent employee-owned (Carpenter and Lazonick 2019). As we discuss in the conclusion of this chapter, the transformation of Silicon Valley from innovation to financialization, as Cisco has done, provides critical insights for building innovative industrial districts in the developing world.
5.5 General Lessons about Economic Development from the Iconic Industrial Districts The analysis of three of the most iconic industrial districts through the integration of his tory and theory has revealed a number of important lessons for countries who look at these experiences as models for their development and policies. Learning from the history of these industrial districts is not about trying to artificially reproduce their experiences, as
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
92 Antonio Andreoni and William Lazonick their context-specific social preconditions and conditions are not reproducible per se. Learning from historical successes (and failures) is more about ways of expanding and pre paring the imagination of firms and policymakers about the dynamics, opportunities, and challenges associated with the development of a regional hub. It is also about realizing how regional hubs are centred around firms, to be understood as social organizations whose business strategies, organizational structures, and committed financial resources contribute to their own success as well as the social, economic, and environmental sustainability of their locality. The linkage perspective originally advanced by Albert Hirschman in development economics remains critical as it was in the experience of early and late industrializers. Developing and middle-income countries are primarily competing with specific regional hubs—more than countries as a whole—located across the old and newly industrialized nations. The most difficult challenge they face is to develop new internationally competitive firms and weave around them a web of production, consumption, technological, and fiscal linkages whereby value is created, retained, and reinvested. In these countries these differ ent types of linkages are mediated by complex political economy dynamics and challenges, which affect the creation, retention, and distribution of value in the still underdeveloped localities. The development of regional hubs is about the transformation of social institu tions by political processes in order to achieve technological change and superior economic performance (Andreoni 2019; Andreoni and Chang 2019). In what follows, to enlighten this challenge, we extract a number of lessons about how social conditions for innovative enterprise were built in each of the three industrial districts reviewed in this chapter.
5.5.1 Lessons from Lancashire Marshall favoured small firms on moral grounds, and from the experience of the Lancashire cotton textile industry, he could argue that a proliferation of small firms and economic development went hand in hand. He understood, moreover, that the competitive advantage of Lancashire was based on the accumulation of skilled labour through on-the-job experi ence, working with advanced machinery. Local finance, including investments by Lancashire’s machinery firms, permitted the rapid construction of new spinning mills in response to growing demand. Indeed, the experience of Lancashire cotton textiles as the central industry of the world’s leading economy at the end of the nineteenth century lent credence to Marshall’s notion that one could analyse an industry in terms of ‘the representative firm’, and in the interwar period, post-Marshallian, i.e. neoclassical, economists translated this view of the world into the notion that a state of affairs known as ‘perfect competition’ represented the ideal of eco nomic efficiency. In the process, neoclassical economists ignored Marshall’s insights into the potential for internal economies of scale to outcompete an industrial district based on external economies of scale. That is precisely what happened to the Lancashire cotton textile industry when chal lenged by the development of higher productivity machine technologies—ring spinning and automatic loom weaving—in the United States and Japan. Especially in Japan, the suc cessful adaptation of these technologies in a highly concentrated cotton–textile industry
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Local Ecosystems and Social Conditions of Innovative Enterprise 93 enabled that nation to become a global competitor in not only yarn and cloth but also textile machinery in the interwar period (Mass and Lazonick 1990). More than that, research has shown that the productive capabilities that the Japanese acquired in textile machinery, first and foremost at Toyoda Automatic Loom Company, provided the foundation for the Japanese to emerge as world leaders in the automobile industry from the 1970s—a position still held by Toyota Motor Corporation, with production of 10.6 million vehicles and $272 billion in revenues in fiscal 2019 (Mass and Robertson 1996). The company has created a formidable industrial district around Toyota City (next to Nagoya, Japan) as the centre of one of the most powerful global companies in the world.
5.5.2 Lessons from the Third Italy and Emilia Romagna The Third Italy and Emilian model of industrial district point to a number of lessons around the evolving dynamics and relationship between firms and their local ecosystem, as well as the important role that regional industrial policies can play in shaping these developments. One of the distinctive features of the Third Italy and Emilia Romagna experiences is that industrial districts emerged organically out of specific social and institutional conditions which sedimented over several years. At the same time, the industrial district model of production was also encouraged and shaped by targeted regional policies and policy medi ation aimed at retaining and reinvesting the value generated in the regional hubs and lever aging multiple set of linkages. From a policy perspective this implies that although these organic development processes are difficult to reproduce out of context and not all regions might have the same ‘social preconditions’ met, all regions—even agricultural ones like Emilia Romagna—have a chance to build them. This will depend on the extent to which productive organizations are attracted, supported, and increasingly linked to the locality— its firms and interests, and the extent to which, over time, this process becomes organic to the economic dynamics of the region. Embarking on an organic process of industrial transformation also means realizing how the development of regional hubs follow a tortuous—far more than linear—trajectory. From a policy point of view this means that triggering the emergence of an industrial dis trict is in fact a continuous process of innovative industrial renewal. Industrial districts are born, and reborn, many times in their life. Emilia Romagna has reinvented its model of production—from a horizontal towards a multi-layered system—in response to the new needs and opportunities offered by internationalization and consolidation. Within this new ecosystem model the provision of real services is critical, as SMEs are disadvantaged and vulnerable to dramatic technological shifts, financial crisis and credit contraction, or the challenges of operating in new markets. Real services tend also to be more effective than financial incentives as they select companies and establish opportunities for building up the collective productive capabilities of the locality. The evolution of the Emilian model into an ecosystem of heterogeneous firms integrated within and across several sectoral value chains also points to the importance of moving beyond the dichotomic policy discussion which seems to suggest that policies should either support small firms or big ones. In the development of an industrial ecosystem, policies impact both directly and indirectly all its firms through their linkages. This opens up oppor tunities for industrial policy packages including instruments supporting small firms
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
94 Antonio Andreoni and William Lazonick indirectly by regulating or allocating resources to big ones; as well as situations where policies support small companies’ continuous upgrading to keep the locality attractive to big companies and avoid the de-linking of big and small firms.
5.5.3 Lessons from Silicon Valley The growth of Silicon Valley in the immediate post-World War II decades was underpinned by the money and research of the US military–industrial complex, bringing high technology from the US East, where MIT was the premier university, to the US West, where Stanford took the lead (Leslie 1993). The proliferation of semiconductor start-ups in the 1960s was directly responsible for the emergence of the Silicon Valley venture-capital industry in the 1970s, which by the last years of that decade became a powerful national force, lobbying for cuts in the capital-gains tax and access to the money held by pension funds (Lazonick 2009). What is distinctive about Silicon Valley’s history, however, is not just that it spawned new ventures but that some of these firms grew to be amongst the largest and most profitable high-tech companies in the world. Intel, founded in 1968, had 21,000 employees two dec ades later, and has had around 107,000 employees since 2013. Apple had almost 18,000 employees by 1995, and, even though it outsources its manufacturing, in 2018 its payroll was at 132,000. Oracle employed 37,000 in 1997, twenty years after its founding, and in 2018 that number was 136,000. Cisco, founded in 1984, grew to 34,000 employees in 2000, and reached a high of 75,900 in 2019. The problem is that, as these companies have become highly profitable, they have pumped out virtually all their profits to shareholders. This financialized behaviour reflects not only the dominance of shareholder-value ideology and institutions in the United States as a whole but also the integral role of the stock market for creation, compensation, and com bination in the NEBM through which these companies became dominant (Lazonick 2018). Combined for 2009–18, these four companies—Intel, Apple, Oracle, and Cisco—have dis tributed $168 billion in dividends (25 per cent of profits) and $474 billion in buybacks (73 per cent of profits). These Silicon Valley companies have become world leaders in what William Lazonick and Jang-Sup Shin call ‘predatory value extraction’, concentrating income amongst the richest households and failing to upgrade the productive capabilities of the US labour force (Lazonick and Shin 2020). Neither these companies nor Silicon Valley will disappear any time soon. But, opting for a financialized business model, they are losing their innovative advantage. Taking their place are other districts in other parts of the world where the social conditions of innovative enterprise are gaining productive strength.5 The lesson is that corporate gov ernance regulation has become a critical dimension of industrial policy (Andreoni et al. 2019). By influencing strategic control and financial commitment, corporate governance regulation can direct business strategies towards a more sustainable logic of retain and reinvest. This approach is not simply relevant for the big companies of the type mentioned above; it also matters for the supply chains of the localities in which they operate. A finan cialized system integrator like Cisco is in fact a threat not only to its own employees 5 See in this volume, Chapter 30.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Local Ecosystems and Social Conditions of Innovative Enterprise 95 operating in a certain locality but also to the entire supply chain of firms whose develop ment depends on the system integrator’s continuous investment in innovation.
References Andreoni, Antonio (2014) ‘Structural Learning: Embedding Discoveries and the Dynamics of Pro duction’, Structural Change and Economic Dynamics 29: 58–74. Andreoni, Antonio (2016) ‘Varieties of Industrial Policy: Models, Packages and Transformation Cycles’, in Akbar Noman and Joseph E. Stiglitz (eds) Efficiency, Finance and Varieties of Industrial Policy. New York: Columbia University Press, pp. 245–305. Andreoni, Antonio (2018) ‘The Architecture and Dynamics of Industrial Ecosystems: Diversifica tion and Innovative Industrial Renewal in Emilia Romagna’, Cambridge Journal of Economics 42: 1613–42. Andreoni, Antonio (2019) ‘A Generalised Linkage Approach to Local Production Systems Develop ment in the Era of Global Value Chains, with Special Reference to Africa’, in Ravi Kanbur, Akbar Noman, and Joseph E. Stiglitz (eds) Quality of Growth in Africa. New York: Columbia University Press, pp. 264–94. Andreoni, Antonio and Ha-Joon Chang (2019) ‘The Political Economy of Industrial Policy: Structural Interdependencies, Policy Alignment and Conflict Management’, Structural Change and Economic Dynamics 48: 136–50. Andreoni, Antonio and Eoin O’Sullivan (2014) ‘Vectors of Innovation in Industrial Ecosystems’, Edith Penrose Centenary Conference, 14–15 November 2014, SOAS University of London. Andreoni, Antonio and Roberto Scazzieri (2014) ‘Triggers of Change: Structural Trajectories and Pro duction Dynamics’, Cambridge Journal of Economics 38: 1391–408. Andreoni, Antonio, Frederico Frattini, and Giorgio Prodi (2017) ‘Structural Cycles and Industrial Policy Alignment: The Private–Public Nexus in the Emilian Packaging Valley’, Cambridge Journal of Economics 41: 881–904. Andreoni, Antonio, Matt Hopkins, and William Lazonick (2019) ‘Governing Financialisation: A UK Industrial Strategy to Retain and Reinvest’. GOFINPRO Working Paper, SOAS University of London. Becattini, Giacomo (1978) ‘The Development of Light Industry in Tuscany: An Interpretation’, Economic Notes 7: 107–23. Becattini, Giacomo (1989) ‘Sectors and/or Districts: Some Remarks on the Conceptual Foundations of Industrial Economics’, in Edward Goodman, Julia Bamford, and Peter Saynor (eds) Small Firms and Industrial Districts in Italy. London: Routledge, pp. 123–35. Becattini, Giacomo, Marco Bellandi, and Lisa De Propris (eds) (2009) A Handbook of Industrial Districts. Cheltenham: Edward Elgar. Berlin, Leslie (2001) ‘Robert Noyce and Fairchild Semiconductor, 1957–1968’, Business History Review 75: 63–102. Best, Michael (1990) The New Competition: Institutions of Industrial Restructuring. Cambridge, MA: Harvard University Press. Bianchi, Patrizio (2013) La rincorsa frenata. L’industria italiana dall’unita’ nazionale all’ unificazione europea. Bologna: Il Mulino. Brusco, Sebastiano (1975) ‘Organizzazione del lavoro e decentramento produttivo nel settore metalmeccanico’, Sindacato e piccola impresa (a cura della FLM di Bergamo), Bari, De Donato. Brusco, Sebastiano (1982) ‘The Emilian Model: Productive Decentralization and Social Integration’, Cambridge Journal of Economics 6: 167–84. Brusco, Sebastiano and Sergio Paba (2014) ‘Towards a History of the Italian Industrial Districts from the End of World War II to the Nineties’. DEMB Working Paper Series No. 37, December. Camagni, Roberto (1994) ‘Il concetto di milieu innovateur’, in Gioacchino Garofoli and R. Mazzoni (eds) Sistemi produttivi locali: struttura e trasformazione. Milano: Angeli, pp. 27–58. Carpenter, Mary and William Lazonick (2019) ‘Why Isn’t Cisco Huawei?’. AIR Working Paper, August.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
96 Antonio Andreoni and William Lazonick Carpenter, Marie, William Lazonick, and Mary O’Sullivan (2003) ‘The Stock Market and Innovative Capability in the New Economy: The Optical Networking Industry’, Industrial and Corporate Change 12: 963–1034. Chandler, Alfred D. Jr. (1962) Strategy and Structure: Chapters in the History of Industrial Enterprise. Cambridge, MA: MIT Press. Dei Ottati, Gabi (2018) ‘Marshallian Industrial Districts in Italy: The End of a Model or Adaptation to the Global Economy?’, Cambridge Journal of Economics 42: 259–84. Elbaum, Bernard and William Lazonick (eds) (1986) The Decline of the British Economy. Oxford: Oxford University Press. Farnie, Douglas (1979) The English Cotton Industry and the World Market, 1815–1896. Cambridge: Cambridge University Press. Hirschman, Albert O. (1958) Strategy of Economic Development. New Haven, CT: Yale University Press. Hirschman, Albert O. (1977) ‘A Generalized Linkage Approach to Development, with Special Refer ence to Staples’, Economic Development and Cultural Change 25 (Supplement): 67–98. ISTAT (2015) ‘I distretti industriali’, Statistiche Report, Febbraio. Kenney, Martin (ed.) (2000) Understanding Silicon Valley: The Anatomy of an Entrepreneurial Region. Stanford, CA: Stanford University Press. Kerstenetzky, J. (2010) ‘Alfred Marshall on Big Business’, Cambridge Journal of Economics 34: 569–86. Lazonick, William (1979) ‘Industrial Relations and Technical Change: The Case of the Self-Acting Mule’, Cambridge Journal of Economics 3: 231–62. Lazonick, William (1981a) ‘Competition, Specialization, and Industrial Decline’, Journal of Economic History 41: 31–8. Lazonick, William (1981b) ‘Factor Costs and the Diffusion of Ring Spinning in Britain prior to World War I’, Quarterly Journal of Economics 96: 89–109. Lazonick, William (1983) ‘Industrial Organization and Technological Change’, Business History Review 57: 195–236. Lazonick, William (1990) Competitive Advantage on the Shop Floor. Cambridge, MA: Harvard Uni versity Press. Lazonick, William (1993) ‘Industry Clusters and Global Webs: Organizational Capabilities in the US Economy’, Industrial and Corporate Change 2: 1–24. Lazonick, William (2009) Sustainable Prosperity in the New Economy? Business Organization and High-Tech Employment in the United States. Kalamazoo, MI: Upjohn Institute for Employment Research. Lazonick, William (2018) ‘The Functions of the Stock Market and the Fallacies of Shareholder Value’, in Ciaran Driver and Grahame Thompson (eds) Corporate Governance in Contention. Oxford: Oxford University Press, pp. 117–51. Lazonick, William (2019a) ‘Apple’s “Capital Return Program”: Where Are the Patient Capitalists?’, In stitute for New Economic Thinking Blog, November. Available at https://www.ineteconomics.org/ perspectives/blog/apples-capital-return-program-where-are-the-patient-capitalists. Lazonick, William (2019b) ‘Innovative Enterprise and Sustainable Prosperity’. AIR Working Paper, January. Available at http://www.theairnet.org/v3/backbone/uploads/2019/03/LazonickIESP-20190118.pdf. Lazonick, William and Jang-Sup Shin (2020) Predatory Value Extraction: How the Looting of the Business Corporation Became the US Norm and How Sustainable Prosperity Can Be Restored. Oxford: Oxford University Press. Lazonick, William, Matt Hopkins, and Ken Jacobson (2016) ‘What We Learn about Inequality from Carl Icahn’s $2 Billion “No Brainer” ’. Institute for New Economic Thinking Ideas & Papers, June. Available at https://www.ineteconomics.org/perspectives/blog/what-we-learn-about-inequalityfrom-carl-icahns-2-billion-apple-no-brainer. Lazonick, William, Mariana Mazzucato, and Öner Tulum (2013) ‘Apple’s Changing Business Model: What Should the World’s Richest Company Do With All Those Profits?’, Accounting Forum 37: 249–67.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Local Ecosystems and Social Conditions of Innovative Enterprise 97 Lécuyer, Christophe (2006) Making Silicon Valley: Innovation and the Growth of High Tech, 1930–1970. Cambridge, MA: MIT Press. Leslie, Stuart W. (1993) The Cold War and American Science: The Military–Industrial Complex at MIT and Stanford. New York: Columbia University Press. Marshall, Alfred (1920) Industry and Trade: A Study of Industrial Technique and Business Organization, and of their Influences on the Conditions of Various Classes and Nations. London: Macmillan. Marshall, Alfred ([1920] 1998) Principles of Economics. Vol. 1. London: Macmillan. Mass, William and William Lazonick (1990) ‘The British Cotton Industry and International Competi tive Advantage: The State of the Debates’, Business History 32: 9–65. Mass, William and Andrew Robertson (1996) ‘From Textiles to Automobiles: Mechanical and Organi zational Innovation in the Toyoda Enterprise, 1895–1933’, Business and Economic History 25: 1–37. Moore, Gordon (1996) ‘Some Personal Perspectives on Research in the Semiconductor Industry’, in Richard Rosenbloom and William Spencer (eds) Engines of Innovation: US Industrial Research at the End of an Era. Cambridge, MA: Harvard Business School Press, pp. 165–74. Packard, David (1995) The HP Way: How Bill Hewlett and I Built our Company. New York: Collins Business. Penrose, Edith T. (1959) The Theory of the Growth of the Firm. Oxford: Blackwell. Porter, Michael (1990) The Competitive Advantage of Nations. New York: Free Press. Regione Emilia Romagna Website: https://www.regione.emilia-romagna.it/. Richardson, George (1972) ‘The Organisation of Industry’, Economic Journal 84: 883–96. Russo, Margherita and Josh Whitford (2009) ‘Industrial Districts in a Globalizing World: A Model to Change or a Model of Change?’, Department of Economics, University of Modena and Reggio Emilia. Sabel, Charles (1982) Work and Politics: The Division of Labour in Industry. Cambridge: Cambridge University Press. Salais, Robert and Michael Storper (1992) ‘The Four “Worlds” of Contemporary Industry’, Cambridge Journal of Economics 16: 169–93. Whitaker, J. K. (ed.) (1975) The Early Economic Writings of Alfred Marshall, 1867–1890. 2 vols. London: Macmillan. Young, Allyn A. (1928) ‘Increasing Returns and Economic Progress’, Economic Journal 38: 527–42.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
chapter 6
I n dustr i a l Policy, I nstitu tiona l Tr a nsfor m ation, a n d the Dev el opm en t of I n dustr i a l Pa r ks Howard Stein
6.1 Introduction The ‘rediscovery’ in orthodox circles of the centrality of industrialization to the development process in Africa and other developing countries has ignited debates on how to best proceed. A central part of the discussion is how to create conditions conducive to the attraction and expansion of investment that will facilitate the growth and development of a manufacturing sector. Standard neoclassical economics draws narrowly on the issue of market failures like the paucity of public goods or externalities which are not efficiently provided by private markets. From their perspective, the key policy initiative is to address these market imperfections by focusing on the provision of public goods like infrastructure in a spatially contained and delimited area or zone thereby minimizing the resource usage while maximizing the impact. Containment also has the added benefit of providing positive externalities such as information, technology, and labour skills within the zones with the potential for spillovers into markets near the zones. This becomes the main rationale for organizing an industrial park and is also key to explaining their failures. In contrast, it will be argued in this chapter that industrial parks should not be seen as an isolated phenomenon trying to address a set of deductively posited spuriously generated impediments to industrialization. Instead they should be conceived as a policy tool which needs to be linked into a broader strategy aimed at transforming the organizational capacities and institutional relationships that underlie the generation of industrial zones, and how they can best be connected to other sectors of the economy. For this purpose, the chapter
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrial Policy, Institutional Transformation 99 will draw on non-neoclassical theoretical traditions and experiences that can help us better understand the prerequisites of successful industrial hubs and the reasons for their failures. The chapter begins with a critical review of the mainstream arguments for designing and running industrial parks, with a focus on an Indian example, before turning to a discussion of industrial policy in other traditions and their relationship to industrial parks.
6.2 Neoclassical Economic Approaches to Industrial Parks As seen in Chapter 2, industrial parks or hubs encompass a variety of forms and continue to proliferate, though the success of these zones has been mixed (Stein 2012). Although the idea of the industrial park originated with the private sector taking advantage of a market opportunity in late nineteenth-century Trafford Park near Manchester (Tuscor Lloyds 2017), the long-term success of parks is linked to their ability to transcend opportunities dictated by the moment by markets. They need to alter their framework in anticipation of changing conditions. Parks have had the largest impact when they can tap into the planning, resources, and regulatory structure of states, particularly their industrial policy (IP) frameworks and priorities.1 What is less understood is how this is done in practice. It also raises the question of the best conceptual tools to comprehend the process. Arguably, neoclassical economics provides a very poor understanding of the organization, transformation, failure, and success of industrial parks along with the policy tools to make parks effective. Neoclassical theory assumes markets are efficient and operate well in allocating goods and services without state intervention. Hence, state policy should focus on maintaining property rights, generating a business-friendly environment with minimal regulation of capital, labour, and product markets, and providing some public and social goods like health, education, and infrastructure. Further government intervention can distort incentives and alter the allocative efficiency of markets (Shleifer and Vishny 1994). In the neoclassical world, there are two possible avenues for state intervention in support of industry. First, while neoclassicists are firmly committed to free trade and the principles of comparative advantage, there is some recognition that new industries are at a distinct disadvantage. With time, domestic industries can lower their production costs through learning by doing, improve their efficiency, and eventually compete with more mature companies. Hence, industrial and trade policy in support of infant industries can help develop latent comparative advantages as firms reach sufficient scale to be able to compete inter nationally (Pack and Saggi 2006). Second, neoclassical theory recognizes the existence of market failures which can impede the industrialization process particularly in the poorly formed markets of developing countries. Information asymmetries are common and can occur when information on opportunities for profitable businesses is costly or unavailable. Coordination problems can
1 White (2008) sees industrial policy as a ‘concerted, focused, conscious effort on the part of government to encourage and promote a specific industry or sector with an array of policy tools’.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
100 Howard Stein also arise from information asymmetries when project success and profitability rely on simultaneous investment. Governments can intervene to reduce the barriers to accessing the information or find ways to generate and disseminate it so the private sector can act (Pack and Westphal 1986). In addition, innovations benefit not only the innovator but the economy at large. However, the costs of any failure are borne privately. A market failure arises as the cost and benefits of the discovery process are misaligned for first movers, with implications for the economy as a whole, and can justify subsidizing new investment (Rodrik 2004). Neoclassical proponents of industrial policy prefer strategic collaboration with the p rivate sector in order to identify and address market imperfections. A more private-sector-driven approach can yield knowledge of the market opportunities and constraints which will help the development of strategies to deal with the gaps created by market imperfections (Rodrik 2004). Policies to design and support industrial zones are defined in similar terms. First, the provision of infrastructure is much easier to plan and finance in a geographically limited space with the advantage of minimizing the usage of scarce government resources with their potential to create distortions in markets through unnecessary taxation. Second, the concentration of firms encouraged by the formation of parks develops knowledge and technology, and a pool of skilled labour with likely spillover inside the park and nearby. Third, it creates the potential for the formation of clusters which can help improve the efficiency of transactions, reduce information asymmetries, and lower transportation costs. The close proximity of firms also provides incentives for a lead firm to overcome free riding and take the initiative to form clusters (Saleman and Jordan 2014). Hence parks are aimed at dealing with common market imperfections including public goods, externalities, free riding, information asymmetries, and transaction and transportation costs, all of which impede the operation of markets. Drawing on neoclassical economics, Saleman and Jordan (2014) of the World Bank define these zones in these narrow terms: ‘involving the provision of common infrastructure to a group of industrial firms in a demarcated area’ (6). The exact nature of the zone will be a product of the ‘forms involving additional regulatory or other measures [which] will be specifically denoted as such’. The key is in developing a particular form of industrial organization which focuses on clustering ‘a concentration of interconnected firms in a particular field. Such clusters can form “organically”, or can be the target of deliberate policies’ (6). However, they warn that the pursuit of cluster development as a policy focus is fraught with danger. Accusations of land grabbing and speculation can arise, particularly when park construction is delayed. Perceptions of special treatment and corruption can arise when park residents get special treatment. Most important is the issue of the counterfactual that the parks would not have risen spontaneously without state support, hence public money was needlessly spent and tax revenues needlessly foregone while companies have received tax breaks for locating in the park. The failure of parks draws on the same neoclassical theory. The largest potential danger is relying on states which are largely conceptualized in public choice terms. States have tendencies towards ‘producing white elephants; eroding the tax base; creating vehicles for land speculation; delivering hand-outs to favored firms; and funneling spending to favored districts. That is if the parks are even completed in less than a decade’ (Saleman and Jordan 2014: 4). Along these lines there are four main types of performance failures: that ‘parks do not get built; the parks are
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrial Policy, Institutional Transformation 101 built but there is little demand from firms to locate and invest in them; the parks are built and generate demand, but with few “cluster effects”; the parks are successful but have neutral or negative side-effects on investment climate outside the park (“negative spillovers” and “crowding out”)’ (7). Hence, there is not only a failure to produce positive effects but the potential for state-generated failures leading not only to large opportunity costs in resources, human capital, and foregone tax revenues, but also to negative externalities with consequences for the operation of private markets. Failures can arise from poor selection of park locations, inability to assess potential demand or to procure sufficient land, poor master planning, paucity of available finance, insufficient procurement of infrastructure construction, or poor construction design, evaluation, monitoring and operation, and maintenance. The key here is determining what role the private and public sectors should play in these various park functions.
6.3 Neoclassical Economics and the Textile Parks of India: An Exploration The remainder of this chapter, drawing on the work of Saleman and Jordan (2014), looks at how India was able to avoid or address failures in the creation of its textile industrial parks. These authors’ predilection for the role of states largely explains their selection of this particular case study: ‘The SITP’s (Scheme for Integrated Textile Parks) most prominent innov ation is . . . the policy makes the entrepreneurs the drivers and ultimate decision-makers of the entire initiative for the creation and functioning of the park’ (Saleman and Jordan 2014: 10). The Ministry of Textiles sets the framework for implementation and funding but then hands over the implementation and monitoring to the Special Purpose Vehicle (SPV) or overseeing group which is self-organized by the future tenants of the park. The SPV selects the project management consultant (PMC) from a Ministry list to prepare a detailed project report (DPR). They are free to use any criteria. The Ministry approves the proposal and allocates funds to cover up to 40 per cent of the infrastructure costs but does not take equity which is also appealing to the authors since it limits government exposure, with fewer spending opportunity costs and government-generated distortions. The approach is also very much in line with the recent agenda of the World Bank which has been pushing private–public partnerships (PPP), seen as the most efficient way to provide infrastructure.2 World Bank authors have a long history of undertaking studies to try to justify existing policy frameworks and agendas and tend to be theoretically embedded in the neoclasssical economic traditions (Stein 2008). Much of the remaining part of Saleman and Jordan’s (2014) study is aimed at empirically illustrating the superiority of this PPP approach to industrial parks. They claim that by October 2012, all forty parks approved had begun construction with eleven fully completed, thirteen starting commercial construction, and sixteen at various stages of construction— hence this superior model had already avoided the first two main performance failures. 2 ‘The World Bank Group helps governments design public–private partnerships (PPPs) and create a balanced regulatory environment to ensure a more efficient and sustainable provision of public services and infrastructure.’ See https://www.worldbank.org/en/topic/publicprivatepartnerships.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
102 Howard Stein Perhaps they collected the wrong data, or things rapidly changed a few years later, or someone told them what they wanted to hear. However, an independent study of the parks commissioned by the Ministry of Textiles presented a radically different picture.3 Of the seventy-four parks approved between 2005 and 2015, only thirty were functional. Nineteen of the seventy-four were cancelled or had applied for cancellation. Causes of cancellation included delays in approval including clearances, disputes over land, or the inability to secure the land, and differences between the members of the SPVs (Wazir Advisors 2016). Saleman and Jordan (2014) also talk of the impressive creation of 42,000 jobs through 2012 ‘which is promising given most of the parks are only now commencing production’. However, by the end of 2015, the numbers had reached only 68,000 compared to the 2016 target of 360,000 which is just under 19 per cent. The numbers are even less overwhelming when one considers that 17,500 jobs are in one park, Brandix India Apparel City, meaning that the other twenty-nine parks only average about 1,700 employees each. Moreover, eleven of the thirty parks had occupancy rates below 50 per cent, with only four reaching 100 per cent occupancy—hardly an indication of overwhelming success (Wazir Advisors 2016). Little data is provided on many of the other purported positive features of this form of the park, a product of deductive reasoning arising from neoclassical principles. For example, Saleman and Jordan (2014) argue that SPVs work because they are able to overcome collective action problems and because the lead firm’s organizational involvement will ensure the benefits of clustering (24). However, they also readily admit that ‘it is difficult to make definitive conclusions on this score’ and hence it is conjecture (15). Based on a survey of parks in places like China, UAE, Oman, Italy, Turkey, Morocco, and South Korea, the study for the Ministry of Textiles finds very weak clustering, partly due to the design of the parks themselves compared to the case studies. First, many of the textile parks are too small, some being only 25 acres, which reduces the critical mass threshold level needed for clustering. The parks in the comparison locations generally exceed 100 acres. Second, many parks have only single value-chain activities (like weaving) and are not integrated. In contrast, the case study parks have been supported by the government to provide full value-chain integration from fibre to finished goods. Third, the parks are impeded by the lack of government support for infrastructure, marketing, and securing land, which elsewhere are generally the responsibility of governments. Internationally, marketing support and investment promotion is a key focal point both domestically and overseas. As discussed above, central government support is limited to partial financing of infrastructure inside the parks. Infrastructure outside the park is up to state governments. The paucity of clustering was also due to other performance dimensions. Parks have filled up slowly, with only four parks fully functional. The remaining parks are only partly functional, limiting real integration. The demand-driven approach so appealing to the World Bank neoclassical economists leaves the parks with no coordination of activities. ‘In many instances, products being manufactured by upstream units are not of the type required by the downstream ones while in some instances the upstream units are unable to meet the downstream demands’ (52). Moreover, SPVs are under financial pressure to fill their parks 3 One reviewer of this chapter argued that there was no independent information and hence it is hard to know which claim about the operation of the park was accurate. It should be emphasized that the government hardly took these reports as equivalent and that the Wazir Associates Report led to a major reassessment of the operation of the parks and their design (Financial Express 2017).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrial Policy, Institutional Transformation 103 and have few incentives to search for manufactures that would better integrate and complete missing components in the value chain. Marketing and raw material sourcing not only lack economies of scale and coordination but also have limited skill and technology spillovers that could lead to anticipated production costs. No park showed evidence of cost reductions. In addition to value-chain integration, it was envisaged that improved technology and skilling of workers would lead to reduced production costs. However, no park has reported success. Even worse, Wazir Advisors (2016) reported costs above those found outside the park due to higher land costs, partly as a result of variations in state laws on norms of space usage and higher construction costs. In sum, neoclassical economists writing on industrial parks are not convinced that ‘the activity in parks would not have happened without their construction and the public money that has been spent to support it’ and cite evidence that clusters often ‘emerged spontaneously’ (Saleman and Jordan 2014). When examining the operation of parks they are wedded to explaining successes in terms of theories that aim to correct slight imperfections in markets with minimal state intervention and expenditures. The weakness of the Saleman and Jordan (2014) analysis is a reflection of the more general flaws embedded in a neoclassical economic understanding of industrial policy. The problem is embedded in the very way that neoclassicists envision markets as exchanges of independent agents coordinated via the price system in markets. However, production largely transpires within interconnected and cooperative networks. In the vision of perfect competition, these forms of interaction are deemed ‘extra-market’ and are antithetical to the operation of efficient markets. However these ‘imperfections’ are often key to the development of new products or providing the financial means for production. Their removal can lead to negative consequences (Sawyer 2000). There are other issues. The conceptualization of imperfections is a contrived product of a theoretical structure and has no meaning outside of these constructs. Philosophers call this concept by postulation. Students of the field of economics are inculcated early on with the vision of perfect markets satisfying utility-maximizing individualistic exchanges generated as the highest ideal of welfare with imperfections postulated as the impediment to reaching this apotheosis. In this world, policy is a precept for achieving collective action goals driven by atomistic models of rational choice. Consequently, prescriptive consequentialism displaces human reasoning in the creation of strategies of intervention for public goals (Bromley 2007). Neoclassical economic authors like Hausmann and Rodrik (2006) are widely cited as inspiring new thinking on industrial policy. However, Chang and Andreoni (2016) consider their arguments on coordination and related failures as a neoclassical repackaging of the ideas of post war development economists like Nurkse and Rosenstein-Rodan. They show, for example, that the Hausmann–Rodrik information externalities is an incomplete version of the argument underlying the need for infant industry support. Their focus is on supporting a firm-specific pioneer with the view that their knowledge will be of use and accessible to other firms in the sector rather than just the firm itself. A more complete version of the infant industry strategy transcends the focus on a pioneering firm to develop approaches that will support the entire sector with the potential for a more systemic impact on the broader economy through management techniques, job training, export strategies, and other interventions. To understand the potential of industrial policy and the role that industrial parks can play in industrial strategies, we need to move away from the chimerical neoclassical world
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
104 Howard Stein driven by a vision of economies operating by putative laws to the real world where e conomies are shaped by the nature of their institutions. A key element is to move away from viewing parks as isolated phenomena addressing local market failures to seeing them as an integral part of the rubric of a country’s industrial policy.
6.4 Alternative Theoretical Constructs for Building Industrial Policy and Parks There are four general theoretical traditions which can be used to help form the building blocks of industrial policy with applications for understanding the potential of industrial parks. Each has its own distinct theoretical contribution but they are also complementary and together provide a more comprehensive picture of industrial policy and parks.4
6.4.1 Structuralism The premise of structuralists is that development requires the structural transformation of economies, and that losses due to wealth and income limit the ability of markets to effectively shift resources to accommodate the transformation. Chapter 3 points to the neoclassical problem with conceptualizing increasing returns which is at the heart of industrialization and draws heavily on structural traditions in development economics to conceptualize industrial policy. I will only briefly review a few relevant points here. Structuralism ‘views changes in the composition of economic activity as amongst the prime movers of growth and employment, and the literature in this tradition therefore focuses on exploring that relationship, taking a holistic approach that considers the role of macroeconomics, trade, technology, and sectoral practices’ (Salazar-Xirinachs et al. 2014). Industrial-sector development is particularly important; high-quality jobs and meaningful poverty reduction ‘hinge on diversifying the economy from low-productivity agriculture and informal sectors to high-productivity sectors such as manufacturing and modern services’ which also absorb labour(UNECA 2014). Competitiveness in new sectors can be generated by government support of skills development, investments in technologies, which can be enhanced through the importation and adaptation of technology, the acquisition of technology via FDI, and the generation of new technology via investment in research and development. Technological changes are catalysed by the emergence of intricate transport and communication systems (Lall 2013). An industrial policy focused on structural transformation is key but says little about how it is to be accomplished. Picking winning industries and sectors is a complex process. Bureaucrats can lack the knowledge to make decisions on what sectors to promote. At the heart of addressing these and other challenges are understanding states, industrial and 4 This section draws on chapter 2 of Stein et al. (forthcoming).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrial Policy, Institutional Transformation 105 industrial policy capabilities, and institutional arrangements. The key question when debating states and industrial policy should be less on why governments have been so weak and more on probing how some developing countries have done very well (Whitfield et al. 2015).
6.4.2 The Developmental State The rise of East Asian economies where the state used industrial policy with great success has led to seminal work by scholars such as Johnson (1982), Wade (1990), Amsden (1989), and Evans (1995) aimed at understanding the developmental state. Two aspects have been highlighted. Politically, the strategic power of the developmental state in East Asia was built on coalitions with domestic industry while limiting the power of organized labour and popular groups. Links to the private sector were crucial but bureaucratic autonomy coexisted with public–private cooperation conditions to enable the development by the state and bureaucratic elites of independent national goals that could translate into effective policies. Wade (1990) argues that East Asian economies owe their success to high levels of investment in key sectors of the economy, encouraged by government intervention and exposure of industries to global competition. Evans (1995) introduces the notion of embedded autonomy and comparative institutions where states undertake four roles: custodian, demiurge, midwife, and husbandry. The state acts as a custodian with its regulatory powers to police and protect, while as a demiurge it assists production by assuming risk and reducing uncertainty in the market. Midwife duties focus on credit to promote new investment activities, and husbandry cultivates, prods, and nurtures entrepreneurship. Evans also introduces the key concept of embedded autonomy with a strong, capable, goal-setting bureaucracy with institutionalized channels to groups with joint projects where goals and policies are negotiated and renegotiated. The need for a developmental state and associated industrial policy is particularly important for dealing with the challenges of late industrialization with its disadvantages relative to so many other countries that have already pursued the path. To compensate, governments must direct their economies to shift into strategic sectors while ensuring perpetuation of capital accumulation not only in competing industries but in new industries with the potential for leapfrogging. This can be even more challenging in today’s international trade climate which is dominated by multinational corporations (MNCs) that control value chains where as much as 50 per cent (in 2014) of all trade is between MNCs and their affiliates (Cadestin et al. 2018). Wade (2015) argues that though today’s climate is more difficult, there are still plenty of opportunities for the developmental state to operate, including investment incentives, currency adjustments, export taxes, and financing trade. Moreover, industrial parks can focus on assisting industrial access to selective areas of the value chain, including bargaining with MNCs to upgrade skills in return for priority access to domestic markets. However, little of this has been done, with many governments captivated by the neoliberal Doing Business indicators which push governments to passively attract FDI. Understanding the political dimension of developmental states is also important. Industrial policy programmes are not merely technical but very political. For the purposes of legitimacy, they require engagement, credibility, and the support of citizens with clear indicators that they benefit everyone, not just a small group. It helps if the bureaucracy
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
106 Howard Stein involved is professional, skilled, and shielded from undue political pressure (Chang 2004). However, they also require a power base and the support of political coalitions that will favour industrialization. Policy arises from the politics of survival of ruling elites which require ‘political settlements’ to remain in power. There are a number of factors underlying the ability of political coalitions to generate an effective industrial policy. First is the vulnerability that arises from internal opposition and from groups outside the ruling faction. Second are the technological capabilities of the industrial capitalists themselves and the extent of their power inside ruling coalitions. Their position is buttressed by their ability to perform economically including their generation of jobs, tax revenue, and foreign exchange. Expansion into new industries alters the terrain of power and allows states to be more developmental in their usage of rents for industrial policy purposes (Whitfield et al. 2015). One serious pitfall of the political settlements approach is the inability to predetermine the configurations of coalitions and capitalist groups that will lead to a positive outcome. However, what is clearly set out is the importance of the technical capabilities of industrialists, yet there is little or no insight into how this is to be developed and supported once developed. In general while the developmental states theory pinpoints important political dimensions and the wide government structures and strategies underpinning industrial policy, other theoretical approaches can be drawn on to better understand the institutional building blocks of an industrial-policy-led drive to industrialization.
6.4.3 The Capabilities Approach The development of capabilities is a core component of industrialization and industrial policy. Wade (2015) points to a particularly challenging ‘capabilities trap’ that tends to afflict countries that cannot compete with low-income countries and have little indigenous capacity to design and innovate into new products. However, the challenge is not only for middle-income countries but at the core of formation of the productive capabilities aimed at catching up. The key elements in building capabilities are to expand knowledge structures along with routines and institutions. Knowledge structures are acquired and embedded as a collective memory in labour forces as individuals acquire a range of conceptual and procedural information. As they expand they also shape economic options and competences for productive transformation. A series of technological knowledge communities each with a unique set of abilities can be distinguished. Tasks and products might draw on one or more of these communities for successful production. The greater the depth and diversity of these knowledge communities, the greater the possibility of drawing on them to produce new goods as circumstances and opportunities change. Knowledge communities that are industry specific can limit this flexibility. Some industries are more problematic. Amongst the worst are the extractive industries so commonly found in sub-Saharan Africa. The skill profiles of their workforce are not readily transferred, cutting down on the possibilities of diversification (Nübler 2014). Routines are also salient to building up procedural knowledge and evolve through experience. They are the carriers of collective competences in firms and organizations and should be differentiated from institutions which carry collective competences in the
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrial Policy, Institutional Transformation 107 economy as a whole. At the economy level, institutions that encourage entrepreneurship, trust, and social consensus can have a powerful impact on collective competences. The production and knowledge spheres are connected through a process of circular caus ation. Productive systems are major sites of learning of routines and technological know ledge while also influencing institutions and skill creation. Hence, productive transformation affects the knowledge sphere which in turn influences the transformation of production through knowledge structures, routines, and institutions. Hence, the process of circular causation and a virtuous circle of productive transformation and learning lead to the potential to enhance employment and the general standard of living. Instruments of industrial policy can be used to improve knowledge structures. Value chains and other industrial networks can be encouraged with incentives and pressures to ensure that the transfer of technology and the enhancement of learning can be fostered in parent–affiliate relationships. India has been successful in setting up a system where parent firms in the auto industry insist on standards and certification from even the smaller firms in the informal economy. China used foreign ownership restrictions and local content rules to encourage the development of knowledge in the auto industry (Nübler 2014). New institutions are also central to constructing the building blocks of industrial policy, parks, and industrialization.
6.4.4 Organizations and the Institutional Approach to Industrial Policy From an institutional perspective, industrial policy has not been conceptualized well. Industrial development is a complex process that involves significant institutional trans formation. In the institutionalist view, society can be seen as a collection of institutional systems or a set of institutions. In the Veblenian tradition, institutions are seen as a ‘set of socially prescribed patterns of correlated behavior’ (Bush 1987: 1076). The idea of socially prescribed patterns is linked to the notion that human tools and intelligence are brought to govern the daily processes of life. Institutions should also be seen as habits of thought common to the generality of men and women. Habits alone are not institutions but predictable responses to contextualized stimuli. However, habits of thought invoke a human dynamic that connects the thinking process to purposive outcomes. They must also be generalized so that they are not idiosyncratic but connected to others with coordination and unity of purpose (Stein 2008). Organizations require common habits of thought that will allow them to operate effectively. However, they also have a host of other dimensions including structures, internal regulations which are affected by external laws, and internal and external power relationships. An institutional approach to organizations recognizes that they are socially and/or politically constituted entities that operate in accordance with norms, rules, and beliefs both within their structures and relative to other entities. If we see states as a set of interlocking organizations then we can see industrial policy frameworks as a substratum of the broader rubric of the state that is not only connected to the state in multiple ways but also has its own semi-autonomous operating structures. Organizations are central to the enhancement of capabilities. They are entities of containment linked spatially and temporally to the shifting
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
108 Howard Stein localized environment, yet have their own internal dynamic. They have the power to expand, shape, and enhance conceptual and procedural capabilities that are vital to both the formation and implementation of industrial-related policies and the actual industrialization process itself. One way organizations are connected institutionally is through ‘organizational fields’. Scott and Meyer (1994) define an organizational field as ‘a community of organizations that partakes of a common meaning system and whose participants interact more frequently and fatefully with one another than with actors outside the field’ (56). The concept is useful in understanding the nature of industrial policy; it can be conceived as a broad network of government ministries, industrial policy organizations (IPOs), and private-sector firms connected by a common set of constructs aimed at a coordinated effort to reproduce and expand industrial production. Industrial parks should be seen as important units within that organizational field. If organizations are conceptually (and often legally) recognized entities that combine groups of people with defined common rules and purposes then an IPO like an industrial park authority is an entity whose purpose is to create, support, alter, or transform industrialization. While IPOs are generally state entities they can also be units that are quasi-state entities or those within the private sector with contingent financial support from the state and/or state-approved mandates to support industrialization. Industrial policy organizations, like government agencies, operate in accordance with behaviour habitual to their institutional environment. Each IPO has a policy or set of policy functions and also a network of informal and formal policy actors which, as argued above, can constitute an organizational field. Understanding the nature of the functions and networks and how they interact is important for diagnosing problems and the potential for change. Within this context, agents are not simply robotically part of the organizational machinery but are capable of reflexive deliberation and can focus both on desirable actions within organizational structures and on changes in the structures themselves. Interactions with other agents through communicative reflexivity allow agents to make decisions on what is perceived to be the correct action. In this context one can perpetuate the policy purpose or on occasion express habit-breaking behaviour which can lead to entrepreneurial policymaking opportunities.
6.5 Applications to Industrial Policy and Industrial Parks: Key Elements We have presented ideas and theoretical constructs from four different approaches beyond mainstream neoclassical economic theory. Where do industrial parks fit within this broad project, and how can we use the theoretical approaches outlined to ensure that they are playing an effective role? To begin with, there needs to be a homology between the economic structures, the IP framework, and the function and nature of industrial parks and other related IPOs. Too often, industrial parks and other IPOs have little to do with the existing structure of the economy or are vestiges of earlier eras. Developmental states are facilitators ensuring sufficient resources for structures and frameworks to operate effectively.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrial Policy, Institutional Transformation 109 Closely related to this is the evolutionary consonance between industrial parks and other related configurations of IPOs. For example, if there is interest in trying to develop an industrial park specializing in electronic assembly, the government could start an IPO dedi cated to locating investors, with offers of financial support and other incentives, organize a training institute for workers and managers, and develop engineering extension services to provide technical services for all aspects of the construction and operation of plants, etc. In the case of the Indian textile parks discussed above, we have seen that the government has done little or nothing to put in place IPOs that could make the parks more effective. Over time n ew IPOs will be needed to deal with structural gaps as parks evolve. A structural gap is one which limits the ability of a country to reach its full developmental potential within the existing structure (static) or to move to a new economic structure (dynamic). A static gap might be one where the producers in the park have not found sufficient markets for their production. One could conceive of the organization of an IPO promoting products and parks overseas. Clearly this could be a great addition to the rubric of industrial policy organizations in India that could make the textile industrial parks more prosperous and developmental. The country has a Cotton Textile Promotion Council (Texprocil), a non-profit entity controlled by individual textile producers, which might help some individual companies in zones that are members. However, as Wazir Advisors (2016: 51) notes, contrary to international experi ences of successful parks, there is no organized government group promoting the park or its products. In dynamic terms, the country will be looking to increase the depth of the economy by specializing in one particular type of assembly or improving the integration of the economy by capturing more localization of the value chain. Indian textile parks production is poorly integrated, often with one or two unconnected levels of the value chain. International experiences have pointed to a strategy of providing incentives to support entire value chains around single products like socks, and a full package of financial services, utilities, logistics, and so on (Wazir Advisors 2016: 51). The effectiveness of the developmental state is not simply measured by the exercise of power but also in the mobilization of coalitions that support industrial policy and the trans formation of industry along with the capacity to demobilize vested interests opposed to industrial change. Coalitions need to be continually altered and reconfigured to deal with the shifting nature of the structure of industry. Industries positively affected by the presence of parks should be organized and drawn into the process of coalition building, including working closely with political parties to help lay the building blocks of a platform for designing, constructing, and maintaining industrial parks. Embedded autonomy also seems particularly applicable to the functioning of successful industrial parks where the state needs a clear vision, and the tools to ensure that vision is implemented with continual feedback from the members of the park to adjust the vision. Parks should have formal councils representing tenants to act as intermediaries of flows from participants through the authorities up to ministerial level and back down. These councils could also elect national representatives to be on ministerial-level industrial policy coordinating units, which in an effective industrial policy framework should work with the highest office of the country (prime minister or president). Parks should also be embedded in local communities which are often alienated from the central state that imposes them. Local community resistance has impeded the operation and expansion of parks by resisting land acquisition (Adunbi and Stein 2019). One approach is to involve local government,
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
110 Howard Stein residents, and businesses in all aspects of the planning and execution stage of parks. Regular consultation and feedback was institutionalized from the beginning in 2010 in Thailand, with considerable success. This is in contrast with the earlier Map Ta Phut Industrial Estate which started in 1990 and which had to suspend operations due to tensions with the local community (Global Green Growth Institute 2017). From the capabilities literature, a key concept for industrial policy is learning to recognize the existence of capability traps, their causes, consequences, and the barriers to climbing the industrial ladder due to the lack or wrong mix of education, training and skills, and technological learning. A key focus of industrial policy is to generate productive transformation which is aimed at diversifying into new products, supply-chain and value-added activities, upgrading technology through imports or local investment, and generating employment and productivity increases in order to sustain a rising standard of living. New thematic industrial parks can be used to help overcome capability traps by providing focal points, product innovation, and technological transformation and upgrading. The Chinese were able to do this with the development of their science and technology parks (Stein 2012). Industrial policy also needs to address both the conceptual and procedural dimensions of the knowledge structure associated with a society’s collective capabilities. This means not only knowing how to perform tasks related to production but also the ability to structure and interpret data and information to make informed decisions. Over time, they are acquired and embedded in the collective memory of labour forces and form the bases of knowledge communities. Industrial policy should identify, categorize, catalogue, and deliberately expand these communities which can be drawn on for the formation of new types of production in line with the transformation of economic structures. A greater depth and diversity of these knowledge communities creates greater flexibility in the selection of the routes towards industrial transformation and the kinds of industrial parks that can succeed. At the heart of industrial policy are the exigencies of coordination and unity of purpose. A state can be seen as a set of interlocking organizations and an IP framework as a substratum of the state: both are connected to the state and have their own operating principles. Industrial parks and IPOs more generally are both entities of containment with their own internal dynamics fully able to shape and enhance procedural and conceptual capabilities, but they should also be seen as institutionally connected to other entities through organizational fields where correlative behaviour arises from common meaning and purpose. In organizational field terms, an IP framework can be conceived as a broad network of government ministries, industrial policy organizations, and private-sector firms connected by a common set of constructs aimed at a coordinated effort to reproduce and expand industrial production. Industrial parks should be seen as organizational units that are concatenated to other members of the IP organizational field. Understanding the linkages of industrial parks to other component parts of the field including other industrial hubs should be a central part of the planning and evolution of industrial parks. As IP frameworks are constructed it becomes central to understand both the norms, rules, and beliefs inherent in each individual IPO and the impact that it has in fulfilling its goals and functions as well as the principles embedded in the operation of related IPOs and the organizational field(s) itself. Industrial policy and IPOs like industrial park
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrial Policy, Institutional Transformation 111 authorities have often performed poorly particularly in African countries (see Chapter 50); understanding the reasons for this outcome is central to learning from these experiences to avoid the same pitfalls. The problem with parks might have little to do with the parks themselves but more with the nature of their linkages to other key components of the field. We can think of an institutional failure in an IPO framework or field as the systemic inability to reach the goals set within the field—the common purpose(s). We can contrast this with an institutional imperfection where, say, a component part is unable to perform effectively but is still able to act in a coordinated fashion towards some field goals and functions. The notion of failures and imperfections can be applied not only to the field itself but to the individual component parts of the field, such as IPO actors, ministerial units, or industrial companies, which can contribute to the field’s failures and imperfections. There are a multitude of other potential sources of institutional failures and imperfections which need to be identified and transcended if IPOs are to lead to the transformation conceptualized by the structuralist approach. Along these lines it is possible to divide industrial policy failures between a more static or momentary analysis of the operations of existing IP organizations like industrial park authorities, and the more dynamic perspective on their failure to evolve. There are a series of intrinsic and extrinsic sources of institutional failure in the industrial policy framework. Industrial policy organizations should be focused on the formulation of goals, the development of strategies, the implementation of strategies, the monitoring of processes, and the evaluation of outcomes relative to original goals. For example, failures can occur in the breakdown of any of these components and an inability to respond in an organic manner to deal with errors and make adjustments to reach these goals. Problems can arise from the scope, design, and authority of industrial policy organizations, and from their incentives and operating rules, and can include a host of connective failures that can impede the operation of IPOs. In particular, outside regulatory structures and rules can have a dramatic effect. There can be poor communication or overlapping authority with other IPOs. In addition, one may witness a breakdown in information flows between the overseeing ministerial and other authorities or with industrial companies. Ministries such as Finance have intervened with neoliberal policies that have deliberately undercut the realm of industrial policy by curtailing the effectiveness of interventions such as providing sufficient finance or reducing tax subsidies needed to assist with industrial parks. In dynamic terms, as one set of industrial challenges is overcome, new ones will be created or additional ones will need to be tackled. In preparation for structurally transforming industry, the government needs a conscious expansion of new knowledge communities that can be drawn upon to expand the capabilities of multiple players inside existing fields or in order to create new sub-fields. Too often we have seen failed IPOs, including industrial parks, operate atomistically with limited connections to broader national industrial priorities, and hence contribute to industrial policy failures. Arguably this is precisely what happened to the textile parks discussed above which were allowed to flounder as independent entities with little connection, direction, incentives, or assistance from the government beyond the initial subsidization of construction costs.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
112 Howard Stein
6.6 Conclusions The widespread recognition of the importance of the growth and expansion of industrialization as a mechanism of raising the standard of living for developing countries has led to debates around the best strategies to stimulate the manufacturing sector, including the use of industrial hubs. This chapter draws on a variety of theoretical tools to better understand the performance of industrial parks and the conditions needed for parks to increase their impact on development. The chapter critically evaluates a World Bank case study of textile parks in India which is seen by its authors as an affirmation of their largely neoliberal principles whereby parks have minimal state support and are constructed and operated by the private sector. In contrast to their claims that the textile parks have been successful, more independent evalu ations point to very disappointing results. One of the key problems of the parks was the lack of supporting industrial policy-related interventions. The chapter has also examined different ways of conceptualizing an industrial policy framework that might make industrial parks more effective. Four different theoretical traditions can be used to better conceptualize industrial policy: structuralist approaches, the development state, capabilities theory, and original institutional economic theory. The final section draws on these insights to both conceptualize the nature of industrial parks and better understand how they can be successfully integrated into the industrial policy strat egies of developing countries.
References Adunbi, Omolade and Howard Stein (2019) ‘The Political Economy of China’s Investment in Africa: Prometheus or Leviathan’, in Arkebe Oqubay and Justin Yifu Lin (eds) China–Africa and an Economic Transformation. Oxford: Oxford University Press, pp. 192–215. Amsden, Alice (1989) Asia’s Next Giant: South Korea and Late Industrialization. New York: Oxford University Press. Bromley, Daniel (2007) ‘Environmental Regulations and the Problems of Sustainability, Moving Beyond “Market Failures” ’, Ecological Economics 3(4): 676–83. Bush, Paul (1987) ‘The Theory of Institutional Change’, Journal of Economic Issues 21(3): 1075–116. Cadestin, Charles, Koen De Backer, Isabelle Desnoyers-James, Sabastien Miroudot, Ming Ye, and Davide Rigo (2018) ‘Multinational Enterprises and Global Value Chains: New Insights on the Trade–Investment Nexus’, OECS Science, Technology and Industry Working Papers No. 2018/05. Chang, Ha-Joon (2004) ‘Institutional Foundations for Effective Design and Implementation of Selective Trade and Industrial Policies in the Least Developed Countries: Theory and Evidence’, in Charles Soludo, Osita Ogbu, and Ha-Joon Chang (eds) The Politics of Trade and Industrial Policy in Africa: Forced Consensus. Trenton, NJ and Asmara, Eritrea: Africa World Press, pp. 305–28. Chang, Ha-Joon and Antonio Andreoni (2016) ‘Industrial Policy in a Changing World: Basic Principles, Issues and New Challenges’, Cambridge Journal of Economics, 40 Years Conference, July. Evans, Peter. B. (1995) Embedded Autonomy: States and Industrial Transformation. Princeton, NJ: Princeton University Press. Financial Express (2017) Government Reviewing Scheme for Integrated Textile Parks Posts Violation of Norms by Some SPVs. Available at https://www.financialexpress.com/economy/governmentreviewing-scheme-for-integrated-textile-parks-post-violation-of-norms-by-some-spvs/575576/.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrial Policy, Institutional Transformation 113 Global Green Growth Institute (2017) Thailand Community-Based Eco Industrial Town Development. Available at www.greengrowthknowledge.org. Hausmann, Ricardo and Dani Rodrik (2006) ‘Doomed to Choose: Industrial Policy as Predicament’, John F. Kennedy School of Government. Johnson, Chalmers (1982) MITI and the Japanese Miracle: The Growth of Industrial Policy: 1925–1975. Stanford, CA: Stanford University Press. Lall, Sanjaya (2013) ‘Reinventing Industrial Strategy: The Role of Government Policy in Building Industrial Competitiveness’, Annals of Economics and Finance 14(2): 785–829. Nübler, Irmgard (2014) ‘A Theory of Capabilities for Productive Transformation: Learning to Catch Up’, in José Salazar-Xirinachs, Irmgard Nübler, and Richard Kozul-Wright (eds) Introduction to Transforming Economies: Making Industrial Policy Work for Growth, Jobs and Development. Geneva: International Labour Organization, pp. 113–49. Pack, Howard and Kamal Saggi (2006) ‘The Case for Industrial Policy: A Critical Survey’. World Bank Policy Research Working Paper No. 3839. World Bank, Washington, DC. Pack, Howard and Larry E. Westphal (1986) ‘Industrial Strategy and Technological Change: Theory versus Reality’, Journal of Development Economics 22(1): 87–128. Rodrik, Dani (2004) ‘Industrial Policy for the Twenty-First Century’. CEPR Discussion Paper No. 4767. Available at SSRN: http://ssrn.com/abstract=666808. Salazar-Xirinachs, José, Irmgard Nübler, and Richard Kozul-Wright (eds) (2014) Introduction to Transforming Economies: Making Industrial Policy Work for Growth, Jobs and Development. Geneva: International Labour Organization. Saleman, Yannick and Luke Jordan (2014) ‘The Implementation of Industrial Parks: Some Lessons Learned in India’. Policy Working Paper No. WPS6799. World Bank, Washington, DC. Sawyer, Malcolm (2000) ‘The Theoretical Analysis of Industrial Policy’, in Wolfram Elsner (ed.) Industrial Policies after 2000. Amsterdam: Springer, pp. 23–57 . Scott, W. Richard and John W. Meyer (eds) (1994) Institutional Environments and Organizations. Thousand Oaks, CA: Sage. Shleifer, Andrei and Robert W. Vishny (1994) ‘Politicians and Firms’, Quarterly Journal of Economics 109: 995–1025. Stein, Howard (2008) Beyond the World Bank Agenda: An Institutional Approach to Development. Chicago, IL: University of Chicago Press. Stein, Howard (2012) ‘Africa, Industrial Policy, and Export Processing Zones: Lessons from Asia’, in Akbar Noman, Kwesi Botchway, Howard Stein, and Joseph E. Stiglitz (eds) Good Growth and Governance in Africa: Rethinking Development Strategies. Oxford: Oxford University Press, pp. 322–44. Stein, Howard, Bethuel Kinuthia, and Adam Elhiraika (forthcoming) ‘Institutions and Industrial Policy in Africa: Toward Structural Transformation and Inclusion’, contracted with Oxford University Press. Tuscor Lloyds (2017) The Industrial History of Trafford Park. Available at https://www.tuscorlloyds. com/industrial-history-trafford-park/. UNECA (United Nations Economic Commission for Africa) (2014) Economic Report on Africa: Dynamic Industrial Policy in Africa. Addis Ababa, Ethiopia: UNECA. Wade, Robert H. (1990) Governing the Market. Princeton, NJ: Princeton University Press. Wade, Robert H. (2015) ‘The Role of Industrial Policy in Developing Countries’, in Alfredo Calcagno, Sebastian Dullien, Alejandro Márquez-Velázquez, Nicolas Maystre, and Jan Priewe (eds) Rethinking Development Policies after the Financial Crisis, Vol 1: Making the Case for Policy Space. Geneva: UNCTAD. White, Lawrence J. (2008) ‘Antitrust Policy and Industrial Policy: A View from the US’. Reg Markets Centre, AEI Centre for Regulatory and Market Studies, Working Paper 08-04, January. Whitfield, Lindsay, Ole Therkildsen, Lars Buur, and Anne Mette Kjær (2015) The Politics of African Industrial Policy: A Comparative Perspective. Cambridge: Cambridge University Press. Wazir Advisors (2016) Review of the Scheme for Integrated Textile Parks (SITP) with a View to Gauge the Impact on Production, Employment and Success in Intended Objective of Clusterisation. Report to the Ministry of Textiles, India, December.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
chapter 7
I n dustr i a l H u bs The Viewpoints of Economic Geography and Empirical Economics Hisaki Kono
7.1 Introduction Industrial clusters are formed to exploit the value of proximity, or agglomeration economies. Successful industrial clusters also need good market access, as the products should eventually be transported outside the cluster to make sufficient profit. Hence the existence of agglomeration economies and the improvement of transportation systems have played a key role in the formation of industrial clusters. Transportation costs have declined substantially in the last one hundred years (Redding and Turner 2015), and this is closely linked to the emergence of large cities and industrial clusters. This chapter provides a framework for understanding the mechanism of agglomeration and formation of industrial clusters, based on models of economic geography and spatial economics. Methodological issues in empirical economics are also illustrated, as the relevance of any models should be checked empirically. Empirical studies of agglomeration economies are reviewed, together with the effects of industrial zones and infrastructure improvement. This chapter also aims at dispelling the misconception about economics amongst noneconomists. Many non-economists criticize ‘mainstream economics’ based on benchmark economic models written in undergraduate textbooks, without knowing the applied ‘mainstream economics’ and its development. Economics has a wide range of models that describe specific aspects of real economies and is useful for policy discussion.1 Further, many empirical economics studies do not actually assume any economic models at all, and statistically investigate the causal impact of policies, programmes, and interventions. These studies intend to provide credible empirical evidence of policy effectiveness, which is expected to help policymakers judge which policies they should implement to achieve a certain goal. This chapter provides a practical guide for policymakers to distinguish reliable empirical studies from unreliable ones. 1 Most of the cited papers are published in top economics journals.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrial Hubs 115
7.2 Economic Geography and Spatial Economics Theories 7.2.1 The Driving Force of Agglomeration Alfred Marshall (1890) pointed out the following factors driving the agglomeration of economic activities:
1) Scale economies (increasing returns to scale); 2) Availability of specialized input goods/services in the region; 3) Availability of a specialized labour force in the region; 4) Flow of new ideas and valuable information, especially through face-to-face communication, in the region; and 5) Existence of infrastructure and public goods. The last four factors—(2) to (5)—are also called Marshallian externalities. Scale economies, Marshallian externalities, and the reduction of transportation costs have together played a crucial role in city and industrial cluster formation.2 Economic geography models consider the location choices of firms and workers to investigate the formation of cities and agglomeration of economic activities. The entry of firms or migration of workers into a particular region have different effects on the local economy, categorized as forward linkage effects and backward linkage effects. Forward linkage effects operate through the increased supply of goods and workers. The entry of firms increases the quantity and variety of the intermediate and final goods in that region, which in turn attracts firms that use the intermediate goods for production, and workers who enjoy a wide variety of consumption goods. Worker migration increases the pool of labour forces available in that region, and the larger pool of skilled labour attracts firms seeking workers with skills that match their production technology. This process is described on the right side of Figure 7.1. Backward linkage effects refer to the effects of the increased demand caused by the entry of firms and migration of workers. As shown on the left side of Figure 7.1, the entry of firms increases the demand for input goods and workers. The higher demand for input goods attracts firms supplying the input goods, and the increased demand for workers results in higher wages, which attract more workers. The migration of workers increases the demand for final goods, which in turn attracts firms producing those goods. The forward linkage effects and backward linkage effects reinforce each other. The forward linkage effects attract firms and workers, which in turn creates additional demand for input 2 Duranton and Puga (2004) categorized the agglomeration effects as (1) sharing, (2) matching, and (3) learning. Sharing effects include the gains from a wider variety of inputs and industrial specialization, the common use of local public goods, and the diversification of risk by having multiple buyers and suppliers. The matching effects include an improvement in either the quality or the quantity of matches between firms and workers, or between suppliers and buyers. The learning effects include the generation, diffusion, and accumulation of knowledge.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
116 Hisaki Kono Increasing demand for Attract firms
Intermediate goods
Increasing supply of Entry of firms
Intermediate goods
Workers Attract workers
Final goods
Attract firms
Final goods Worker migration
Backward linkage effects
Attract workers
Workers Forward linkage effects
Figure 7.1 Backward linkage effects and forward linkage effects Source: Author’s elaboration.
goods, labour, and final goods, increasing the number of firms and workers in this region (backward linkage effects). These increases in firms and workers generate additional forward linkage effects. This cycle of the forward linkage effects and backward linkage effects generates a strong agglomeration force. The increases in firms and workers also accelerate the flow and creation of new ideas and valuable information, which also works as an agglomeration force. Interacting with these agglomeration forces, the reduction of transportation costs plays a crucial role in the formation of big cities and industrial clusters. To understand the under lying mechanisms, consider an economy consisting of many regions with similar but slightly different population sizes. Notice that in the process of agglomeration depicted in Figure 7.1, firms producing intermediate goods and migrating workers play a similar role in generating the backward and forward linkage effects. Hence, for simplicity, assume that workers are immobile and only firms make location choices. In this case, we only need to consider the agglomeration of firms. Firms’ production technology is characterized by scale economies,3 and they can sell their products in their home region without transportation costs but need to pay the transportation costs when selling to other regions. If not employed by firms, workers will produce agricultural goods that do not exhibit scale economies. Hereafter, firms are treated as the industrial sector, and industrial clustering refers to the clustering of firms. When the transportation costs are high, the distribution of economic activities will be almost equal. With high transportation costs, firms opt to sell products only in their home region, hence their production levels are constrained by local demand, which in turn depends on the size of the local population. Because population sizes are similar across regions, the number of firms will be almost the same across regions in equilibrium. Now consider a reduction in transportation costs. This enables firms to sell their products to other regions at lower costs, which generates the following three effects on the firm’s profit: 1) Increased profits from other regions (market expansion effect); 2) Decreased profits from home region owing to severe competition caused by the increased imports from other regions (competition effect); and
3 Usually the scale economies are modelled on the decreasing average production cost owing to the existence of fixed production costs. In such models, the firm size is constrained by the demand for the firm’s product. Fujita, Krugman, and Venables (1999) and Baldwin et al. (2003) provide a wide variety of economic geography models which are helpful for understanding the mechanisms of the agglomeration.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrial Hubs 117 3) Increased productivity owing to the availability of input goods produced in other regions (productivity effect). As the transportation costs decline, these effects become stronger. With transportation costs, firms have a cost advantage in their home region, where their market share will be the largest. Firms located in the region with the largest population size have the advantage of accessing the largest local market demand, which in turn reduces their average production costs owing to scale economies. With this cost advantage, firms in the largest region will (1) enjoy the largest market expansion effect as they can sell the products at a lower price than their rivals in other regions, and (2) face less competition effect as their prices are much lower than those of rivals in their home region. Therefore, the firms located in the largest region benefit the most from a reduction in transportation costs. The resultant increase in profits induces the entry of new firms, generating backward and forward linkage effects. The cycle of these backward and forward linkage effects creates a strong agglomeration force. In equilibrium, the region with the largest population size will have a disproportionately greater number of firms, even with only a slight difference in the population size across regions. Eventually, industrial production is concentrated in the largest regions and industrial goods from these regions are exported to other regions. The effects that industrial production is disproportionately concentrated in the region with the largest local market size are called home market effects. With home market effects, the reduction in the transportation costs magnifies the difference in the level of local economic activities. The entry of firms producing input goods also benefits the firms in that region as the latter can now access a variety of input goods without incurring additional transportation costs. Although the reduction in the transportation costs enables firms in other regions to purchase input goods from other regions (as noted above), firms in the region with various input suppliers have an advantage in procuring these input goods without any transportation costs. Hence, the productivity improvement caused by the increase in local suppliers further reinforces the agglomeration. So far, we have assumed that workers are immobile. If we allow for migration, skilled workers will migrate to the larger regions because they can easily find jobs that match their skills and offer higher wages. They can also purchase final goods produced there at lower prices. Hence, migration of skilled workers further reinforces the agglomeration. We should note that the impacts of the transportation cost reduction are heterogeneous across regions: it will increase industrial output in the regions with the largest population sizes while shrinking economic activities in other regions through the relocation and exit of firms. Hence, when we evaluate any policy that facilitates agglomeration, such as infrastructure improvement, we should look at the impact not only on the agglomerated regions but also on the regions that are adversely affected. Finally, we consider an extreme case of zero transportation costs. Then, firms can access any market without any costs, irrespective of their location. Hence, the location choice will depend only on the local production costs, especially labour costs and land rent, and not on the local market size. With the limited land supply, agglomeration simply increases the production costs, and in equilibrium, the distribution of economic activities will be equalized. This implies that the relationship between transportation costs and degree of agglomeration is not monotonic.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
118 Hisaki Kono In this way, Marshallian externalities and transportation costs play crucial roles in the formation of cities and industrial clusters. These externalities are often classified as localization economies and urbanization economies. Localization economies are defined as the benefits generated by the proximity of firms in the same industries or related industries. The localization economies include lower transaction costs, concentration of specialized input goods/services and a specialized labour force, and frequent information exchange. They are considered the main factors explaining industrial clustering. Urbanization economies are defined by all the merits generated from the agglomeration of overall industrial activities, such as lower costs for transaction and communication across industries, availability of higher-quality inputs and labour force, and local amenities and public goods. The localization economies increase the productivity of specific industries in the given region, whereas the urbanization economies improve the productivity of overall industries in that region.
7.2.2 Economic Development and Industrial Clustering Economic geography models also help us understand how economic development and industrial clustering are interconnected. Applying the standard economic geography framework to multi-region and multi-sector cases with exogenous labour productivity growth, Fujita, Krugman, and Venables (1999: ch. 15) simulate the pattern of economic development and the sequence of industrial clustering. Hence, workers are assumed to be immobile while firms are mobile, and firms use outputs of other firms as their inputs. Figure 7.2 describes the simulated evolution of the relative real wages in three leading regions (ω1 / ω , ω2 / ω , and ω3 / ω ) in their baseline five-region and seven-sector model, where the horizontal axis represents labour productivity, which is common to all the regions.4 As the agglomeration increases the real wage level, the relative real wage captures the distribution of economic activities across regions. We start from the situation where all the industries are concentrated in one region, say region 1 (left of the graph). Due to the higher labour demand in region 1, its real wage is higher than other regions, as expressed by its relative real wage ω1 / ω being higher than 1. The increase in labour productivity translates to the increase in labour income, which generates a greater demand for industrial goods, as suggested in Engel’s law. This increased demand for industrial goods further expands industrial production in region 1, which generates backward linkage effects (greater demand for inputs and labour) and forward linkage effects (greater supply of inputs and final goods), reinforcing the agglomeration of economic activities in region 1. The real wage gap between region 1 and other regions gets larger in this phase (Phase I in Figure 7.2). However, the increase in wages in region 1 makes production in this region expensive. Once the wage exceeds a certain point, the most labour-intensive sector relocates to another region, say region 2. This corresponds to the beginning of Phase II in Figure 7.2. The relocation of the most labour-intensive sector to region 2 generates backward and forward linkage effects, inducing relocation of the second most labour-intensive sector. This further reinforces the agglomeration and relocation from region 1, and eventually the real wage level 4 It is assumed that five regions have the identical technology, preferences, and endowments, and seven sectors are identical except for their labour intensity.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrial Hubs 119 Relative real wages
ω1/ω¯ 1.1
1.0 ω2/ω¯ ω3/ω¯ 0.9 I
II
III
IV L
Figure 7.2 Economic development and spread of industrial clusters Note: The Roman numerals at the bottom indicate the phase of the development described in the main text. Source: Masahisa Fujita, Paul Krugman, and Anthony J. Venables, The Spatial Economy, figure 15.3, © 1999 Massachusetts Institute of Technology, by permission of The MIT Press.
converges with that in region 1. In this process, economic development in region 1 increases production costs in region 1, triggering industrialization in region 2. However, other regions are yet to start industrialization. Actually, at the beginning of Phase II, many regions had a chance to industrialize as they had the same economic structure, but the backward and forward linkage effects amplified the small advantage of one region (region 2 in this case), and other regions were left behind. Phase III starts when region 2 perfectly catches up with region 1. As labour productivity increases, these two regions experience an increase in the real wage until it becomes so expensive that the most labour-intensive sector relocates to region 4, which initiates Phase IV when region 3 catches up with regions 1 and 2. This indicates that the process of industrialization is not uniform across regions. It starts in the largest city, followed by subsequent development of other areas. Increasing production costs in the existing industrial clusters are the driving force behind formation of new industrial clusters. The model predicts that the most labour-intensive sector is the first industry to relocate, which is consistent with the empirical patterns in many Asian and African countries, where labour-intensive textile and light industries are the precursors to industrial development. The entry of labour-intensive industries paves the way for the entry of other industries by creating backward and forward linkage effects. In general, the relocation decision depends on the degree of the backward and forward linkages that the firm receives from the industrial clusters. For example, the final goods sector will receive little backward linkage effect from industries other than the intermediate goods sector. While the demand for intermediate goods is concentrated in the industrial cluster, the demand for final goods exists in every region, making them the first to relocate when production costs increase. Industries that depend little on intermediate goods produced in the cluster are also quick to relocate when production costs increase for the same reason.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
120 Hisaki Kono It is also implied that industries with strong input–output linkages have a strong agglom eration force. Such hub industries will attract other firms through backward and forward linkage effects, and the existence of production networks there makes them reluctant to relocate to other places. Many successful industrial zones (IZs) have in fact attracted hub firms characterized by strong backward and forward linkage effects.
7.2.3 History and Expectations The strong backward and forward linkage effects magnify slight differences in the initial economic conditions. After industrial clusters are formed, individual firms have little incentive to relocate until production costs become too high, as shown in Figure 7.2. Further, relocation of their establishments incurs large sunk costs, which makes firms hesitant to relocate from their existing location. Hence, once industrial clusters are formed, firms tend to be locked in in that region, and a moderate change in economic conditions in other regions will not change the spatial distribution of firms. The lock-in effect implies that the distribution of economic activities in a given economy is regulated by its historical path and that changing it will require a drastic change in the economic conditions across the region. The strong backward and forward linkage effects also imply that the firm’s optimal location choice will crucially depend on where other firms will be located. This creates a coordination problem, and expectations matter a lot. Hence, the strong commitment by governments to attract related industries to that region will play a crucial role in cluster formation.
7.3 Methodological Issues in Empirical Studies on Policy Evaluation We have briefly introduced implications from economic geography models, focusing on the cause and process of industrial cluster formation. However, theories in themselves do not ensure success or failure of policies and interventions a priori as they rely on a set of assumptions, and the validity of the models should be empirically tested. To identify effective policies for economic development and industrial hub formation, careful empirical investigation of existing policy experiences is required. This section introduces some important concepts in empirical analysis. Many existing policy evaluations are flawed because they fail to take account of these concepts. Policymakers need the knowledge to discern which evaluations are empirically valid and trustworthy rather than unconditionally trusting the results of existing policy evaluations.
7.3.1 Causal Inference Framework A key to policy evaluation is the understanding of the counterfactual, irrespective of whether one uses a quantitative or qualitative approach. The counterfactual is the state of the world that is contrary to the fact. For example, suppose we are interested in estimating the impact
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrial Hubs 121 of IZs on local employment. Then, the counterfactual is the employment level that would have been realized if the IZs had not been established. Because the IZs were actually established, we cannot observe the counterfactual employment level. Although not observed, the counterfactual plays a key role in defining policy impact. The impact is actually defined by the difference between the state of the actual world (fact) and the counterfactual:
Impact = Fact − Counterfactual
Because the counterfactual is the state of the world that would have been realized if the policy had not been implemented, the difference between the fact and the counterfactual gives us the causal impact of the policy. This implies that evaluating the policy impact requires a valid construction of the counterfactual. In the qualitative approach, researchers should pin down what the outcome would have been if the policy had not been implemented (counterfactual), based on various sources of local information, including statistics, related documents, and interviews with local stakeholders. However, the construction of the counterfactual will be quite subjective. Even with the same sets of information, different researchers will construct different counterfactuals, resulting in different conclusions. Faced with the difficulty of judging which counterfactual would be more plausible, policymakers tend to choose the analysis that sounds good to them. In the quantitative approach, researchers typically construct the counterfactual using observations that were not influenced by the policy. In this case, different researchers can construct the same counterfactuals if they use the same quantitative methods. However, the validity of the constructed counterfactual depends, crucially, on the methods used, and many policy evaluations are invalidated by the use of invalid methods. The following section focuses on the quantitative approach, and explains several issues that are important for understanding the validity of a constructed counterfactual.5
7.3.2 Selection Bias As an example, consider the impact of IZs established in year T0 on local employment. In typical evaluation studies, researchers compare either: 1) the level (or trend) of employment before and after year T0 when the IZs are established, or 2) the level (or trend) of employment in the regions with IZs and in the regions without IZs. While case (1) uses the employment before year T0 as the counterfactual for the employment after year T0 , case (2) uses the employment in the regions without IZs as the counterfactual. Sometimes researchers compare the level (or trend) of employment in IZs with national or regional employment levels, in which case they use the national or regional employment levels as the counterfactual. 5 For a thorough treatment of causal inference, refer to Imbens and Rubin (2015) and Hernán and Robins (2019).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
122 Hisaki Kono As noted above, the validity of the analysis crucially depends on the validity of the constructed counterfactual. In case (1), the required assumption is that without the IZs, the level (or trend) of employment after year T0 would have been the same as that of employment before year T0 on average. However, it is possible that the prospect of future economic growth induced the establishment of IZs, or other national-level economic policies or reforms were implemented at the same time. In both cases, the employment level before year T0 will not serve as the valid counterfactual for the employment level after year T0 . In case (2), the required assumption is that the level (or trend) of employment in the regions without the IZs would have been the same as that of employment in the regions with IZs on average. However, IZs are often established either in advanced regions to ensure their success, or in lagged regions to alleviate regional inequality. In either case, regions with IZs are systematically different from regions without IZs, and employment in the regions without IZs cannot be a valid counterfactual. The problem is caused by the fact that the periods or regions with IZs are different from the periods or regions without IZs, hence they cannot be simply compared. The timing or the regions for establishing IZs will be carefully selected by the authorities, hence simple before/after or with/without analysis would generate biased estimates of the policy impact, and therefore would have no empirical validity. This is called selection bias, and one of the main issues in empirical economics is how to overcome selection bias by finding valid counterfactuals.
7.3.3 Relocation and Spillover In addition to selection bias, the analysis of IZs or industrial policies faces the challenges of relocation and spillover. As we have seen, firms and workers make location choices. If IZs attract workers and firms to their region from other regions, then employment and economic activity in the latter regions should shrink. Ignoring this relocation effect will overstate the impact on social welfare. In some cases, we can arrive at a false conclusion, finding a positive impact of a policy that is actually harming the aggregate economy. To clarify this, consider an economy consisting of four regions, each of which has fifty immobile workers, fifty mobile workers, and ten firms. The distribution of the workers and firms is described in the left panel of Figure 7.3. For simplicity, assume that the mobile workers and firms will be located in the region where the IZs are established, and that the IZs have no aggregate impact, that is, neither the total numbers of workers nor firms in the economy are fixed, irrespective of the establishment of IZs. Suppose that the IZ is established in region 1. Then, all the mobile workers and firms are now located in region 1, as described in case (a) in the right panel of Figure 7.3. Hence, we would observe: • 250 workers (fifty immobile workers and 200 mobile workers) and forty firms in the region with the IZ, and • fifty workers and zero firms in each region without the IZ. If we compare the region with the IZ and the regions without the IZ after the IZ was established, we would conclude that the IZ generated 200 more jobs and forty more firm
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrial Hubs 123 (a) Industrial zones in region 1 Region Immobile Mobile Firms workers workers 1 50 50 10 2 50 50 10 3 50 50 10 4 50 50 10
Region Immobile Mobile Firms workers workers 1 50 200 40 2 50 0 0 3 50 0 0 4 50 0 0 (b) Industrial zones in regions 1 and 2 Region Immobile Mobile Firms workers workers 1 50 100 20 2 50 100 20 3 50 0 0 4 50 0 0
Figure 7.3 Illustrative example of relocation effects and biases in empirical studies Source: Author’s elaboration.
entries. Hence, establishing IZs is beneficial to the economy. However, as we have assumed, the IZ has no aggregate impact and simply changes the location of economic activities. Ignoring relocation guides us to a false conclusion. Alternatively, if we compare the economic activity before and after the establishment of the IZ in region 1, we observe: • 100 workers and ten firms before the establishment of the IZ, and • 250 workers and forty firms after the establishment of the IZ. As we observe 150 more jobs and thirty more firms after the establishment of the IZs, we would again conclude that establishing an IZ is beneficial, which would be wrong. We could also look at the difference between the region with IZ and the regions without IZ before and after the establishment of the IZ,6 observing: • 150 more workers and thirty more firms in the IZ region, and • fifty less workers and ten less firms in each region without IZs. Again, one would conclude that establishing the IZ generated 200 more jobs (= 150 − (−50)) and forty more firms (= 30 − (−10)) . These simple exercises clearly show the danger of ignoring the possibility of relocation. Any policy evaluation on industrial hubs should seriously consider the issue of relocation. Note also that with relocation, establishing fewer IZs will result in a greater measured impact than establishing more IZs. For simplicity, assume again that IZs have no aggregate impact. Instead of establishing an IZ in region 1, consider the policy that establishes IZs in regions 1 and 2, as described in case (b) in the right panel of Figure 7.3. Suppose that half 6 Comparing the difference between the regions with IZs and without IZs in the difference of the outcome between the pre- and post-periods of the IZ establishment is called the differences-in-differences approach.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
124 Hisaki Kono of the mobile workers and firms are located in region 1 and the other half in region 2. Then, we would observe:
• 150 workers (fifty immobile workers and 100 mobile workers) and twenty firms in each of the regions with the IZs, and • fifty workers and zero firms in each of the regions without IZs. Now, the difference between the regions with the IZs and the regions without the IZs is 100 for employment and twenty for the number of firms, which is half of those in case (a). Even though the IZs have no aggregate effects and therefore the aggregate impact is the same in cases (a) and (b), one would find a higher impact of IZs in case (b), where the number of IZs is smaller. This can mislead the policy discussion. Even when establishing more IZs has more aggregate impact, any analysis that ignores relocation would conclude that establishing fewer IZs is better. These examples are quantitative, but the same argument equally applies to qualitative studies. Without considering the possibility of relocation, merely comparing regions with IZs and regions without IZs, comparing industries with policy interventions and other industries without them, or looking at the trend of economic activities before and after the policy would produce misleading conclusions. The same argument holds true for evaluation of industrial policies. Supporting one industry will help that industry, but it can harm another industry and reduce aggregate welfare. Success in boosting a certain industry does not necessarily mean that the policy has benefited the economy. Without considering the potential relocation effects, one cannot provide plausible policy suggestions. To allow for relocation effects, one needs to know which regions would not be affected at all by the IZs through relocation. However, such regions are often quite far away from those with IZs, and their economic conditions could be quite different. Therefore, those regions can hardly be the appropriate counterfactuals for the regions with IZs. Some recent economic studies utilize economic geography models to incorporate the relocation effects, which we will address below.
7.3.4 Sample Selection Bias Another issue in empirical studies is the danger of focusing on successful examples only. Investigating only successful examples will not help us identify the cause of the success. Suppose a country established 100 IZs, and one IZ, say Penang, was a success. If one looked at Penang, then one would conclude that IZs are quite effective. However, to evaluate the effectiveness of IZs, we should also look at the remaining ninety-nine IZs. Even if Penang succeeds, it is possible that the remaining ninety-nine IZs have failed. Further, suppose one investigates the case of Penang and concludes that the governance of IZs is important. However, it is possible that the remaining ninety-nine IZs implemented the same governance policies, only to fail. Analyses focusing only on success or failure cases are inappropriate and often misleading. To summarize, when evaluating the effectiveness of IZs or identifying important factors for policy success, one should collect a sample that completely represents the whole economy.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrial Hubs 125
7.3.5 External Validity When evaluating policies, we always consider two types of validity: internal and external. Internal validity is the validity of the empirical analysis conducted for a given society; the argument above is all about internal validity. External validity concerns the generalizability of the empirical results to other settings. For example, Singapore is often cited as a successful example of industrial policies, backed by comprehensive analyses of key industries by public and private evaluators. But such analyses will not be possible for most other countries, especially when the country is quite large, consists of geographically distant and heterogeneous regions, and has a wide variety of industrial sectors for consideration. Depending on the economic and social circumstances, the same policy can result in different outcomes, and policymakers should consider such differences while learning from the empirical results obtained from other countries. To understand the degree of external validity and to consider the applicability of the results to a particular country, knowledge of economic systems and incentive structures is helpful. At least, one should evaluate: • The degree of change in incentives and economic conditions caused by the policy, • The level of enforcement of the policy, and • Constraints in the economy. Even with the seemingly same policies, the degree of change caused by them will be different, depending on the initial condition or the existing economic structure. Further, the effective change will be different depending on how strictly the policy was implemented. The policy will also have different impacts depending on the constraints and bottlenecks faced by the economy, such as input availability, quality of the labour force, and level of transportation costs. Economic theories will help understand how and which differences are important for the policy purpose. Further, theories of economic geography indicate that market failures should exist for industrial policies or any interventions to improve welfare. Hence any policy discussion should first identify major market failures preventing economic development in that country. Whether the industrial policies or the proposed interventions are in fact the best policy options should then be discussed. For example, if the main market failure is the malfunctioning of the credit market, then direct intervention in the credit market might be preferable. If coordination failures amongst investors are the problem, then the government can simply act as a coordinator instead of investing millions of dollars in establishing IZs. This argument suggests that even if research shows positive impacts for some policy interventions in other countries, whether those policy interventions are necessary and the best option for that country should be explored.
7.3.6 Reduced-form Approach and Structural Approach There are two approaches in empirical economics: reduced-form approach and structural approach. In the reduced-form approach, we statistically investigate the causal effect of some variable, say policy, on outcome variables. In this approach, we do not assume any
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
126 Hisaki Kono economic models and we simply focus on the statistical causal relationship. The validity of the empirical results does not rely on any economic models, and its validity should be judged from the viewpoint we have mentioned in this section. In the structural approach, we impose a particular economic model, and estimate the parameters of that model to simulate the impact of some counterfactual policies, or to quantify the magnitude of each channel described in the economic model. This approach is especially useful for estimating relocation effects and computing welfare effects. However, this approach relies heavily on the economic model, and the validity of the analysis crucially depends on the plausibility of the economic model.
7.4 Empirical Evidence on Economic Agglomeration and Policy Effectiveness 7.4.1 Agglomeration Economies First, consider the evidence in support of agglomeration economies. Ellison et al. (2010) found support for Marshallian externalities when they analysed the agglomeration pattern of US manufacturing sectors, finding that the input–output linkage, availability of specialized labour force, and flow of ideas and information causes agglomeration, with the input– output linkage having the greatest effect. Combes et al. (2008) found that higher-skilled workers are more often located in denser cities. Using French establishment-level data, Combes et al. (2012) showed that agglomeration improves the firm’s productivity, and more efficient firms benefit more from the agglomeration. Generally existing empirical literature supports the existence of agglomeration economies, ensuring the validity of the theoretical models we have discussed. Combes and Gobillon (2015) provide a detailed review of empirical studies on agglomeration economies.
7.4.2 Industrial Zones Given the existence of agglomeration economies, we now turn to the effectiveness of public policies intended to facilitate agglomeration. First we consider the impact of industrial zones and other similar interventions, such as enterprise zones and special economic zones. As we have seen, firms and workers tend to be locked into industrial clusters once they are formed, hence starting a new cluster requires sufficient change in the production costs and expectations of relevant economic agents. This implies that the success of industrial zone policies depends on the degree of backward and forward linkage effects in existing industrial clusters, the change in production costs induced by the policies, and the government’s commitment and coordination for relevant investors. Additionally, in evaluating any industrial zone policy, the relocation effects discussed above should not be ignored. Neumark and Kolko (2010) evaluated the impact of California’s enterprise zone programme on local employment. To address selection bias, they compared the change in employment in
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrial Hubs 127 the newly established areas with the employment in the surrounding areas and the pre-existing enterprise zone. They found no significant difference between these areas. Although this study does not perfectly control the relocation effects, the fact that ignoring the relocation effects overestimates the positive effects implies that the enterprise zones did not contribute to generating local employment.7 Freedman (2013) studied the Texas enterprise zone programme. He exploited the fact that enterprise zone designations are determined in part by a cut-off rule based on census block group poverty rates. Notice that areas where poverty rates are slightly below the threshold can be considered similar to areas where poverty rates are slightly above the threshold, except that the former are more likely to be designated as enterprise zones. Hence, if one finds significant differences between these two groups, then this can be attributed to the differences in the impact of the enterprise zone. He found that the Texas enterprise zone programme increased local employment, though it is still possible that the relocation effect might have simply driven this positive impact. Givord et al. (2013) investigated Zones Franches Urbaines in France and found that while the programme increased employment and business establishment in the zones, it also reduced employment and business establishment in surrounding areas, suggesting a relocation effect. Neumark and Simpson (2015) reviewed studies investigating the impact of industrial and enterprise zones, concluding that the empirical evidence on the effectiveness of these policies is mixed.
7.4.3 Infrastructure Finally, we review studies investigating the impacts of infrastructure improvement that reduces transportation costs. As we have observed, the impact of transportation cost reduction is heterogeneous (expanding large locations and contracting small locations) and nonmonotonic (inducing agglomeration when transportation costs are at intermediate levels but reducing it if transportation costs get close to zero). Chandra and Thompson (2000) investigated the effect of the interstate highway system on non-metropolitan US counties, finding a positive impact on earnings in connected counties. However, they also found a negative impact on earnings in neighbouring unconnected counties (relocation effect), and cannot reject the hypothesis that its aggregate impacts sum to zero. Kline and Moretti (2013) evaluated the long-run effect of a large-scale regional development programme in the United States, the Tennessee Valley Authority (TVA). By comparing the programme’s regions with the regions that were proposed but never approved by Congress, they found that the TVA led to large gains in manufacturing employment, and the impact continued after federal transfers had lapsed, a pattern consistent with the presence of agglomeration economies. The induced shift from agriculture to manufacturing raised aggregate income in the TVA region. To consider the relocation effect and assess the aggregate
7 There might be positive spillover effects on the surrounding areas, but they discard this possibility by comparing areas close to the newly established zones with areas further from these zones.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
128 Hisaki Kono impact, they also developed an economic geography model, finding that agglomeration gains in the TVA region were offset by losses in the rest of the country (relocation), and the TVA’s aggregate impact stemmed only from the positive direct impact of its infrastructure investment on national manufacturing productivity. Some recent literature investigates the effect of infrastructure on agricultural sectors, which will not have agglomeration economies. Donaldson (2018) studied the effect of railways in India during the period 1870–1930, finding that districts with access to the railway recorded 16 per cent higher real agricultural incomes. Further, by using an economic model of trade, he concluded that more than half of the real income effect of the railways is attributable to newly exploited gains from trade owing to trade cost reduction. Donaldson and Hornbeck (2016) examined the historical impact of railways on the US agricultural sector in 1890, showing that railway network expansion increased agricultural land values. However, Asher and Novosad’s (2018) study on India’s national rural road construction programme found no major changes in agricultural outcomes, income, or assets. They conclude that even with better market connections, remote areas may continue to lack economic opportunities.
7.5 Conclusion We have reviewed representative models of economic geography to understand the process of agglomeration, and introduced important issues in empirical work. Invalid empirical analysis will lead to a false conclusion, which will mislead policy discussion. Hence one cannot overemphasize the importance of rigorous empirical studies that overcome the problems of selection bias, sample selection bias, and relocation effects. Results of existing empirical studies on the effects of industrial zones and infrastructure improvement are rather mixed, suggesting that one cannot be unconditionally optimistic about policies that facilitate agglomeration or the formation of industrial hubs. Careful diagnosis of economic conditions and an understanding of the underlying economic structure are essential for policy discussion and policy formation.
References Asher, Samuel and Paul Novosad (2018) ‘Rural Roads and Local Economic Development’, American Economic Review, forthcoming. Available at https://www.aeaweb.org/articles?id=10.1257/ aer.20180268&&from=f. Baldwin, Richard, Rikard Forslid, Philippe Martin, Gianmarko Ottaviano, and Frederic RobertNicoud (2003) Economic Geography and Public Policy. Princeton, NJ: Princeton University Press. Chandra, Amitabh and Eric Thompson (2000) ‘Does Public Infrastructure Affect Economic Activity? Evidence from the Rural Interstate Highway System’, Regional Science and Urban Economics 30(4): 457–90. Combes, Pierre-Philippe and Laurent Gobillon (2015) ‘The Empirics of Agglomeration Economies’, in Gilles Duranton, Vernon Henderson, and William Strange (eds) Handbook of Regional and Urban Economics, Vol. 5. Amsterdam: Elsevier, pp. 247–348. Combes, Pierre-Philippe, Gilles Duranton, and Laurent Gobillon (2008) ‘Spatial Wage Disparities: Sorting Matters!’, Journal of Urban Economics 63(2): 723–42.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrial Hubs 129 Combes, Pierre-Philippe, Gilles Duranton, Laurent Gobillon, Diego Puga, and Sébastien Roux (2012) ‘The Productivity Advantages of Large Cities: Distinguishing Agglomeration from Firm Selection’, Econometrica 80(6): 2543–94. Donaldson, Dave (2018) ‘Railroads of the Raj: Estimating the Impact of Transportation Infrastructure’, American Economic Review 108(4–5): 899–934. Donaldson, Dave and Richard Hornbeck (2016) ‘Railroads and American Economic Growth: A “Market Access” Approach’, The Quarterly Journal of Economics 131(2): 799–858. Duranton, Gilles and Diego Puga (2004) ‘Micro-foundations of Urban Agglomeration Economies’, in J. Vernon Henderson and Jacques-François Thisse (eds) Handbook of Regional and Urban Economics, Vol. 4. Amsterdam: Elsevier, pp. 2063–117. Ellison, Glenn, Edward L. Glaeser, and William R. Kerr (2010) ‘What Causes Industry Agglomer ation? Evidence from Coagglomeration Patterns’, American Economic Review 100(3): 1195–213. Freedman, Matthew (2013) ‘Targeted Business Incentives and Local Labor Markets’, Journal of Human Resources 48(2): 311–44. Fujita, Masahisa, Paul R. Krugman, and Anthony J. Venables (1999) The Spatial Economy: Cities, Regions, and International Trade. Cambridge, MA: MIT Press. Givord, Pauline, Roland Rathelot, and Patrick Sillard (2013) ‘Place-based Tax Exemptions and Displacement Effects: An Evaluation of the Zones Franches Urbaines Program’, Regional Science and Urban Economics 43(1): 151–63. Hernán, Miguel A. and James M. Robins (2019) Causal Inference. Boca Raton: Chapman & Hall. Imbens, Guido W. and Donald B. Rubin (2015) Causal Inference in Statistics, Social, and Biomedical Sciences. Cambridge: Cambridge University Press. Kline, Patrick and Enrico Moretti (2013) ‘Local Economic Development, Agglomeration Economies, and the Big Push: 100 Years of Evidence from the Tennessee Valley Authority’, The Quarterly Journal of Economics 129(1): 275–331. Marshall, Alfred (1890) Principles of Economics. Neumark, David and Jed Kolko (2010) ‘Do Enterprise Zones Create Jobs? Evidence from California’s Enterprise Zone Program’, Journal of Urban Economics 68(1): 1–19. Neumark, David and Helen Simpson (2015) ‘Place-based Policies’, in Gilles Duranton, Vernon Henderson, and William Strange (eds) Handbook of Regional and Urban Economics, Vol. 5. A msterdam: Elsevier, pp. 1197–287. Redding, Stephen J. and Matthew A. Turner (2015) ‘Transportation Costs and the Spatial Organization of Economic Activity’, in Gilles Duranton, Vernon Henderson, and William Strange (eds) Handbook of Regional and Urban Economics, Vol. 5. Amsterdam: Elsevier, pp. 1339–98.
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
chapter 8
I n dustr i a l H u bs, U r ba n Systems, a n d Economic Dev el opm en t Edlam Abera Yemeru
8.1 Introduction Cities have been at the heart of economic development dynamics for millennia and remain so today. With most of the world’s population and economic activities concentrated in cities, their role in shaping development outcomes is ever more prominent. Historically, economic and urban growth has followed interconnected pathways with cities being both enablers and products of economic transformation. Industrialization and urbanization are especially mutually reinforcing transitions underpinning economic growth in most coun tries, with the onset of industrialization in the eighteenth century having taken root in cities. Industrial activity is known to gravitate towards cities given the advantages, or scale economies, present in urban areas that increase efficiency and productivity. Scale econ omies are differentiated across cities in a national urban system and so is the type of indus trial activity. A hierarchy of cities forms within an urban system where scale economies and industrial activities vary at each layer of the system. Cities and urban systems are therefore critical factors shaping the performance of industrial hubs as instruments fostering eco nomic development. At the same time, industrial hubs development significantly influences the evolution of cities as well as the overall structure of the urban system. This chapter reviews the relationship between cities and industrialization, with a specific focus on the role of national urban systems in industrial hubs development. While industri alization itself significantly shapes the growth and transformation of cities, and urban sys tems, this is addressed elsewhere in the Handbook, including through a focus on the social and human effects of industrial hubs development. This chapter starts by highlighting the historical association between urbanization and industrialization and outlines the present-day scale and magnitude of global urban dynamics. It then discusses how cities enable industrial
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
Industrial Hubs, Urban Systems, and Economic Development 131 productivity through scale economies. Subsequent sections present the differentiated ways in which cities within an urban system influence industrial activity. The policy measures taken to promote a productive urban industrial economy by some countries are highlighted. The concluding remarks recap the key reasons why cities and urban systems constitute crit ical determinants of industrial hubs development and outcomes.
8.2 Industrialization and Urbanization: A Historical Perspective Cities have been central to the evolution of human society for millennia. Their ‘generative power’ has underpinned economic, social, and political advances from the first known cit ies to have emerged 10,000‒12,000 years ago (Soja 2010; Tilly 2010). Urbanization has been associated with major developmental milestones in history including the agricultural and industrial revolutions, as well as creative and technological advances. The notion of an ‘urban revolution’ has been used to portray the city as a symbol of the evolution of the eco nomic structure and organization of human societies (Childe 1950). While less than an esti mated 1 per cent of humans lived in cities for the first 6,000 years of their existence, rising through the next 5,000 years, the major urban shift that unfolded in the eighteenth century was driven by the advent of industrial cities (Soja 2010). The Industrial Revolution of the eighteenth and nineteenth centuries, when production took a major leap towards mechanization and centralized manufacturing, happened simul taneously with an urban revolution that saw unprecedented growth in the size and eco nomic structure of cities. Industrialization during this period was both a product and a driver of urbanization. Terms such as ‘industrial urbanism’ and the ‘urban-industrial econ omy’ have been used to describe the conjoined processes of industrialization and urbanization that drove rapid economic development for much of the nineteenth century (Goheen 1973; Blumin 2006). The nineteenth century was the ‘take-off ’ period for urbanization, driven by three main factors related to industrialization (Blumin 2006). First, with expanding industrial output, labour shifted out of agriculture to manufacturing, which was predominantly located in and around urban areas. This attracted large numbers of people into urban areas who sought proximity to the ever-growing factories and production sites. Second, industrial firms chose to locate closer to one another to maximize efficiencies through access to cap ital, labour, services, and other inputs. Third, the mechanization of agriculture drove large numbers of farmers out of rural areas into cities to join the industrial labour forces. These shifts remain similar to the processes through which countries worldwide continued to industrialize and urbanize after the nineteenth century, as highlighted later in this chapter. In Britain, considered the birthplace of the Industrial Revolution, major changes to the economic and urban profile of the country unfolded from 1770 to 1840. Existing and new cities emerged as manufacturing hubs with a growing industrial workforce and total popu lation continued to increase steadily while rapidly shifting out of rural areas (Morris 1971).
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
132 Edlam Abera Yemeru Cities grew at unprecedented rates, such that by 1851 the majority of Britain’s population were urban dwellers (Stobart 2000). Differentiated growth across urban centres in the country resulted in an urban hierarchy of major manufacturing cities surrounded by emerging networks of smaller urban centres (Stobart 2000). The American industrial revolution took off with the mechanization of cotton mills on the east coast from 1790 onwards, such that by the end of the nineteenth century manufac turing accounted for more than half of the value of goods in the economy and urban dwell ers accounted for around 40 per cent of the population (Blumin 2006). High rates of urban growth accompanied the shift of the economy towards manufacturing, with a number of major and smaller urban cities emerging as hubs for industrial activity. While the initial wave of manufacturing was concentrated in larger cities of more than 100,000 people, which became centres of trade and commerce, it moved increasingly towards smaller cities with a population of 20,000 to 38,000 at the end of the nineteenth century, creating an urban system with differentiated city economies at each level (Weber 1904). Canada and other countries and regions in Western Europe also underwent rapid indus trialization and urbanization in the nineteenth century (Goheen 1973). However, most countries in other parts of the world, including Latin America, Africa, and Asia, remained largely rural during this period, though by 1960 many had undergone rapid urbanization and industrialization (e.g. Japan, Argentina, Chile) (Blumin 2006; Hoyt 1963). While urban growth gradually slowed in countries where manufacturing had first taken hold, more rapid urbanization was experienced in newer industrial economies and developing countries from 1890 to 1960 (Hoyt 1963). Urbanization and industrialization continued to advance in tandem in the twentieth century, although with some exceptions, most notably in Africa where rapid urban growth has been taking place amidst low or weak levels of manufacturing and economic growth (Bloom et al. 2008; Gollin et al. 2016). Nevertheless, urbanization is today a global phenomenon setting the conditions within which economic development takes place.
8.3 Global Urban Dynamics The twenty-first century is undoubtedly an urban one. In 1950, only 29.6 per cent of the world’s population was urban. In 2008, the global urban transition passed a major milestone with more than half of the world’s population being urban. In 2018, 55.3 per cent of global population was urban and was estimated to reach 68 per cent by 2050. While the pace of urban growth will be slower in the decades up to 2050, the absolute number of p eople living in cities will continue to rise. An additional 2.3 billion persons will live in cities between 2020 and 2050, reaching 6.68 billion persons. During this period, the global rural popula tion will decline (by 324,421,000 persons) to 3 billion by 2050 (UNDESA 2018). This is an enormous shift with far-reaching implications, especially for economic development. The urbanization process has not been a uniform one globally. Trends by world region illustrate the differentiated levels of urbanization (see Figure 8.1). Africa is currently the least urbanized (40 per cent), followed by Asia (48 per cent), with other world regions hav ing high levels of urbanization (Oceania 68.1 per cent, Europe 73.9 per cent, Latin America and Caribbean 79.9 per cent, North America 81.6 per cent). However, the least urbanized
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
Industrial Hubs, Urban Systems, and Economic Development 133 100.0
Percentage of Urban Population
90.0 80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0
1950
1960
1970
WORLD
1980 AFRICA
1990
2000
ASIA
LATIN AMERICA AND THE CARIBBEAN
2010
2020
2030
2040
2050
EUROPE NORTHERN AMERICA
OCEANIA
Figure 8.1 Percentage of population at mid-year in urban areas, 1950‒2050 Source: UNDESA (2018).
regions have the highest urban growth rates, namely Africa and Asia at 3.58 per cent and 2.16 per cent respectively in the period 2015 to 2020. Africa and Asia will also concentrate 90 per cent of the world’s future urban growth. The global urban hierarchy is also shifting in several ways (UNDESA 2018). First, in terms of the distribution of urban population across cities of different sizes: a growing n umber of people are living in larger cities. In 1950, 24 per cent of the global urban population resided in cities of above 1 million inhabitants. This will rise to 48 per cent by 2035. In the category of cities below 1 million inhabitants, most of the population is in intermediate cities of less than 300,000 inhabitants. In 2018, such cities accounted for 41 per cent of the global urban population. Second, in terms of where cities of different size are located globally, the largest cities are located in developing countries. In 2018, 80 per cent of the megacities with more than 10 million inhabitants were in developing countries. Third, in terms of the pace of growth of cities within the global urban hierarchy, smaller ones will grow faster. In 2018, the highest rates of growth are observed in smaller size settlements of less than 300,000 inhab itants in Asia and Africa. At the national scale, a high degree of concentration of urbanization in one or two primate cities is observed mostly, but not only, in developing countries. Latin American capitals such as Asuncion (Paraguay) and Montevideo (Uruguay) concentrate more than 50 per cent of the total urban population and African capital cities like Nouakchott (Mauritania), Brazzaville (Congo), and Libreville (Gabon) have 30‒50 per cent of urban population. However, Tokyo concentrates a third of the urban population (30.4 per cent), and European cities like Tiblisi (Georgia), Lisboa (Portugal), and Helsinki account for onethird to one-fourth of the total urban population (UNDESA 2018). Essentially, the fastest and greatest magnitude of urban growth and change will take place in developing countries, and mainly in smaller settlements. Many national urban systems will likely also continue to be dominated in demographic and economic terms by one or
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
134 Edlam Abera Yemeru two primate cities. These trends have particular implications for the promotion of industrialization as an economic development instrument in developing countries, given the strong interactions between cities and economic productivity.
8.4 Urban Scale Economies and Productivity Cities are significant in today’s world because they concentrate not only the global popula tion, but also economic activities. Historically, a strong economic rationale underpinned the emergence of early cities given their core function of facilitating trade and exchange (Soja 2010). Today, just 600 cities generate 60 per cent of global GDP, with more Asian cities expected to drive this growth up to 2025 (Cadena et al. 2012). ‘The simple correlation coefficient across countries between the per cent urbanized in a country and, say, GDP per capita (in logs) is about 0.85’ (Henderson 2003a). A 10 per cent increase in levels of urbanization causes an estimated 61 per cent rise in per capita GDP (Glaeser 2011). The economic significance of cities derives from the close association between urbanization and economic productivity, which in turn explains the concentration of industrial activity in and around urban areas. The tendency for industry to be concentrated in and around cities makes the study of space a critical dimension of the field of economics. Space has not been a core focus in eco nomics, having received sporadic treatment with the field of spatial economics having a ‘thin history’ (Derycke and Huriot 1998; Krugman 1998). This has been changing rapidly with a robust and growing body of work examining the interrelations between the organ ization of space and economic growth (see Gill and Goh 2010). In the 1990s, specifically the New Economic Geography headed by Paul Krugman’s work sought to explain unevenness in distribution and growth of economic activities in space, intensifying the focus on geog raphy and spatial dynamics in the field of economics. The urban economics field seeks to explain these dynamics by interrogating why cities exist and why dense areas are signifi cantly more productive (Glaeser and Gottlieb 2008). There is ample evidence linking cities and economic productivity (Duranton 2008). Urban scale economies, or the benefits that cities offer to firms by cutting costs and maxi mizing returns, are critical for industrial firms. Industrial firms co-locate and do so often in cities to take advantage of such scale economies. If location relative to cities did not matter for industrial firms, they would be arbitrarily located (Glaeser and Gottlieb 2008; Hoogstra and Van Dijk 2004). Urban scale economies, also referred to as agglomeration economies, occur in two main ways. First, through the concentration of industries from within the same sector, also known as localization economies, and second as a result of the concentration of a diversity of firms across sectors, known as urbanization economies. The analysis of localization econ omies has its roots in the work of Alfred Marshall in 1920 that identified factors for success ful industrial clusters (Krugman 1991). By concentrating in cities, firms and workers experience gains in productivity through efficient sharing of infrastructure, amenities, and inputs; pooling of skilled workers; better matching between firms and workers, buyers and suppliers, or business partners; and learning and knowledge spillovers including through uptake of new technologies and practices (Duranton 2008; Puga 2009).
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
Industrial Hubs, Urban Systems, and Economic Development 135 Urbanization economies gained prominence through the work of Jane Jacobs (1969) which showed how cities nurture the flow of ideas and human capital development, facili tating innovation (see Glaeser and Gottlieb 2008). There is evidence of both effects in devel oping countries, although much of the studies and evidence so far have focused on developed countries (Duranton 2008 for references; Fan and Scott 2003). However, the degree of importance and influence of these advantages may vary, such that some effects are more important than others for certain industries (Ellison et al. 2010). While there is broad consensus on the magnitude of urban agglomeration economies, this is less the case on the sources (Puga 2009). Questions have been raised as to what really drives agglomeration effects. For instance, is higher productivity in large cities a result of the survival of the fittest firms rather than agglomeration effects per se? Is it the greater presence of skills in more dense areas that is driving the positive effects of density rather than the other way around? (See reference in Puga 2009; Glaeser 2011). There has been debate as to whether agglomeration effects are a consequence of the advantages innate to certain locations, such as proximity to a natural harbour or resources (Glaeser and Gottlieb 2008). Nonetheless, what is clear is that industrial activity continues to be concen trated in urban areas. However, this is differentiated within the system of cities in a country, as the productive advantages offered by cities also vary. Urban systems theory has sought to interrogate and explain the reasons for such differentiation.
8.5 Urban Systems: Theoretical Underpinnings An urban system is a group of cities of different size and function distributed within a defined geographical space—in the case of this chapter, the national boundaries of a specific country. Several significant propositions from across a variety of disciplines, including eco nomics, geography, and sociology, have influenced urban systems theory (Aiken et al. 1987). In the urban economics field, theories of urban systems seek to explain the distribution of economic production and consumption across cities of different size and function (AbdelRahman and Anas 2004). Before delving into the association between economic produc tion, city size, and function, it is worth considering some of the theoretical roots of urban systems theories. Early nineteenth-century thinking around the relationship between cities and economic production include the work of Johann Heinrich von Thünen, which interrogated the organization and location of agricultural production and land use in relation to cities (Grotewold 1959). He proposed the concept of the ‘isolated state’ where concentric rings of agricultural production surround and supply a central city, with distance from the city’s market, price of products, and land rent determining the type of products and costs. This provided an early foundation for examining the space‒economy interface across the fields of geography and economics (Block and DuPuis 2001). Theoretical reflections by geographers made a significant contribution towards the understanding of how economic activities are concentrated in space, and in specific locations within an urban system. Particularly influential has been central place theory, championed
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
136 Edlam Abera Yemeru by geographers Walter Christaller and August Losch, which contends that goods require different scale economies so their location across cities varies based on this need (Hsu 2012). In the 1950s and 1960s, spatial concentration of economic activity was explored by geographers such as Harris and Pred who demonstrated that manufacturing in the United States concentrated in areas with good access to markets, and in turn access to markets was better in locations where more firms chose to produce (Krugman 1998). Other works that have shaped urban systems theory include those related to the size distribution of cities explain ing how cities of different size are located within a hierarchical urban system, and the asso ciation with levels of economic development (Berry 1962). In the field of economics, the work of Henderson (1974) is considered to have had a par ticularly critical role in propelling forward the concept of systems of cities as an analytical lens through which the economy could be understood (Abdel-Rahman and Anas 2004). Henderson examined the conditions and factors that determine the size and types of cities in relation to the production and consumption of goods. He posits that cities exist and grow given economies of scale in production and consumption of goods. Given that different goods require different scale economies, the size of city is defined by the type of goods in which it specializes or produces (Henderson 1974). The size and function of cities is con sidered both a driver and an outcome of the type of goods produced and consumed within those cities.
8.6 Industrial Location and the Urban System Industrial and manufacturing activities are unevenly distributed across space with a clear geographic concentration in some locations. Economies of scale drive the geographic con centration of economic activities in space (Krugman 1998). In particular, firms and workers cluster in close proximity to cities because of the scale economies arising in those cities and the associated benefits. Because different cities support different scale economies, industrial activities vary by city size and type (Henderson 1974, 1983). Depending on the type of scale economies needed for the products in which they specialize, industries are drawn to differ ent types of cities. A national urban system is composed of cities in different locations, varied in size as well as spatial and economic structure, and able to support different scale economies. In an urban system composed of different cities, an interdependent hierarchy emerges with a division of labour between cities of different size in terms of their economic structure and functions (Aiken et al. 1987). In a sense, cities within an urban system will have distinct economic functions. As the population size of an economy grows, the urban system selforganizes into a hierarchical system (Fujita et al. 1999). In this structure, the layers within the urban hierarchy have similar functions and host the same kinds of industries (Hsu 2012). The size distribution of cities across a hierarchical urban system is associated with pat terns of industrial production. Henderson (1983) makes clear linkages between national industrial production and city size. He observes a direct link between patterns of goods produced within a national economy and the size composition of cities in that economy.
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
Industrial Hubs, Urban Systems, and Economic Development 137 As such, he argues, the size of cities within an urban system is associated with the type of goods produced within those cities, although it may be weakened by technological change. Depending on the type of industrial hubs development, and the types of goods produced, different cities offer varied advantages and conditions necessary for firm productivity. Goods requiring higher scale economies are located in larger cities, while those requiring lower scale economies locate in smaller cities. As such, larger cities provide more expensive and specialized goods, while smaller ones focus on goods that meet basic needs (Aiken et al. 1987). Larger cities are associated with financial and business services and light and high-tech manufacturing, while smaller ones may be linked to natural resource-dependent industries (Henderson 1983). Larger cities with a more diverse industrial structure, and therefore generating urbanization economies, may be more attractive for service sectors such as entertainment, financial services, and IT, while smaller more specialized cities con centrate industries such as textiles, steel, and food processing where localization economies are effective (Henderson 2003b; Fujita et al. 1999; Abdel-Rahman and Anas 2004). Diversified and specialized urban economies are of differentiated importance across the life cycle of a product. Diversified urban economies could serve as ‘nurseries’ for the early stages of a product life cycle and more specialized ones concentrate goods that are fully developed with standard production procedures (Duranton and Puga 2001). City size influences productivity while the size of a city is itself shaped by the location decisions of workers and capital owners as they seek to maximize their respective advan tages (Henderson 1974). There is strong evidence that the productivity of firms and workers is higher in larger and denser urban areas (Duranton 2008). The skills of the workforce are also more diversified in larger cities than in smaller ones. As cities grow, so does their prod uctivity (Sveikauskas 1975). With a doubling of city size, productivity has been found to increase by 3 to 8 per cent (Rosenthal and Strange 2004). Although firm selection, where only the most productive firms survive in larger cities, is considered a factor for this, it is not always the case (Combes et al. 2012). Learning may determine the location decisions of firms related to different-sized cities and urban system. New and inexperienced firms learn and grow better in large and diverse cities, but later relocate to smaller more specialized cities with lower costs (Henderson et al. 1995; Henderson 2003b). The accumulated knowledge and specialization of a city is also an important consideration. Cities with rich knowledge and a history of concentration on specific products may be more attractive to some firms than cities with no such experience (Henderson 2003b). The pattern of concentration of urban population in cities of different sizes may also have implications for economic and industrial activity. Specifically, the extent to which urban population in a country is concentrated in one or two large primate cities impacts product ivity growth, with over- or under-concentration having costly impacts (Henderson 2003a). Henderson argues that in earlier stages of economic development, higher urban and indus trial concentration is an advantage and is later followed by eventual deconcentration as economic development hastens (Henderson 2003b). Deconcentration of manufacturing to urban areas beyond the primate city fosters specialization of those cities. This requires suf ficient intercity transport and communications for efficient and timely connection to national and international markets, and a degree of local fiscal autonomy for cities to better support local industrial needs and compete with primate cities for industry and population (Henderson 2003b).
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
138 Edlam Abera Yemeru As industrial hubs gain prominence to accelerate structural change, it becomes critical to look at the ways in which the configuration of national urban systems could constrain or enable their success. The size of cities, and their structure of their economies, has direct implications that can either foster or hinder industrial competitiveness. As much as cities offer economic of scale, they can also undercut the productivity of industrial activity through diseconomies of scale.
8.7 Urban Diseconomies and Industrial Activity In as much as scale economies are desirable for industrial activity and lead to concentration in cities, diseconomies can arise representing centrifugal forces against concentration (Krugman 1998; Nathan and Overman 2013). As they become excessively large and dense, cities offer diminishing productivity benefits while diseconomies related to congestion, rent, wages, and the like increase, eventually leading firms to relocate (Wheeler 2003). The way in which a city is functioning, or the extent to which externalities are minimized, is just as critical as the city size (Gill and Goh 2010). So manufacturing firms and industries with high transport costs tend to avoid dense, urban areas (Glaeser and Gottlieb 2008). Yet even when diseconomies take effect, firms may remain in a city. For instance, despite rising wages and higher rent in more dense and large urban areas, firms stay if the productive advantages outweigh costs (Puga 2009). Diseconomies manifest at different stages of urban development and industrialization. In developing-world regions, urban constraints undercut efficiencies, and constrain the performance of manufacturing firms even before agglomeration benefits come into play. For instance, premature diseconomies are observed in African cities where gaps in urban infrastructure, services, and financing occur (Kessides 2005; UNECA 2017), regardless of size and density and at early stages of industrialization. The case of Africa is of particular relevance as this is where 90 per cent of future urban growth will take place, together with Asia, and industrial hubs are priorities for economic development. The size of cities in Africa has been considered a challenge. Freire et al. (2014) note that cities in sub-Saharan Africa are too small to generate economies of scale, thus show limited productivity gains and economic growth. They argue that there are not enough large cities to offer agglomeration benefits for the economy. Lall et al. (2017) underscore three main urban challenges that constrain economic gains in African cities. First, they are crowded rather than economically dense. They concentrate population without commensurate investments in infrastructure, services, and housing, such that the costs of such concentration outweigh the benefits. Second, cities are disconnected without sufficient transport connectivity which undercuts scale and agglomeration gains for firms. Third, labour costs are high due to costs of living, which in turn reduces returns on investment. Under such conditions, the potential of cities to support the development of industrial hubs is limited. Africa’s experience with respect to industrialization and urbanization is of particular importance, given that it departs from the historical experience of cities and industries growing in mutually reinforcing ways. Africa has the world’s highest rates of urban growth
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
Industrial Hubs, Urban Systems, and Economic Development 139 in the period 2015 to 2020 (3.58 per cent) while the contribution of manufacturing to the economy is stagnant and weak (UNDESA 2018; UNECA 2017). African cities and national urban systems face numerous barriers that undercut their productivity and ability to sup port industrial activity, including the following: high costs of mobility, housing, energy, and other infrastructure; low spatial and economic density; and disconnected and imbalanced urban system (UNECA 2017). Industrial hubs development being a major instrument for economic transformation in developing countries, and especially in Africa, whether or not cities in those regions have the necessary conditions for industrial activity, becomes critical. Also necessary is a consideration of how cities function and interact within a global urban system where a hierarchy of cities shapes economic and industrial activity in differentiated ways. In such a context, cities in developing countries faced with the constraints highlighted above become even less competitive.
8.8 The Global Urban System and Industrial Hubs Beyond the national urban system, the global urban system also shapes the structure of operations of industrial firms. In an urban era where cities concentrate populations and economic activity, a networked hierarchy beyond national borders is prominent, most notably advanced through the work of Saskia Sassen on the ‘global city’ (Sassen 2005). With the globalization of the global economy, cities serve as major locations for economic pro duction and exchange, interacting and transacting with one another across national borders. The operations of industrial firms may be located in a network of global cities, especially across a multiplicity of sectors and types of products. In this sense, the global urban system offers different locational options for industrial firms, with cities in more than one country becoming important factors for their performance. In a sense, the scale economies that firms seek would be spread in cities beyond countries, and cities would compete in a global urban system based on the productive advantages they offer. As global capital and industrial activity continues to become ‘footloose’, the competitive ness of cities within a global urban system is shifting. The location preferences of industrial firms within a global urban system would also differ by the type of products and the specific advantages offered by cities within the system. The growth of high-tech firms in cities within the global urban system is a case in point. Florida (2002) examines the ‘economic geography of talent’, or the ways in which the spatial distribution of talent and industrial activity are interrelated. He theorizes that in a globalized economy where innovation is a key competitive advantage, the extent to which a city or region is endowed with a ‘creative class’ is a determinant of its ability to attract hightech industries and result in higher incomes. The creative class is a different segment of the labour force from highly skilled labour, in that the creative class has a specific preference to locate in cities and regions with high levels of cultural assets as well as ethnic and lifestyle diversity (Lorenzen and Andersen 2009). In addition, high-quality housing, work enablement, and specialized consumption are important factors that draw the creative class to specific cities. Florida therefore concludes that cities with the ability to attract a creative class become
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
140 Edlam Abera Yemeru competitive centres attracting high-tech firms. This is pertinent not only when considering the national urban system but also the global network of cities, whereby high-tech indus trial firms take location decisions based on the creative class endowments of cities. On the other hand, globalization and technological advances may make the scale econ omies arising from spatial proximity less critical for some industrial activity. Innovations in communications and IT in particular have diminished the need for physical co-location to facilitate knowledge and learning exchange between industrial firms, enable workers to operate away from firms, and enable virtual trading and exchange. However, for many types of products and firms, scale economies arising from locating in cities remain fundamental to their productivity and performance.
8.9 Enabling the Urban-industrial Economy In order to harness the advantages of cities for industrial hubs development, deliberate policies are needed at different scales, but with a strong national vision and direction. Experiences from some countries demonstrate how national government deliberately tar geted existing urban agglomerations, or promoted others, in order to enable a mutually reinforcing and highly productive ‘urban-industrial economy’. One of the illustrative cases of deliberate policies to link urban and industrial growth in mutually beneficial ways is the case of China. China’s urban transformation has been heav ily impacted by the country’s national industrial strategy and economic growth model (Chan 2010). In the post-reform period after 1978, industrialization was pursued in tandem with and through institutional restructuring of cities and a nationally defined spatial frame work (Qin and Han 2012). Rapid urban growth was both a vehicle and outcome of the expansion of export-oriented manufacturing (Hu and Zhang 2018). Manufacturing in the Pearl River Delta led to a massive growth of major cities in the eastern region, but also those close by, given the spillover of industrial and urban growth effects (Chan 2010). A mix of public policies and investments shaped urban and industrial growth in inter related ways. The creation of special economic zones and channelling of FDI to coastal provinces in the east fostered rapid urban growth in those areas. Urban policy interven tions, such as the devolution of administrative power and prioritization of city economic performance targets, created the conditions for manufacturing growth. Migration policies managed the inflow of workers in a way that met labour needs for growing industries, while controlling urban growth, with likely productivity losses (Au and Henderson 2006). Following a strong concentration of manufacturing in and around urban centres in coastal regions, rising agglomeration externalities and improvements in infrastructure and facil ities enabled new entrants to consider inland locations to enter the market (Zhu and He 2014). China’s twin urban and industrial growth trajectory has not been without chal lenges, including environmental costs and imbalances in the national urban system leading to a call for a new model of urbanization (Hu and Zhang 2018). Malaysia is another case where spatial considerations and the role of cities were of importance in the country’s industrialization. Since independence, Malaysia has registered
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
Industrial Hubs, Urban Systems, and Economic Development 141 impressive economic growth, a shift towards manufacturing, with per capita income rising several-fold, underpinned by a strong development planning culture since the 1950s (Lee and Chew-Ging 2017). While post-independence development plans prioritized investments in agriculture, the government progressively and explicitly adopted a coordinated regional development approach to plan economic and urban development. As manufacturing took prominence as a national development planning priority, the government sought to disperse industries towards smaller urban centres while at the same time enhancing infrastructure and other facilities in urban areas to better serve industry, and improving connectivity between urban centres (Hutchinson 2017). Progressively and more so in Malaysia’s latest national development plan, integrated spatial and economic planning has been key, with regional economic corridors and cities considered key engines of growth and innovation, and thereby seen as investment priorities. The current five-year development plan prioritizes investing in competitive cities as one of the ‘game changers’ to accelerate development through competitive city master plans for four cities (Kuala Lumpur, Johor Bahru, Kunching, and Kota Kinabalu) and corridor plans to boost regional development (Federal Department of Town and Country Planning & Ministry of Urban Wellbeing, Housing and Local Government of Malaysia 2016). A key focus of the city master plans is to facilitate agglom eration and economic density through promoting knowledge-based clusters and improved infrastructure for efficiency. The vision is to expand the approach to more cities over time. Malaysian cities are thus considered central to the country’s economic transformation agenda through strengthening the performance and competitiveness of industry. South Korea’s export-oriented industrialization is also closely tied to its urbanization. Cities were at the heart of this transformation. Initial industrialization was directed in and around the cities of Seoul and Pusan with two-thirds of the non-agricultural and non-mining labour force working in those two regions (Ho 1979). Industrial concentration in and around Seoul and Pusan has been attributed to three main reasons. First, Seoul-Inchon and Pusan were central to the contacts between Korean producers and foreign buyers necessary to advance exportoriented industrialization. Second, businesses preferred to be located near Seoul, closer to the central government and its decision-making machinery. Third, infrastructure was concen trated in and near Seoul and Pusan so spatially dispersed industrialization across the country could not be supported (Ho 1979). Later, from 1970 onwards, the government decided to limit the growth of the country’s three largest cities (Seoul, Pusan, and Taegu) while dispersing industries to other locations. To do so, the government focused on developing infrastructure and industrial zones away from existing urban areas, offering financial and tax incentives in those locations. In due course, this led to the growth and emergence of other urban and indus trial hubs. The South Korean experience reaffirms the importance of considering the location of industrial hubs relative to urban areas in order to maximize advantages arising from agglom eration, but also under resource constraints curtailing dispersed investments. The highlighted experiences underscore the interlinked evolution of cities and industrial hubs, implying the need for a spatial perspective in planning industrial hubs. Nathan and Overman (2013) question whether industrial policy should be spatially targeted (clustering and concentration) or space neutral. They query whether it is necessary to infuse a spatial dimension in industrial policy and what the spatial scope of industrial policy would encompass. They underscore the importance of industrial policies that are ‘place sensitive in design and delivery’ and recommend horizontal policies that target place-specific determinants of productivity as more effective than sector-focused priorities.
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
142 Edlam Abera Yemeru
8.10 Concluding Remarks Cities and urban systems are critical factors for industrial hubs development and perform ance. Historically, industrialization and urbanization have evolved along mutually reinforc ing pathways. Industrial activities gravitate towards cities as they are drawn to the productive advantages present in urban areas, or scale economies. Within a national or global urban system, industrial firms make different location decisions as cities offer differentiated scale economies for different products. At the same time, cities can also constrain the perform ance of industrial hubs through agglomeration diseconomies. Fostering a productive urban-industrial economy requires deliberate policy measures backed by a long-term vision for economic transformation. This calls for strong linkages and coherence between policies for urban, spatial, and economic development. Yet, all too often, the tendency has been for industrial policies and strategies to be formulated and implemented from a sectoral perspective, rarely considering cities and urbanization. While industrial hubs are by definition a spatial construct in that they co-locate industries, it is their location and linkage to cities that can unlock productive advantages. At earlier stages of industrialization and development, amidst resource constraints, it becomes par ticularly critical to consider how to leverage urban advantages and concentration for indus trial hubs. At the same time, the type of cities and urban system, and their functional constraints have a bearing on industrial hubs. The question is how to match cities and industrial hubs within a national urban system to optimize the mutually beneficial effects on urban and industrial development. This calls for a more explicit, deliberate policy articu lation, and corresponding strategies as countries plan, design, and invest in industrial hubs.
References Abdel-Rahman, Hesham, and Alex Anas (2004) ‘Theories of Systems of Cities’, in Vernon Henderson and Jacques-François Thisse (eds) Handbook of Regional and Urban Economics, Vol. 4 A msterdam, North Holland. See also https://www.elsevier.com/books/handbook-of-regional-and-urban- economics/henderson/978-0-444-50967-3. Aiken, Michael, Kenneth Newton, Roger Friedland, and Guido Martinotti (1987) ‘Urban Systems Theory and Urban Policy: A Four-Nation Comparison’, British Journal of Political Science 17(3): 341‒58. Au, Chun-Chung and J. Vernon Henderson (2006) ‘Are Chinese Cities Too Small?’, The Review of Economic Studies 73(3): 549–76. Berry, Brian J. L. (1962) ‘City Size Distributions and Economic Development’, Ekistics 13(76): 90‒7. Block, Daniel and E. Melanie DuPuis (2001) ‘Making the Country Work for the City: Von Thünen’s Ideas in Geography, Agricultural Economics and the Sociology of Agriculture’, American Journal of Economics and Sociology 60(1): 79‒98. Bloom, David E., David Canning, and Günther Fink (2008) ‘Urbanization and the Wealth of Nations’, Science 319(5864): 772‒5. Blumin, Stuart M. (2006) ‘Driven to the City: Urbanization and Industrialization in the Nineteenth Century’, OAH Magazine of History 20(3): 47–53. Cadena, Andres, Richard Dobbs, and Jaana Remes (2012) ‘The Growing Economic Power of Cities’, Journal of International Affairs 65(2): 1‒17. Chan, Kam Wing (2010) ‘Fundamentals of China’s Urbanization and Policy’, China Review 10(1) Special Issue: Urbanization in China: Processes and Policies: 63‒93.
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
Industrial Hubs, Urban Systems, and Economic Development 143 Childe, Gordon V. (1950) ‘The Urban Revolution’, The Town Planning Review 21(1): 3‒17. Combes, Pierre-Philippe, Gilles Duranton, Laurent Gobillon, Diego Puga, and Sébastian Roux (2012) ‘The Productivity Advantages of Large Cities: Distinguishing Agglomeration from Firm Selection’, Econometrica 80(6): 2543–94. Derycke, Pierre-Henri and Jean-Marie Huriot (1998) ‘A Brief History of Spatial Economics’, Recherch es Économiques de Louvain/Louvain Economic Review 64(1): 3‒10. Duranton, Giles (2008) ‘From Cities to Productivity and Growth in Developing Countries’, The Can adian Journal of Economics/Revue canadienne d’Economique 41(3): 689‒736. Duranton, Gilles and Diego Puga (2001) ‘Nursery Cities: Urban Diversity, Process Innovation, and the Life Cycle of Products’, The American Economic Review 91(5): 1454–77. Ellison, Glenn, Edward L. Glaeser, and William R. Kerr (2010) ‘What Causes Industry Agglomera tion? Evidence from Coagglomeration Patterns’, American Economic Review 100: 1195–213. Fan, Cindy C. and Allen J. Scott (2003) ‘Industrial Agglomeration and Development: A Survey of Spatial Economic Issues in East Asia and a Statistical Analysis of Chinese Regions’, Economic Geography 79(3): 295‒319. Federal Department of Town and Country Planning & Ministry of Urban Wellbeing, Housing and Local Government of Malaysia (2016) Malaysia National Report for the 3rd United Nations Conference on Housing and Sustainable Urban Development. Available at http://habitat3.org/wp-content/ uploads/Malaysia-National-Report-28092016.pdf. Florida, Richard (2002) ‘The Economic Geography of Talent’, Annals of the Association of American Geographers, 92(4): 743‒55. Freire, Mila E., Somik Lall, and Danny Leipziger, Danny Leipziger (2014) ‘Africa’s Urbanization: Chal lenges and Opportunities’. The Growth Dialogue, Working Paper No. 7. Fujita, Masahisa, Paul Krugman, and Tomoya Mori (1999) ‘On the Evolution of Hierarchical Urban Systems’, European Economic Review 43: 209‒51. Gill, Indermit S. and Chor-Ching Goh (2010) ‘Scale Economies and Cities’, The World Bank Research Observer 25(2): 235‒62. Glaeser, Edward L. (2011) ‘Productivity and the Quality of Life’, Science 333(6042): 592‒4. Glaeser, Edward L. and Joshua Gottlieb (2008) ‘The Economics of Place-making Policies’, Brookings Papers on Economic Activity (Spring): 155–239. Goheen, Peter G. (1973) ‘Industrialization and the Growth of Cities in Nineteenth-century America’, American Studies 14(1): 49‒65. Gollin, Douglas, Remi Jedwab, and Dietrich Vollrath (2016) ‘Urbanization with and without Industri alization’, Journal of Economic Growth 21(1): 35‒70. Grotewold, Andreas (1959) ‘Von Thunen in Retrospect’, Economic Geography 35(4): 346‒55. Henderson, J. Vernon (1974) ‘The Sizes and Types of Cities’, The American Economic Review 64(4): 640‒56. Henderson, J. Vernon (1983) ‘Industrial Bases and City Sizes’, The American Economic Review 73(2): 164‒68. Henderson, Vernon (2003a) ‘The Urbanization Process and Economic Growth: The So-What Ques tion’, Journal of Economic Growth 8(1): 47‒71. Henderson, Vernon (2003b) ‘Urbanization in Developing Countries’, The World Bank Research Observer 17(1): 89‒112. Henderson, Vernon, Ari Kuncoro, and Matt Turner (1995) ‘Industrial Development in Cities’, Journal of Political Economy, 103(5): 1067‒90. Ho, Samuel P. (1979) ‘Rural–Urban Imbalance in South Korea in the 1970s’, Asian Survey 19(7): 645‒59. Hoogstra, Gerke J. and Jouke Van Dijk (2004) ‘Explaining Firm Employment Growth: Does Location Matter?’, Small Business Economics 22(3/4): 179‒92. Hoyt, Homer (1963) ‘The Growth of Cities from 1800 to 1960 and Forecasts to Year 2000’, Land Eco nomics 39(2): 167–73. Hsu, Wen-Taid (2012) ‘Central Place Theory and City Size Distribution’, The Economic Journal 122(563): 903‒32.
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
144 Edlam Abera Yemeru Hu, Biliang and Kunling Zhang (2018) ‘New Urbanisation in China: A Multidimensional Perspective— Empirical Analysis of 289 Prefecture and Higher-level Cities’, in Ross Garnaut and Ligang Song (eds) China’s 40 Years of Reform and Development: 1978–2018. Canberra: ANU Press, pp. 455–86. Hutchinson, Francis E. (2017) ‘Evolving Paradigms in Malaysia’s Regional Development Policy’, Journal of Southeast Asian Economies 34(3): 462‒87. Jacobs, Jane (1969) The Economy of Cities. New York: Random House. Kessides, Christine (2005) ‘The Urban Transition in Sub-Saharan Africa: Implications for Economic Growth and Poverty Reduction’. Africa Region Working Paper Series No. 97. Available at http:// web.worldbank.org/archive/AFRtrade/WEB/PDF/WP97.PDF. Krugman, Paul (1991) ‘Increasing Returns and Economic Geography’, Journal of Political Economy 99(3): 483‒99. Krugman, Paul (1998) ‘Space: The Final Frontier’, The Journal of Economic Perspectives 12(2): 161‒74. Lall, Somik V., J. Vernon Henderson, and Anthony Venables (2017) Africa’s Cities: Opening Doors to the World. Washington, DC: World Bank. Lee, Cassey and Lee Chew-Ging (2017) ‘The Evolution of Development Planning in Malaysia’, Journal of Southeast Asian Economies 34(3): 436‒61. Lorenzen, Mark and Kristina Vaarst Andersen (2009) ‘Centrality and Creativity: Does Richard Florida’s Creative Class Offer New Insights into Urban Hierarchy?’, Economic Geography 85(4): 363–90. Morris, A. E. J. (1971) ‘History of Urban Form-I: The Industrial Revolution’, Official Architecture and Planning 34(2): 141, 143, 145. Nathan, Max and Henry Overman (2013) ‘Agglomeration, Clusters, and Industrial Policy’, Oxford Review of Economic Policy 29(2): 383‒404. Puga, Diego (2009) ‘The Magnitude and Causes of Agglomeration Economies’. IMDEA Social Sciences and CEPR. Available at https://diegopuga.org/papers/jrs50agg.pdf. Qin, Bo and Sun Sheng Han (2012) ‘The Location Dynamics of Firms in Transitional Shanghai, 1978–2005’, Urbani Izziv 3(Supplement 2): S161‒71. Rosenthal, Stuart S. and William C. Strange (2004) ‘Evidence on the Nature and Sources of Agglom eration Economies’, in Vernon Henderson and Jacques Francois Thisse (eds) Handbook of Regional and Urban Economics, Vol. 4. Amsterdam: North Holland. Available at https://www.elsevier. com/books/handbook-of-regional-and-urban-economics/henderson/978-0-444-50967-3. Sassen, Saskia (2005) ‘The Global City: Introducing a Concept’, The Brown Journal of World Affairs 11(2): 27‒43. Soja, Edward (2010) ‘Cities and States in Geohistory’, Theory and Society 39(3/4): 361‒76. Stobart, J. (2000) ‘In Search of Causality: A Regional Approach to Urban Growth in Eighteenthcentury England’, Geografiska Annaler, Series B, Human Geography 82(3): 149‒63. Sveikauskas, Leo (1975) ‘The Productivity of Cities’, The Quarterly Journal of Economics 89(3): 393‒413. Tilly, Charles (2010) ‘Cities, States, and Trust Networks: Chapter 1 of Cities and States in World History’, Theory and Society 39 (3/4): 265‒80. UNECA (United Nations Economic Commission for Africa) (2017) Urbanization and Industrializa tion for Africa’s Transformation. Addis Ababa: UNECA. UNDESA (United Nations, Department of Economic and Social Affairs), Population Division (2018) World Urbanization Prospects: The 2018 Revision. Available at https://www.un.org/development/ desa/publications. Weber, Adna Ferrin (1904) ‘The Significance of Recent City Growth: The Era of Small Industrial Centres’, The Annals of the American Academy of Political and Social Science 23: 29‒42. Wheeler, Christopher H. (2003) ‘Evidence on Agglomeration Economies, Diseconomies, and Growth’, Journal of Applied Econometrics 18(1): 79–104. Zhu, Shengjun and Canfei He (2014) ‘Global, Regional and Local: New Firm Formation and Spatial Restructuring in China’s Apparel Industry’, GeoJournal 79(2): 237‒53.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
pa rt I I
H ISTOR IC A L C ON T E X T A N D A NA LY T IC A L THEMES
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
chapter 9
I n dustr i a liz ation, H u bs, a n d Catch-u p The World Economy in Historical Perspective Deepak Nayyar
The object of this Handbook is to analyse the role of industrial hubs in economic development, in terms of theories, policies, and practices, with a focus on the national context. The object of this chapter is somewhat different. Even if it is a variation on the theme, there is a slight twist. It considers industrialization as an essential part of catching up by countries that are latecomers to development, situated in a long-term historical perspective, but its focus is on the global context. The past 250 years have witnessed profound changes and momentous shifts in the world economy with the rise and fall of nations. Catching up and falling behind has been an inevitable part of this process. It is no surprise, therefore, that manufacturing activities and industrial production have been concentrated in different countries, and different continents, at different times. And, if we think of countries as hubs of manufacturing in the world, such industrial concentrations have moved across geographical space, with profound implications not only for national development but also for the global balance of economic and political power. This chapter seeks to analyse the striking changes in the geographical distribution of manufacturing production amongst countries and across continents since 1750, over a period that spans more than two-and-a-half centuries, which could be described as a movement of industrial hubs in the world economy over time. While doing so, it does not even attempt to analyse the underlying factors, for that would require several chapters. The discussion is structured as follows. Section 9.1 outlines the rationale for industrialization as an imperative in catching up and explains the role of industrial hubs in the process, to suggest that, in the wider context of the world economy, it is possible to think of countries as industrial hubs. Section 9.2 examines the dramatic changes in the distribution of world manufacturing production during the period from 1750 to 1913, as the Industrial Revolution in Britain, followed by the advent of European colonialism and imperialism, led to the industrialization of Western Europe, with
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
148 Deepak Nayyar Britain as the new industrial hub of the world, and the deindustrialization of Asia, with a devastating impact on China and India. Section 9.3 considers the six decades from 1913 to 1973, during which the United States was the dominant industrial hub in the world economy, while Western Europe diminished in relative importance, as some new industrial hubs emerged, particularly Japan, and developing countries arrested their deindustrialization. Section 9.4 analyses the changes in the distribution of manufacturing output in the world economy, during the period from 1970 to 2017 in which 2000 was an important turning point, that border on the transformative, as the concentration of manufacturing in the old industrial hubs diminished significantly and Asia, particularly China, emerged as a new industrial hub. Section 9.5 concludes.
9.1 Industrialization and Industrial Hubs Development is about creating production capabilities in economies, by extending production-possibility frontiers, and ensuring the well-being of people in countries, by extending consumption-possibility frontiers. Economic history suggests that no country has managed such a transformation without going through a period of sustained industrialization. There are some exceptions where countries have relied on an exploitation of nat ural resources such as petroleum to attain high income levels, but these examples are few and have not been associated with a transformation of production capabilities. The underlying reason is simple. It is possible for an economy to transform its production capabilities only if it shifts the composition of output and employment from low productivity activities to higher productivity activities. In this process, there is a lead role for the manufacturing sector as the key driver of economic growth, which is essential for catching up on the part of countries that are latecomers to development. This proposition was also implicit in the Lewis (1954) model. It started from the premise that, in countries at low levels of income and at early stages of development which are capital scarce, there is surplus labour that is underemployed, irrespective of their endowments of land or natural resources. In these situations, the marginal productivity of labour in the agricultural sector, as also the social opportunity cost of labour, is zero or close to zero, so that its withdrawal does not lead to any reduction in output, while it mobilizes their most abundant, yet underutilized, resource—people—for development. The transfer of such surplus labour from the agricultural sector to the manufacturing sector, at a subsistence-plus wage,1 increases the profits of capitalists, reinvestment of which is the key source of capital accumulation and economic growth. In effect, the transfer of surplus labour with near-zero marginal productivity in traditional agricultural sectors to modern industrial sectors is a source of economic growth. The Kaldor (1966) model went much further in developing this causation to suggest that the manufacturing sector is the engine of growth in economies. This was set out in terms of three laws. First, there is a positive relationship between the growth of manufacturing out1 Surplus labour in the agricultural sector earns a subsistence wage. If workers move to the urban sector, they need to earn a subsistence-plus wage that would compensate them for the cost of moving and the higher cost of living.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrialization, Hubs, and Catch-up 149 put and the growth of GDP, which is explained by the absorption of surplus labour from the agricultural sector into the manufacturing sector. Second, growth of manufacturing output leads to growth of productivity in manufacturing, which is attributable to static and dynamic scale economies, where the former depend on plant size or output levels at any point in time, while the latter derive from learning-by-doing that is a function of cumulative past output (Arrow 1962) or cumulative production experience (Kaldor 1962) over time. Third, growth of manufacturing output is associated with an overall increase in productivity in the economy driven by spillover effects elsewhere. Industrialization has, therefore, always been emphasized as an imperative in the quest for catching up with countries at higher levels of income or at a more advanced stage of development. The spread effects of industrialization are associated with higher investment levels on the demand side and scale economies on the supply side, together with backward and forward linkages within and between sectors. This enables economies to capture external ities that arise in the process. But that is not all. Industrialization also leads to an interaction between the demand side and the supply side. An increase in market size facilitates the realization of scale economies, thus bringing about a cost reduction, just as a reduction in costs, hence prices, induces a demand expansion. In an industrializing economy, a rapid increase in the share of manufacturing in total output may then be associated with a steady decline in average costs over time. The cost reduction in the form of lower prices would stimulate demand expansion. This circular and cumulative causation, which is virtuous even if complex, is often termed the Kaldor-Verdoorn Law. Industrial hubs have evolved as a means to the end of industrialization. This notion entered the lexicon of economists in the late nineteenth century when Alfred Marshall developed the concept of industrial districts, in the context of competitive entrepreneurial capitalism. In the most elementary sense, it could be thought of as an area where a concentration of firms has settled down. Such a localization of manufacturers may be attributable to a geographical proximity to sources of raw materials, or markets for goods, or even access to land and water. This primitive form of localization, if it lasts long enough, gets embedded and evolves into an industrial district.2 The competitive strength of small or medium firms in such a district derives from external economies that ‘depend on the general organization of trade, on the growth of the knowledge and appliances common to the trade, on the development of subsidiary industries and so on’ (Marshall 1898: 50) In this process, access to labour skills, whether hereditary or through a local market for special skills, is particularly important. Such industrial districts are characterized by a peculiar combination of competition and cooperation that creates an atmosphere or a milieu (Marshall 1919), which is conducive to the spread and growth of manufacturing activities, where agglomeration economies and entrepreneurial learning play an important role. The late twentieth century witnessed the creation of industrial hubs across the developing world in countries that sought to promote industrialization. This was an integral part of industrial policy adopted by pro-active governments that sought to promote industrialization in the context of a global managerial capitalism. It took the form of industrial parks, export processing zones, or special economic zones. The outcomes have been mixed—success, muddling-through, or failure—in different countries at different times. The objectives, too, have spanned a wide range. In some countries, it has been limited to creating infrastructure, 2 For a lucid discussion on the origin and evolution of industrial districts in the thinking of Alfred Marshall and his contemporaries in Cambridge at the time, see Belussi and Caldari (2009).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
150 Deepak Nayyar providing connectivity, and access to imported inputs, in sequestered geographical spaces. In other countries, the effort has extended to creating a milieu that is conducive to industrialization. In doing so, every country has hoped to exploit agglomeration economies, impart skills to labour, develop managerial capabilities in individuals, and create technological capabilities in firms. Industrial hubs are thus seen as catalysts or leaders. In the vocabulary of economists, the concept of industrial hubs might have surfaced in the late nineteenth century. But the geographical concentration of manufacturing activities goes back in time, much before the advent of production in factories which followed the Industrial Revolution, to traditional production by craftsmen centuries earlier during the second millennium ever since manufacturing began life. The underlying economic factors were very similar: the proximity to sources of raw materials, the availability of hereditary skills in labour, and access to markets for the goods produced. Such geographical concentrations in craft– manufacturing–production existed not only in Asia, both China and India, but also in Europe. It is clear that the existence of industrial hubs within countries is widely recognized and extensively discussed in the national context. However, in the literature on this subject, it is not quite recognized that countries might also be hubs of manufacturing in the global context. Geographical concentration of industrial production is embedded in the nature and logic of manufacturing activities. Therefore, spatial distribution is inevitably unequal across geographical space within countries. It is, then, hardly surprising that the distribution of manufacturing production is also unequal between countries across geographical space in the world economy. And if we think of countries as manufacturing hubs in the world, these industrial concentrations have witnessed dramatic shifts across countries over time. Such movement of industrial hubs in the world economy is the focus of this chapter.
9.2 The Era of Dramatic Changes: 1750–1913 Manufacturing activities were not always concentrated in the Western world. Indeed, during the seventeenth and eighteenth centuries, the world economy was characterized by a flow of manufactured goods from Asia to Europe that was paid for by a flow of silver from Europe to Asia (Findlay and O’Rourke 2007). Trade in spices was just one part of the story. Cotton textiles from India and porcelains or silks from China were much sought after in Europe. And some of the most dynamic sectors in eighteenth-century Europe were seeking to imitate and compete against goods from Asia (Parthasarathi 2011). In a study of industrialization levels across the world since 1750, Bairoch (1982) provides estimates of manufacturing production for selected regions, country-groups, and countries. Table 9.1, based on these estimates, outlines the changes in the distribution of manufacturing production in the world economy from 1750 to 1913. It presents evidence on two groups of continents and countries: the first is made up of Europe, North America, and Japan, described as ‘The West’, while the second is made up of Asia, Africa, and Latin America, described as ‘The Rest’.3 It also presents evidence on the changing shares of nine countries that were the leading manufacturing hubs in the world during that era. 3 Bairoch (1982) describes the former as the ‘Developed Economies’ and the latter as the ‘Third World’.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrialization, Hubs, and Catch-up 151
Table 9.1 The distribution of manufacturing production in the world economy: 1750–1913, in percentages: triennial annual averages 1750
1800
1830
1860
1880
1900
1913
0.1 1.9 2.9 4 2.4 5 3.8 32.8 24.5 27
0.8 4.3 3.5 4.2 2.5 5.6 3.5 33.3 19.7 32.3
2.4 9.5 3.5 5.2 2.3 5.6 2.8 29.8 17.6 39.5
7.2 19.9 4.9 7.9 2.5 7 2.6 19.7 8.6 63.4
14.7 22.9 8.5 7.8 2.5 7.6 2.4 12.5 2.8 79.1
23.6 18.5 13.2 6.8 2.5 8.8 2.4 6.2 1.7 89
32 13.6 14.8 6.1 2.4 8.2 2.7 3.6 1.4 92.5
23.2 0.1 3.8
28.1 0.8 3.5
34.2 2.5 2.8
53.2 7.5 2.6
61.3 15.1 2.4
62 24.2 2.4
56.6 32.9 2.7
73
67.7
60.5
36.6
20.9
11
7.5
47.4 100
28.3 100
15.3 100
7.9 100
United States United Kingdom Germany France Italy Russia Japan China India THE WEST of which Europe North America Japan THE REST of which China and India WORLD
57.3 100
53 100
5 100
Notes: The figures for 1913 are for the year and not a triennial average. Apart from the countries listed in the table, Europe includes Austria, Belgium, Hungary, Spain, Sweden, and Switzerland while North America includes Canada. Source: Bairoch (1982).
Manufacturing production covers the entire range of output without differentiating between technology levels or organizational structures. Thus, it includes both the traditional sector with production by craftsmen and modern industry with production in factories. The figures are based on triennial annual averages to eliminate the impact of short-term fluctuations. Table 9.1 shows that Asia, Africa, and Latin America, taken together, were dominant during the period 1750–1830, accounting for 73 per cent of world industrial output in 1750, much of which was concentrated in Asia with just China and India accounting for 57 per cent of world industrial output. The share of ‘The Rest’ remained high, even if lower, at 68 per cent in 1800 (with China and India contributing 53 per cent) and 61 per cent in 1830 (with China and India contributing 47 per cent). Over the same period, the share of Europe, North America, and Japan in world industrial output increased from 27 per cent in 1750 to almost 40 per cent in 1830. However, the situation changed significantly during 1830 to 1860. By 1860, the share of ‘The Rest’ plummeted to 37 per cent (with China and India contributing 28 per cent), while the share of ‘The West’ rose to 63 per cent. In fact, the proportional contribution of the two groups of continents and countries to world manufacturing production was almost reversed in just thirty years between 1830 and 1860. Yet, in 1860, China ranked second and India third, just below Britain ranked first, but above France fourth, United States fifth and Germany sixth, in the world in terms of total manufacturing output (Bairoch 1982: 284). After 1860, however, the share of India in world manufacturing output dropped
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
152 Deepak Nayyar faster than that of China. During the period 1880 to 1913, this dramatic shift of manufacturing production in the world continued. By 1913, the share of ‘The Rest’ fell further to a mere 7.5 per cent (with China and India contributing 5 per cent), whereas the share of ‘The West’ rose to 92.5 per cent. The decline and fall of ‘The Rest’ was obviously concentrated in China and India. But the nature of concentration in the rise of ‘The West’ was somewhat mixed and changed over time. The share of Britain in world manufacturing output, barely 2 per cent in 1750, rose to 4 per cent in 1800. The impact of the Industrial Revolution led to a rapid increase in this share to 10 per cent in 1830 and 20 per cent in 1860 to peak at 23 per cent in 1880, which made Britain the hub of manufacturing in the world. This changed as the Industrial Revolution spread to latecomers in Europe such as Germany, France, and Russia. Indeed, by 1913 the share of Germany in world manufacturing output exceeded that of Britain. The United States began to catch up with Britain even earlier, as its industrialization gathered pace from the mid-nineteenth century, so that its share of world manufacturing output rose close to one-fourth by 1900, surpassing Britain, and almost one-third by 1913, making it the hub of manufacturing in the world. Despite such rapid industrialization across the Atlantic, even in 1913, countries in Europe, taken together, accounted for 57 per cent of world manufacturing output as compared with North America at 33 per cent.4 It is not possible to compare how productivity per worker in manufacturing changed over time in these two parts of the world because there is no evidence on the number of persons employed in the industrial sector. Thus, Bairoch studied levels of industrialization in terms of manufacturing production per capita. The results of this exercise are just as striking. The ratio of manufacturing production per capita in Asia, Africa, and Latin America to that in Europe, North America, and Japan, dropped from 7:8 in 1750 and 3:4 in 1800 to 1:4 in 1860, 1:8 in 1880, 1:17.5 in 1900, and 1:27.5 in 1913.5 In the same selected years, these ratios were almost the same for China and India, although India fared somewhat worse after 1860. But China and India were not even amongst the top twenty countries in the world, in terms of manufacturing output per capita, in 1860 let alone in 1913, given their large populations. The revolutionary change in methods of manufacturing, which were developed in Britain in the late eighteenth century and spread to countries in Western Europe through the early nineteenth century, meant profound changes in the economic life of Europe (Allen 2009). Innovation, followed by continuous improvements in technologies, yielded sharp increases in productivity, output, and incomes. The rapid diffusion of the new technologies combined with their geographical spread brought about rapid industrialization in Britain, Belgium, Netherlands, France, and Germany. This industrialization in Western Europe, associated with scale economies that sharply reduced prices of manufactured goods, led to the demise of traditional industries in Asia, particularly India and China, so that the outcome was deindustrialization in Asia except for Japan. Consequently, the knowledge and skills that had been developed in Asia over centuries were slowly but surely eroded and diminished. 4 The remaining share of ‘The West’ (2.5 per cent) was attributable to Japan. Its share in world manufacturing hovered around 2.5 per cent during 1800 to 1913, as compared with 3.5 per cent during 1750 to 1800. 5 These ratios are calculated from Bairoch (1982: 294 and 302), who estimates per capita industrialization levels for country-groups and countries, based on triennial averages, as index numbers with the United Kingdom in 1900 as 100.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrialization, Hubs, and Catch-up 153 Thus, the nineteenth century witnessed a divergence not just in incomes but also in labour productivity, skill levels, and technological capabilities. The consequent path dependence had long-term consequences for development (Nayyar 2013). The obvious cause of this deindustrialization in Asia was the much greater competitiveness of industry in Britain and Western Europe. But that was not all. The transport revolution of the nineteenth century dismantled the natural protection, which was provided by the distance and time implicit in geographical barriers, to handicrafts or manufacturing industries in countries such as India and China (Nayyar 2006; 2013; Findlay and O’Rourke 2007; 2013). The advent of the steamship reduced the cost of ocean freight by two-thirds between 1870 and 1900 (Lewis 1978). The spread of the railways, everywhere, brought the hinterland of countries into the world economy, not only to source raw materials but also to sell manufactured goods (Nayyar 2006). It would have required high tariffs, possibly even an exclusion of imports through prohibitive protection, for India and China, or countries in Asia, to neutralize the impact of the revolutions in industry and transport on prices of manufactures imported from Britain or Western Europe. Of course, colonialism and imperialism meant that countries in Asia did not have the freedom to use tariffs for protecting domestic industries. The mix of gunboat diplomacy and colonial dominance imposed free trade on China, India, Indonesia, Japan, and Korea.6 There was sustained productivity growth with industrialization in Western Europe and a steady productivity decline with deindustrialization in Asia. This widening productivity gap was the basic factor underlying the divergence in per capita incomes between Western Europe and Asia. The world economy came to be divided into ‘The West’ (mostly countries with temperate climates, some of which also had white settlers) that industrialized and exported manufactures, and ‘The Rest’ (mostly countries with tropical climates in Asia and Africa) that did not industrialize and exported primary commodities. The ‘Great Divergence’ in incomes between these countries was closely related to the ‘Great Specialization’ in division of labour between countries (Nayyar 2013). Slowly but surely, these countries became dependent on the industrializing countries in Western Europe, not simply for markets and finance but also as their engine of growth (Lewis 1978). Much of Asia, except Japan, was colonized de jure or de facto, placing it on the wrong side of this divide, which led to its deindustrialization and underdevelopment. In fact, the industrialization of Western Europe and the deindustrialization of Asia were two sides of the same coin.7
9.3 The Decades of Continuity: 1913–73 The six decades from 1913 to 1973 were characterized by more continuity and less change, in the movement of industrial hubs in the world economy, as compared with the era from 1750 to 1913 discussed above. More than half this period, interspersed as it was by two World Wars and the Great Depression, was difficult if not disruptive for the world economy. In sharp contrast, the quarter century from 1948 to 1973, which has been described as the 6 For Japan, this was so only until 1911. But Japan imposed an unequal treaty on Korea in 1876, even when it was subject to unequal treaties with the West. 7 For a more detailed discussion on this issue, see Nayyar (2013). See also Williamson (2006).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
154 Deepak Nayyar ‘Golden Age of Capitalism’ (Marglin and Schor 1990), was a period in which rapid growth combined with full employment meant unprecedented economic prosperity in the industrialized West, both North America and Western Europe. The other parts of the world also fared somewhat better. The communist countries—USSR and Eastern Europe—experienced economic progress juxtaposed with political stability. For the developing countries—Asia and Africa—it was the beginning of the post-colonial era, as they attained political independence, which enabled them to pursue their national development aspirations. The ‘Great Divergence’ in per capita incomes came to a stop. And the process of reviving, or kickstarting, industrialization began. Latin American countries, which attained political independence in the early nineteenth century, were exceptions during the period 1870–1950. For one, they did not experience such a massive divergence in per capita incomes. For another, high tariffs enabled them to sustain some industrialization.8 In view of these developments, it would be instructive to consider how different continents, country groups, or countries fared at industrialization during this period. Table 9.2 sets out the available Bairoch (1982) estimates of industrial production for six benchmark years during 1913 to 1973. The countries and country-groups are not quite the same as in Table 9.1. In a changed context, Russia is replaced by the USSR, Brazil and Mexico are added, while the
Table 9.2 Changes in the distribution of industrial production in the world: 1913–73, in percentages United States United Kingdom Germany France Italy Japan USSR China India Brazil Mexico Western Europe North America Japan USSR and Eastern Europe Developing Countries WORLD
1913
1928
1938
1953
1963
1973
32.0 13.6 14.8 6.1 2.4 2.7 8.2
39.3 9.9 11.6 6.0 2.7 3.3 5.3
31.4 10.7 12.7 4.4 2.8 5.2 9.0
44.7 8.4 5.9 3.2 2.3 2.9 10.7
35.1 6.4 6.4 3.8 2.9 5.1 14.2
33.0 4.9 5.9 3.5 2.9 8.8 14.4
3.6 1.4 0.5 0.3
3.4 1.9 0.6 0.2
3.1 2.4 0.6 0.2
2.3 1.7 0.6 0.3
3.5 1.8 0.8 0.4
3.9 2.1 1.1 0.5
40.8 32.9 2.7 15.8 7.5 100.0
35.4 40.8 3.3 12.5 7.2 100.0
37.3 32.8 5.2 16.3 7.2 100.0
26.1 46.9 2.9 16.0 6.5 100.0
26.5 37.2 5.1 20.9 8.5 100.0
24.5 35.1 8.8 20.1 9.9 100.0
Notes: (a) In the bottom panel, for geographical regions and country-groups, the columns may not add up to 100 because figures are rounded off. (b) The figures for 1928 and 1938 are triennial averages. Source: Bairoch (1982).
8 For a more detailed discussion, see Nayyar (2013). See also Clemens and Williamson (2002) and Bertola and Ocampo (2012).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrialization, Hubs, and Catch-up 155 country groups are different. And the distinction between the industrialized parts of the world and developing countries is clearer. In analysing this evidence, the six decades are divided into two periods: 1913–38 and 1953–73, while the period 1938–53 is treated as an interregnum dominated by World War II and its aftermath.9 The quarter century from 1913 to 1938 did not witness any significant changes. The share of Western Europe and the United States, together, in world industrial production remained dominant at about three-fourths, with the latter higher than the former only in 1928. The share of Japan doubled from one-fortieth to one-twentieth. The shares of the USSR and Eastern Europe changed little and added up to one-sixth. The share of developing countries— Asia, Africa, and Latin America—remained almost unchanged around its 1913 level at one-fourteenth. During the interregnum of 1938–53, the share of the United States jumped from 31 per cent to 45 per cent, partly because the shares of most countries in Western Europe had not quite recovered after the War. In fact, the shares of Germany and Japan contracted by one-half. The share of USSR and Eastern Europe together remained at onesixth, while that of developing countries dropped a little. During the two decades, 1953–73, the United States remained dominant with around one-third of world industrial production, while its 1953 peak level was the exception. The share of Western Europe declined to about one-fourth, as compared with almost two-fifths until 1938, a decline that was mirrored in the shares of the leading industrial countries in the region. The share of Eastern Europe remained unchanged at 5–6 per cent, while the share of USSR rose from 11 per cent to 15 per cent, so that the share of the communist countries rose from one-sixth in 1953 to one-fifth in 1973. The beginning of the rise of Japan as an industrialized nation is also discernible in these decades as its share rose from 3 per cent to 9 per cent. Thus, even in 1973, industrialized countries—North America, Western Europe, Japan, the USSR, and Eastern Europe—accounted for nine-tenths of world manufacturing production. It is clear that the United States was the manufacturing hub, and the leading industrial nation, of the world economy during 1913 to 1973. Western Europe, taken together, was a larger industrial hub during 1913 to 1938, but the United States surpassed it thereafter. However, the decline of Western Europe was relative rather than absolute. The rise of the United States was attributable in part to its geographic and economic size. But it was also significantly attributable to state intervention in the form of industrial policy.10 The countries that emerged as industrial hubs during these decades of continuity were the early late industrializers—USSR and Japan—where the role of the state in promoting industrialization was critical.11 Developing countries, particularly in Asia, arrested the precipitous decline in their share of world industrial production, after they regained political independence during the period from the late 1940s to the late 1950s, which restored their economic autonomy, enabling them to pursue their national development objectives. The colonial era was associated with openness and markets, which were perceived as factors responsible for deindustrialization 9 Ideally, this should have been 1939–45 but 1953 is the next benchmark year after 1938 in the Bairoch estimates. 10 For an analysis of the United States as a late industrializer, see Bairoch (1993), Kindleberger (1996), Chang (2002), Maddison (2007), and Best (2018). 11 The pioneering work on Russia and the USSR was that of Gerschenkron (1962), which emphasized the ‘advantages of relative economic backwardness’. This mode of thinking also influenced studies in the economic history of Japan (Okawa and Rosovsky 1973). On the role of the state and industrial policy in Japan, the pioneering work was that of Johnson (1982).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
156 Deepak Nayyar and underdevelopment. Thus, reviving industrialization through government intervention and support was seen as an imperative in their quest for catching up. Such state-led industrialization, which was an integral part of the ‘Development Consensus’ at the time, did have its limitations. Even so, it provided the foundations for the catching up through industrialization, the results of which surfaced slowly but surely in the subsequent decades.12
9.4 Beginnings of Transformative Change: 1970–2017 There is more systematic and complete evidence on the distribution of manufacturing output in the world economy, since 1970, in national accounts statistics compiled by the United Nations. Table 9.3 presents the shares of fourteen economies, seven from the industrialized world and seven from the developing world, in world manufacturing value added, for selected benchmark years during this period of almost half a century. For reference and comparison, it also provides the corresponding shares of three country groups—industrialized countries, developing countries, and transition economies—in the world economy. In analysing the evidence, it is necessary and appropriate to divide this longish span of time into two sub-periods: 1970–2000 and 2000–17. During the last three decades of the twentieth century, the United States broadly maintained its dominant share of world manufacturing value added, on average, at about one-fourth. The share of the leading West European countries, Britain, Germany, France, and Italy, taken together, was also maintained at somewhat less than one-fifth. Japan was the rising industrial power in this period, as its share of world manufacturing value added rose from less than one-tenth to almost one-fifth. The erstwhile centrally planned economies— USSR and Eastern Europe—experienced a dramatic decline in their share of world manufacturing output from more than one-fifth in 1970 to barely one-twentieth in 1990. In their incarnation as transition economies, this share collapsed further to a mere one-fortieth in 2000. The share of the seven leading developing economies in world manufacturing output jumped from 4.5 per cent in 1970 to 16.3 per cent in 2000, while that of developing countries almost doubled from 10.1 per cent to 19.5 per cent. Obviously, industrial production in the developing world came to be increasingly concentrated in these seven countries. Between 1970 and 2000, the share of both industrialized countries and developing countries in world manufacturing value added increased by about 10 percentage points each, essentially at the expense of transition economies. During the first two decades of the twenty-first century, the shifts in industrial production in the world economy were far more dramatic. Between 2000 and 2017, the share of the United States in world manufacturing output dropped from 28 per cent to 17 per cent, while that of Britain, Germany, France, and Italy, taken together, dropped from 18 per cent to 12 per cent. The share of Japan fell from almost 20 per cent to 7.5 per cent. Consequently, during 2000 to 2017, the share of industrialized countries in world industrial production fell from more than three-fourths to less than one-half, by as much as 33 percentage points. In sharp contrast, the share of developing countries in world industrial production rose 12 For a more detailed discussion of these issues, see Nayyar (2013, 2019).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrialization, Hubs, and Catch-up 157
Table 9.3 Share of selected countries and country-groups in world manufacturing value added, in percentages 1970
1980
1990
2000
2010
2017
United States United Kingdom Germany France Italy Russian Federation Japan
29.5 3.8 8.0 3.5 3.2 – 8.8
21.8 4.3 9.5 4.8 4.6 – 11.7
22.9 4.0 9.9 4.5 5.2 3.1 18.6
27.9 3.8 7.3 3.5 3.6 0.9 19.8
17.1 2.1 6.5 2.6 2.9 1.9 11.3
16.7 1.8 6.0 2.0 2.2 1.4 7.5
China India Indonesia Republic of Korea Taiwan Brazil Mexico
(0.8) 1.1 0.1 0.2 0.2 1.0 1.1
(0.6) 1.2 0.4 0.5 0.5 2.2 1.8
(1.3) 1.3 0.6 1.5 1.1 2.2 1.4
(6.0) 1.4 0.8 2.6 1.5 1.6 2.4
18.3 2.7 1.6 2.9 1.2 2.7 1.6
27.1 3.0 1.6 3.2 1.3 1.7 1.5
67.7 10.1 22.2 100.0
69.6 14.8 15.6 100.0
78.4 16.2 5.4 100.0
78.3 19.5 2.3 100.0
53.3 42.5 4.1 100.0
45.6 50.9 3.5 100.0
Industrialized countries Developing countries Transition economies World
Notes: (a) The percentages have been calculated from data on manufacturing value added and GDP in current US dollars at market exchange rates. (b) The figures for China, in 1970, 1980, 1990, and 2000, which are not available in UN National Accounts Statistics, are calculated from the Groningen Growth and Development Centre online database and are not strictly comparable with figures for subsequent years. (c) The figures for Taiwan in 1970, 1980, 1990, 2000, 2010 are taken from Nayyar (2019). The data in column 2017 relate to 2016. Source: Author’s calculations from UN National Accounts Statistics (https://unstats.un.org/ unsd/snaama/downloads).
from barely one-fifth to more than one-half, by as much as 31 percentage points, entirely at the expense of industrialized countries. The most important reason for this dramatic transformation was the increase in the share of China in world manufacturing output from 6 per cent in 2000 to 27 per cent in 2017, by 21 percentage points, which accounted for about two-thirds of the rise in the share of developing countries as well as the fall in the share of industrialized countries. Such massive change in a short time span of just seventeen years is surprising, and difficult to explain, but that is what national accounts statistics reveal. Of course, the share of developing countries, excluding China, in world manufacturing output, also increased from 13.5 per cent in 2000 to 23.8 per cent in 2017, by more than 10 percentage points. In this context, it is worth noting that during 2000 to 2017, the entire increase in the share of developing countries in world manufacturing value added was attributable to Asia.13 In other words, over this period, 13 The share of Asian developing countries in world manufacturing value added increased from 11.1 per cent in 2000 to 43.3 per cent in 2017 (calculated from United Nations National Accounts Statistics). For a more detailed discussion on this issue, with supporting evidence, see Nayyar (2019). See also Nayyar (2013).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
158 Deepak Nayyar there was little change in the share of Latin America and Africa in world manufacturing value added. During 1970 to 2000, there were three striking changes in the distribution of manufacturing production in the world economy. The first change was that the share of the erstwhile centrally planned economies—USSR and Eastern Europe—plunged from a stable high around 22 per cent to an unprecedented low at 2 per cent. The decline began in the late 1970s with their economic slowdown. These difficulties were compounded by political strains. The crisis was precipitated by the political collapse of communism, which had devastating economic consequences leading to a sharp contraction in total output. The transition to market economies turned out to be most difficult in these countries which did not have the necessary institutions.14 The situation did stabilize after a decade of turmoil, but the recovery in their share of world manufacturing output was at best modest. The second change was the emergence of Japan as a leading industrialized country. Its share in world manufacturing production rose from 9 per cent to 20 per cent. Of course, this rise of Japan was a continuation of the process that began around 1950, in which industrialization was led by the state and promoted by a pro-active industrial policy. The third change was that the share of developing countries in world manufacturing output increased from 10 per cent to 20 per cent, essentially at the expense of USSR and Eastern Europe. This revival of industrialization in the developing world, in sharp contrast with deindustrialization in the colonial era, was attributable to the initial conditions that were created and the essential foundations that were laid through a pro-active role of the state in the early post-colonial era, the results of which surfaced after a time lag.15 The changes in the distribution of world manufacturing output which surfaced during 2000 to 2017 border on the transformative. The share of the United States dropped from more than one-fourth to about one-sixth. In 2017, it was 16.7 per cent, as compared with a share that ranged from one-fourth to one-half throughout the twentieth century. It was lower at 14.7 per cent as far back as 1880. The share of Britain, Germany, France, and Italy, taken together, dropped from somewhat less than one-fifth to less than one-eighth. In 2017, it was less than 2 per cent for Britain, about the level it was at in 1750, whereas for Germany it was 6 per cent, distinctly lower than it was from 1880 to 1938 and about the same as it was in 1953. The share of Japan dropped from one-fifth to one-thirteenth. In 2017, at 7.5 per cent, it was lower than in 1973. The decline of manufacturing in industrialized countries was relative rather than absolute. Given the timing, it was attributable partly to the global financial crisis and the Great Recession that persisted in its aftermath (Nayyar 2011). However, it was also attributable, perhaps more so, to the continuing structural change, with manufacturing giving way to services as per capita incomes reached high levels creating post-industrial economies.16 The increase in the share of developing countries in world manufacturing output, which rose from 20 per cent in 2000 to 51 per cent in 2017, was the other side of the coin, possibly 14 For an extensive discussion on problems of transition in the economies of the Russian Federation and Eastern Europe, see Amsden et al. (1994) and Cornia and Popov (2001). 15 This process of industrialization in the developing world is analysed, at some length, by the author elsewhere. See Nayyar (2013). 16 This argument about deindustrialization in advanced capitalist economies was first developed by Rowthorn and Wells (1987).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrialization, Hubs, and Catch-up 159 returning to its level around 1840.17 About two-thirds of this massive increase was attributable to China. In 2017, its share at 27 per cent had returned to its level around 1830. The remaining one-third was attributable essentially to developing economies in Asia (excluding China), amongst which India, Indonesia, South Korea, and Indonesia were most significant. Of these, in 2017, the share of India at 3 per cent had returned to its level around 1870. This success at industrialization in Asia was driven by sensible industrial policy implemented by effective governments.18 In this context, it is worth noting that Japan, with its developmental state and industrial policy, was a role model not only for South Korea and Taiwan, but also for China.
9.5 Conclusion Economic history reveals that countries which are latecomers to development have relied on industrialization in their quest for catching up in terms of production capabilities and income levels. This chapter shows that, in the process, the past 250 years have witnessed momentous shifts in hubs of manufacturing activities or industrial production across con tinents and geographies. Such shifts in industrial hubs in the world economy have, in the past, been associated with industrialization in some countries and deindustrialization in other countries, and might, in the future, be associated with premature deindustrialization in yet other countries. During the second half of the eighteenth century and the first quarter of the nineteenth century, China and India in Asia were the hub of manufacturing production in the world economy. The Industrial Revolution in Britain in the late eighteenth century, followed by the advent of European colonialism and imperialism in Asia, transformed this situation in fifty years. By 1870, Britain was the hub of industrial production in the world. This hub spread thereafter to some countries such as Germany and France in Western Europe and across the Atlantic to the United States. By 1913, the industrialization of Western Europe had led to an almost complete deindustrialization of China and India. However, the share of the United States in world manufacturing production surpassed that of Britain by 1900. Consequently, the United States, as a single country, was the dominant industrial hub in the world economy throughout the twentieth century. Of course, Western Europe also remained an industrial hub until World War II, after which it declined in relative importance. The period from the late 1940s to the early 1970s witnessed the emergence of two new industrial hubs in the world. The USSR and Eastern Europe was one. It accounted for 22 per cent of world manufacturing production by 1970. However, the communist countries were a short-lived industrial hub. Their relative importance declined steadily during 1970 to 1990 and the collapse of communism led to a situation where it was negligible by 2000. Japan was the other. It emerged as the new industrial hub during the second half of the twentieth century, as its share in world manufacturing production rose from 2 per cent to 20 per cent. 17 The share of Asia, Latin America, and Africa in world manufacturing production was 60 per cent in 1830 and 37 per cent in 1860 (Table 9.1). 18 This argument is developed at some length in Nayyar (2019). See also Chang and Zach (2019) and Wade (2019).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
160 Deepak Nayyar The United States continued its dominance, while Western Europe maintained its significance, as industrial hubs of the world during 1950 to 2000. The developing countries kick-started, or revived, their industrialization during the second half of the twentieth century, and trebled their share of world manufacturing production, for Asia to emerge, once again, as a possible industrial hub in the world economy. The first two decades of the twenty-first century have witnessed changes that border on the transformative. During 2000 to 2017, the shares of the United States, Western Europe, and Japan in world manufacturing production declined significantly and rapidly in a relatively short span of time. This fall corresponds almost exactly to the rise in the share of developing countries in world manufacturing production, which in turn is attributable largely to Asia. China, which accounted for 27 per cent of world manufacturing production in 2017, as a single country, is the new industrial hub of the world economy. Developing countries in Asia, excluding China, accounted for 16 per cent of world manufacturing production in 2017. Obviously, it is not possible to predict how the future will unfold. Even so, it is plausible to suggest, though impossible to prove, that Asia might once again become the industrial hub of the world economy as it was two centuries ago.
References Allen, Robert C. (2009) The British Industrial Revolution in Global Perspective. Cambridge: Cambridge University Press. Amsden, Alice H., J. Kochanowicz, and Lance Taylor (eds) (1994) The Market Meets its Match: Restructuring the Economies of Eastern Europe. Cambridge, MA: Harvard University Press. Arrow, Kenneth J. (1962) ‘The Economic Implications of Learning by Doing’, Review of Economic Studies 29: 155–73. Bairoch, Paul (1982) ‘International Industrialization Levels from 1750 to 1980’, Journal of European Economic History 11: 269–333. Bairoch, Paul (1993) Economics and World History: Myths and Paradoxes. Chicago, IL: Chicago University Press. Belussi, Fiorenza and Katia Caldari (2009) ‘At the Origins of Industrial Districts: Alfred Marshall and the Cambridge School’, Cambridge Journal of Economics 33(2): 335–55. Bertola, Luis and Jose Antonio Ocampo (2012) The Economic Development of Latin America since Independence. Oxford: Oxford University Press. Best, Michael H. (2018) How Growth Really Happens: The Making of Economic Miracles through Production, Governance and Skills. Princeton, NJ: Princeton University Press. Chang, Ha-Joon (2002) Kicking away the Ladder: Development Strategy in Historical Perspective. London: Anthem Press. Chang, Ha-Joon and Kiryl Zach (2019) ‘Industrialization and Development’, in Deepak Nayyar (ed.) Asian Transformations: An Inquiry into the Development of Nations. Oxford: Oxford University Press, pp. 186–215. Clemens, Michael A. and Jeffrey G. Williamson (2002) ‘Close Jaguar Open Dragon: Comparing Tariffs in Latin America and Asia before World War II’. NBER Working Paper No. w9401, National Bureau of Economic Research, Cambridge, MA. Cornia, G. Andrea and Vladimir Popov (2001) Transition and Institutions. Oxford: Oxford University Press. Findlay, Ronald and Kevin H. O’Rourke (2007) Power and Plenty: Trade, War, and the World Economy in the Second Millennium. Princeton, NJ: Princeton University Press. Gerschenkron, Alexander (1962) Economic Backwardness in Historical Perspective. Cambridge, MA: Harvard University Press.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrialization, Hubs, and Catch-up 161 Johnson, C. (1982) MITI and the Japanese Miracle: The Growth of Industrial Policy, 1925–1975. Stanford, CA: Stanford University Press. Kaldor, Nicholas (1962) ‘Comment on Economic Implications of Learning by Doing’, Review of Economic Studies 29: 246–50. Kaldor, Nicholas (1966) Causes of Slow Rate of Growth in the United Kingdom. Cambridge: Cambridge University Press. Kindleberger, Charles P. (1996) World Economic Primacy: 1500–1990. New York: Oxford University Press. Lewis, W. Arthur (1954) ‘Economic Development with Unlimited Supplies of Labour’, The Manchester School 22: 139–91. Lewis, W. Arthur (1978) The Evolution of the International Economic Order. Princeton, NJ: Princeton University Press. Maddison, Angus (2007) Contours of the World Economy, 1–2030 AD: Essays in Macroeconomic History. Oxford: Oxford University Press. Marglin, Stephen and Juliet Schor (eds) (1990) The Golden Age of Capitalism. Oxford: Clarendon Press. Marshall, Alfred (1898) ‘Distribution and Exchange’, Economic Journal 8(29): 37–59. Marshall, Alfred (1919) Industry and Trade. London: Macmillan. Nayyar, Deepak (2006) ‘Globalization, History and Development: A Tale of Two Centuries’, Cambridge Journal of Economics 30: 137–59. Nayyar, Deepak (2011) ‘The Financial Crisis, the Great Recession and the Developing World’, Global Policy 2: 20–32. Nayyar, Deepak (2013) Catch Up: Developing Countries in the World Economy. Oxford: Oxford University Press. Nayyar, Deepak (2019) Resurgent Asia: Diversity in Development. Oxford: Oxford University Press. Okawa, Kazushi and Henry Rosovsky (1973) Japanese Economic Growth: Trend Acceleration in the Twentieth Century. Stanford, CA: Stanford University Press. Parthasarathi, Prasannan (2011) Why Europe Grew Rich and Asia Did Not: Global Economic Divergence, 1600–1850. Cambridge: Cambridge University Press. Rowthorn, Robert E. and John R. Wells (1987) De-Industrialization and Foreign Trade. Cambridge: Cambridge University Press. Wade, Robert H. (2019) ‘East Asia’, in Deepak Nayyar (ed.) Asian Transformations: An Inquiry into the Development of Nations. Oxford: Oxford University Press, pp. 477–503. Williamson, Jeffrey G. (2006) Globalization and the Poor Periphery before 1950. Cambridge, MA: MIT Press.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
chapter 10
I n dustr i a l Distr icts i n Eu rope Lessons for Industrial Hubs and Industrialization Gioacchino Garofoli
10.1 The Emergence of the Question The emergence of industrial districts (IDs) in Italy in the second half of the 1970s was contemporary with other great changes in regional development and firm size. New industrial regions were emerging in semi-peripheral areas and countries (Scott 1988), the role of small and medium-sized firms (SMEs) increased, especially in late-developed countries (Fuà 1980), and Perroux’s development poles tool started to show great difficulties in peripheral regions. All these phenomena were working simultaneously in Italy (Garofoli 2009: 37–9). The success story of small firms and their territorial organization in Italy, and the rediscovery (Becattini 1979) of the alternative industrial organization model introduced by Alfred Marshall in the late nineteenth century obliged scholars to reflect not only on the Italian case but even on the presence of this model of production organization in other European countries. Even in an historical perspective, industrial districts were important in the second half of the nineteenth century and the beginning of the twentieth not only in Great Britain but even (if not especially) in the second comers (the French and German cases). The progressive disappearance of the industrial district model in northern European countries in the 1920s–1930s and 1950s–1960s was linked to the extension of the model of large Fordist firms (Piore and Sabel 1984; Sabel and Zeitlin 1982; Zeitlin 1985), whereas small-scale production survived in peripheral areas. At the beginning of the publication of ID literature in the late 1970s and early 1980s, the enthusiasm prompting Italian scholars to organize field research,1 visit firms, hold discussions with social actors and local public bodies looking for information and evidence, and analyse 1 In addition to Becattini’s Florence group, there were two other research groups working (at the Universities of Modena and Pavia) on small businesses and industrial districts in those years (Brusco 1982; Garofoli 1981, 1983a).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrial Districts in Europe 163 informal economic data and inter-firm links to understand the workings of an industrial production model based on a high density of small firms and flourishing entrepreneurship was regarded as strange by northern European scholars. At the same time Italian scholars looked beyond Italy, seeking to understand whether this model of production organization was specific to the Italian case or should be a more general alternative industrialization and development path. The first mapping of industrial districts in Europe was produced in 1979 with a research report commissioned by Prato industrial district2 and presented at an international conference at Durham University in 19813 and—in a wider and more complete analysis—the first publication in an Italian journal in 1983 (Garofoli 1983b).4 International discussion of the Italian model based on IDs started in 1984 with the publication of Piore and Sabel’s book The Second Industrial Divide. The analysis of IDs produced another novelty for the debate and for policymaking: the economic structure and the development features of IDs were underlining the crucial role of territorial organization in this particular model.
10.2 The Dynamics of Industrial Districts in Europe 10.2.1 The Success Stories of Southern European Industrial Districts In the middle of the 1970s many Italian small firms showed they were able to work in an autonomous way in relation to large firms, working not only for local and regional outlet markets but even for national and international markets. Often they were highly concentrated in specific locations (frequently with more than 1,000, sometimes 3,000–5,000 firms). The economic game was quite complex, involving a set of different firms specialized in specific items or in specific phases of production along the production filière. Almost 100 industrial districts were recognized in Italy at the end of the 1970s and shown in the first map of IDs (Garofoli 1981). External conditions were favourable: European economic integration and the formation of a European market helped the progressive international openness of small and medium-sized Italian firms.5 The progressive openness of the European market acted as 2 The commission, officially from the Prato local bank, was addressed to Censis (a leading Italian research centre on social issues) and subcontracted to a young (risk-taking) researcher. The results were presented in Artimino (near Prato) in 1981. This story is important because it underlines the links between innovative research (in this case inductive research analysis) and the need for IDs and LPSs to understand their international position. 3 The paper was submitted to an international journal but it was rejected because neither the referee reviewers nor the editor understood the significance of the proposed paper. 4 The article was soon published in Spanish and a few years later in Portuguese (in a Brazilian journal). 5 It is important to remember that the European market for Italian district firms in the furniture industry only began to open at the beginning of the 1970s, but even for footwear and clothing European exports had already started. The only postwar open sector (amongst Italian ID production) was t extiles.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
164 Gioacchino Garofoli an enlargement of the market for companies, giving space to division of labour amongst firms and helping the organization of flexible relationships amongst local firms, producing a balance of cooperative and competitive forces. All this provoked a critical mass of the territorial production system, giving opportunities to SMEs without any threshold obstacles for each specific firm. This transformation process provided opportunities for a parallel growth in other competences (technological issues, organizational mechanisms, attention to markets, product differentiation orientation, production quality). This gradually led to a production quality orientation and an upgrading strategy by SMEs in the industrial districts. In the golden age of Italian and European industrial districts markets were enlarging (rising competitiveness of existing firms in industrial districts) while dynamic pursuit of continuous upgrading was also occurring (moving away from competition based on production costs). The dynamics of territorial development led to the production of new competences and increasing strategic opportunities. There has been another crucial and structural factor: the lack of concentration in the retail sector, especially for textiles and clothing, footwear, and furniture products, has created a multiplicity and variety of manufacturing firms (in accordance with the rules of monopolistic or imperfect competition), especially in Italy in the 1970s and 1980s (and later on even in Spain and Portugal), favouring access of small and medium firms to final markets (cf. Brusco 1989). In the 1980s and 1990s, this enabled (especially in southern European countries) different and cooperative strategies amongst firms with the creation of alternative distribution chains to reach final markets. All this explains the development and rise of the industrial district model in Italy, and later in Spain and Portugal. The lack of competition from extra-European countries till the early 1990s favoured the maintenance of high levels of manufacturing employment in labour-intensive sectors in southern Europe, with only a limited progressive shifting (displacement) of jobs from French and partially Italian districts towards Spanish and Portuguese ones. The agglomeration of firms in industrial districts is supported by the realization of external economies mainly linked to the historical localization of skills and technical knowledge and capabilities to which local firms have access. The large variety of labour skills coherent with labour demand leads to a well-behaved local labour market: firms have free access to a wide and articulated labour supply with low recruitment costs. This also allows high labour mobility which fosters the diffusion of technical and organizational knowledge amongst local firms. The large number of local firms and their increasing openness to international markets (especially to the European market) favours division of labour amongst firms and increasing specialization at the firm level. The variety and high-quality levels of technical knowledge and capabilities support both the division of labour amongst firms and the reproduction of entrepreneurial capabilities, because the majority of the new entrepreneurs within industrial districts come from the ranks of technicians and highly qualified workers. The dynamics of the model and the ability of firms to enter new markets without competing on the low cost of labour showed the efficiency of SMEs in industrial districts and the innovation capability embedded within this organizational model. This definitively
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrial Districts in Europe 165 broke the old paradigm under which small firms represented the weak side of dualistic economies. As the success story of Italian industrial districts started to be diffused amongst international scholars, thanks to international conferences and seminars,6 and visits to firms and districts by foreign scholars and practitioners, the idea of the existence of a new (or the rediscovery of an) alternative way of industrial organization started to enter both the debate and local policymaking.
10.2.2 The Different ID Dynamics in Northern and Southern European Countries After the emergence of industrial districts in Italy, some scholars started to look at the presence of this model of production organization around other European countries. In Spain and Portugal (Costa Campi et al. 1993; Silva 1992; Figueiredo et al. 1994) but also in southern France (Courlet and Judet 1986; Courlet and Pecqueur 1992; Courlet and Hsaini 1997) and other Mediterranean countries, a high presence of local SME systems was found (at least twenty-five to thirty areas in Spain, ten to fifteen in Portugal, and around thirty in France),7 often corresponding to true industrial districts. In some cases they resulted from the progressive transformation of diffused handicraft productive culture (for example, Ubrique in Spain, specializing in leather production, or Kastoria in Greece, specializing in fur production) or of a historical tradition of local raw materials processing (cf. Porriño in Spain, specializing in granite processing). In other cases they came directly from an ancient professional specialization (for example, cutlery in Thiers, France), from a long and consolidated experience of international markets (see Mazamet in France, specialized in the first phases of wool processing), or from the progressive accumulation of professional experience in mechanical engineering (see Cluses in France, specialized in the production of micro mechanical components). Some regions are characterized by the enduring presence of important industrial districts: the Valencia region of Spain with the shoe district of Elche and Elda; the toys district of Ibi-Onil; the textile district of Alcoy; the tiles district of Castellón de la Plana; the Rhône-Alpes and Jura (France) with Cluses, Oyonnax (plastic materials), Roanne (textiles), Morez (spectacle frames), St. Claude (wood products and transformation), and Romans (shoes until the end of the 1980s); northern Portugal with textiles and clothing districts (Guimarães, Santo Tirso and Vila Nova de Famalicão in the Ave valley; Feira and Agueda in Aveiro region); shoe districts (Felgueiras, S. João da Madeira, Feira and Oliveira de Azemeis in the Entre-Douro-e-Vouga region); and furniture districts (Paredes and Pacos Ferreira in the Porto region).8 6 At the first two Aegean seminars on regional development in Naxos in 1983 and Lesvos in 1985, several scholars belonging to different schools, social sciences disciplines, and countries were present; amongst them the future editors of Regional Studies, European Urban and Regional Studies, and European Planning Studies—some of the journals in which the theme of industrial districts has been often discussed in the last twenty years. 7 It is always difficult to make reliable estimates of the number of industrial districts in various countries, due to limited accurate statistical information and the different methodologies of scholars. 8 The production and employment level in some of these industrial districts is quite significant. More than 2,000 textile, clothing, and footwear firms with around 100,000 employees are localized (at least
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
166 Gioacchino Garofoli By contrast, in northern European regions there are few examples of IDs. In the United Kingdom, for instance, of the historical industrial districts existing in Marshall’s time or even during the 1930s and 1950s, few have preserved elements of a district organization in the 1970s and 1980s, for example, Nottingham with knitting, Northampton and Leicester with footwear, Sheffield with cutlery, and Bradford and Halifax with the wool industry. Germany’s situation is similar: some district elements were still at work in Solingen (cutlery), Pforzheim, Göttingen, and Tüttlingen, whereas historical districts such as Pirmasens (footwear) had disappeared.9 At the same time, new dynamic agglomerations are forming, in both the United Kingdom and Germany,10 based on high-tech professional skills, research, and creativity, with typical interactions amongst firms and organizations. Interactive learning, complementary and cooperative links, and the formation of new professional skills and competences are the specific features of these new agglomerations of firms. The increasing importance of large firms in industrialized countries and regions in the postwar period forced many small firms to become their sub-contractors (Sabel and Zeitlin 1985; Zeitlin 1985). Furthermore, the increase in dependent workers and the changed role of SMEs reduced their ability to maintain specific resources and variety of production which, in the long term, reduced the entrepreneurial formation rate. The job opportunities in large firms have changed the expectations of the younger generations, reducing the potential for autonomous jobs. The industrial policy of some central states added to these tendencies and favoured the expansion of large firms looking for scale economies. In France, industrial policies aimed at financial concentration and financial incentives for the localization of large firms in peripheral regions followed Perroux’s ‘growth pole’ theory (cf. Sabel and Zeitlin 1982; Zeitlin 1985; Courlet 2008). All this reduced the role of SMEs and created a negative environment for the reproduction and the survival of existing industrial districts in Northern Europe between the two world wars and especially during the 1950s and 1960s. Finally, the de-localization process of large firms looking for cost reduction and unlimited labour supply in Mediterranean countries during the 1970s completed the de-structuring process in the northern IDs, with a reduction of linkages amongst firms at the local level and the opportunity to enlarge relationships amongst firms looking for specific competences.
10.2.3 The Introduction of Industrial Districts into the Economic Debate in Europe We will now briefly discuss the roles of scholars and policymakers in introducing the concept of industrial districts into different European countries. were so till the end of the 1990s) in Portugal’s Ave valley. But even the size of Spanish districts is quite significant in the same years: the footwear district in Elche counted around 20,000 employees and the textile district in Alcoy showed an agglomeration of more than 600 firms with 10,000 employees. 9 In Scandinavian countries there are few cases: Gnösjio is almost the only one in Sweden; and Jæren, Horten, and Arendal are cases of some relevance in Norway (see Johannisson 2009). 10 The areas characterized by new technologies, such as Cambridge, but also media services in Cardiff, are interesting cases. Some interesting cases in Germany are biosciences in Heidelberg-Mannheim, hightech machinery in Aachen, medical electronic equipment in Freiburg and Tübingen, and multimedia technology in Köln.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrial Districts in Europe 167 In France, Claude Courlet and the research group working in Grenoble (Courlet and Judet 1986; Courlet and Pecqueur 1992), and the sociologist Bernard Ganne (1990, 1992) were promoting the awareness of the role of IDs, and geographer Pierre Houssel was able to write a paper on the industrial district of Prato and on Third Italy at the end of the 1970s. Maria Teresa Costa Campi (Costa Campi et al. 1993) in Barcelona and Josep-Antoni Ybarra (1991) in Alicante were the first economists to introduce the concepts and analysis of IDs in Spain, although Antonio Vazquez Barquero in Madrid had previously written on policy interventions on IDs. Later Joan Trullen worked on this subject in detail with his group in Barcelona; the first scholar to engage with the subject was a geographer (Bernabé Maestre 1983). In Portugal the scholars who introduced the debate on IDs were Mario Rui Silva (cf. Silva 1988, 1992; Figueiredo et al. 1994) and José Reis (Reis 1992), but also João Ferrão with detailed work on new forms of regional development. Regional policymakers in some countries soon understood the role of industrial districts in their economies and the opportunity to intervene in their transformation processes. Spanish autonomous regions played a crucial role in regional development, especially the Valencia Community, which introduced several industrial policy tools fostering solutions for SMEs in IDs, such as services centres and technology centres often specializing in specific sectors and localized within the IDs.11 National policymakers entered the debates later: the Ministry of Industry in Spain financed research on industrial districts throughout the country at the beginning of the 1990s (cf. Costa et al. 1993); DATAR—now DIACT—having commissioned research on industrial districts and local productive systems (LPSs) in France (Courlet and Hsaini 1997) and later in Europe (Courlet et al. 2000), worked with IDs on restructuring, favouring an innovation orientation. State institutions at every level were involved, enabling a balance between national planning and ‘bottom-up’ approaches. During the 1990s in France, following the Italian experience, several meetings between scholars, practitioners, and policymakers were organized to discuss the problems and prospects of IDs and related policies.12 The participation of local actors, both public and private, in the governance process of industrial districts in France, has been quite high. The need for mobilization of local actors and, at the same time, local coordination through regional networks, was clear enough. This explains the success story of the Industrial Districts Club (CDIF), which was established in 1997 with its first presidency in Cluses. Although IDs never attained the high level of growth experienced in Italy, Spain, or Portugal, France was able to orientate industrial and development policies. Since the end of the 1980s and beginning of the 1990s, the majority of policymakers at both national and regional levels have perceived the opportunities to follow different development paths and the existence of alternative models to large-firm organization.
11 Impiva, a public company created in 1984, played a crucial role, acting as a coordinator of a network of eleven technology centres in the region. 12 See the meetings in Cluses in 1992 (combined with an important international conference in Grenoble), in Mazamet in 1994 (attended by representatives of thirty European areas based on small firms), and in Beziers in 1998, organized by a network of European researchers/scholars and universities.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
168 Gioacchino Garofoli
10.3 The Industrial District Model 10.3.1 The Working Mechanisms Industrial districts explain much of the success of small-scale industry as well as much of the size structure of Italian industry in the 1970s and 1980s. Nevertheless, the productive organization of the industrial district should not be confused with the decision-making behaviours, structural characteristics, or organizational methods of small business. The company in the industrial district exhibits behaviour and adopts types of relationship (with other companies, with the market, with the external social-economic context and environment) completely dissimilar to those of ‘isolated’ small businesses that do not operate within a specific system of interactive relationships (see, for example, Brusco and Sabel 1981). The industrial district (ID) represents a social construction: it is a model of both production organization (alternative to those based on the predominant role of the large enterprise) and social and human relations. The district is ‘a socio-territorial entity characterized by the active coexistence . . . of a community of people and a population of industrial enterprises’ (Becattini 1989). From the point of view of production organization, there are typically many (sometimes thousands of) SMEs specializing in different stages of the production cycle of a good, which requires cooperation and horizontal integration of firms. All this generates a complex production system based on the balance of competition and collaboration between companies. Analyses of IDs show the role of the territory as a place of social interactions, reproduction of knowledge, and the production of specific skills and resources which cannot be used elsewhere and which therefore represent competitive advantages for companies located there. The territory is the meeting point between development actors, a cooperative space between companies, the place where the division of labour between companies is decided, and the meeting point of market forces and forms of social regulation (Garofoli 1991: 53–5, 1992: 4–5). IDs offer opportunities for dynamic competitive advantages to companies that are located there. They provide external economies to the enterprise that are internal to the territory, like a public good with free access for the various economic actors, often at zero cost (or close to zero) to businesses. The presence of specialized companies for both specific services and production phases enables the accumulation of complementary knowledge and skills, allowing not only small and medium-sized enterprises to focus on the areas where they have greater competitive advantages but also to solve problems for complementary businesses in the ID. As a result, hierarchy is replaced by market, vertically integrated production by horizontal economic integration and specialization (as a result of widespread division of labour between com panies), and exclusive use of competences inside each firm by access to a pool of specialized resources. This model is more flexible and presents lower fixed costs. The local firms’ system, through the dialectical game between competition and collaboration, facilitates attention to change and innovation, produces social capabilities at the level of the community of businesses, organizations, and institutions operating in the local area, and offers small firms (through the reduction of transaction costs) opportunities for ‘external growth’.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrial Districts in Europe 169 Interactive relationships between businesses, as well as between businesses and the local milieu and environment, become the alternative strategy to size growth; economic organ ization is increasingly becoming a relational economy based on the production of public goods. The local business system is a model of local regulation that determines cooperative relationships between firms and the local system which can be added to the collaborative relationships between companies. The third relevant issue in the analysis of IDs concerns the role of external economies for the firms (internal to the district). External economies represent the set of factors which constitute Marshall’s ‘industrial atmosphere’ and which are difficult to assess ex ante but which can explain ex post why companies in regions and countries with higher costs in standard factors perform well and gain market share. External economies are produced by historical investments (public and private) in specific knowledge and skills that determine the existence of specific resources in the territory of the district and which are not transferable to other areas. They are incorporated, therefore, in tacit knowledge (Becattini and Rullani 1993) which is transmitted exclusively through interpersonal and social relationships, in values and rules of conduct, in trust relationships, in low costs of access to specific resources, and low transaction costs. In other words, there are real public goods, access to which is guaranteed to all local firms (often at no cost), which therefore have a strictly contextual character (i.e. linked to the history, culture, and behaviour of the community of people and of the business population that belong to that territory). The working mechanisms of industrial districts have been described several times in the specific literature (see Becattini 1987, 2000; Brusco 1989; Garofoli 1983a, 1991). It is sufficient to list the following structural features: (a) High production specialization (b) High division of work between local firms, with substantial production interdependencies (c) A large number (sometimes thousands) of firms and the absence of a leading company (guaranteeing a high level of firms’ autonomy in the system) (d) High specialization at the enterprise and production plant level (which often allows large economies of scale for each single process to be achieved) (e) Efficient information circulation system (knowledge is a common asset for companies) (f) High level of skills and professional qualifications of workers (g) Face-to-face relations between economic actors facilitating the dissemination of organizational and technological improvements. In section 10.3.2 we highlight opportunities or trajectories for transformation, emphasizing the dynamic features of the ID as a model in continuous evolution, and drawing some general conclusions on the relationship between small firms and economic development.
10.3.2 The Dynamic Model The first dynamic feature of the ID is the growing division of labour between local firms. This generates increasing production interdependences between companies, with the
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
170 Gioacchino Garofoli involvement of production branches and complementary sectors in relation to the original production specialization of the area. This means that production interdependences are both intra-sectoral and cross-sectoral, involving new sectors and different production branches, both from a processing and technological point of view (cf., for instance, the introduction of artificial fibres and machine tools production in textile districts). This allows for a horizontal spread between sectors of technological innovation, allowing interactive and cumulative effects that strengthen the efficiency (and competitiveness) of the entire territorial production system. The diffusion/imitation effects unintentionally spread knowledge between companies and produce critical mass in the local system. Moreover, the effects of complementarity between companies (through voluntary decisions on formal and informal collaboration and the circulation of information) are determined by the continuous introduction of problem-solving processes. Intermediate institutions (as interface structures) which usually play the role of ‘versatile integrator’ (Becattini 2000: 106) also intervene, combining different know-how and skills to solve problems common to many companies and enabling the upgrading of businesses and the local system as a whole. In other words, the interrelationship game of local institutions and organizations enables the governance of complex processes such as those of continuous change and transformation of the ID. The progressive accumulation of knowledge and technical skills at the local level leads to the emergence of local dynamic competitive advantages that differentiate local production organization from that of other areas and do not commit firms to a competitive game based solely on costs (and prices). The external economies enable use of specific rather than standard resources. The third dynamic feature is the high rate of formation of new enterprises, which has long differentiated the behaviour of IDs from other areas, with the possibility of production and reproduction of organizational-entrepreneurial (OE) skills reinforced by learning mechanisms in the field, high social mobility, and the spread of economic calculation capabilities (Fuà 1983) and social values that reward and promote social innovators and risk-taking people. The fourth and final strategic feature of the ID is the increasing complexity of the local economic system, with the progressive involvement of new production sectors and new branches (both in industrial production and services) directly linked and stimulated by the interaction with the specialization sector. The increasing complexity of the territorial production system, which would statistically appear to give rise to a weakening of the specialization sector, leads in reality to a deepening and strengthening of the local business system. This complexity accentuates dynamic interaction, increases the strategic capacity of local firms, and ultimately multiplies external economies, allowing dynamic competitive advantages in comparison with non-district enterprises (Garofoli 1991).
10.3.3 Social Capability and Some Preliminary Conclusions In IDs, the low barriers to entry, favourable market conditions, and greater production of OE skills (Fuà 1980) due to social dynamics and on-the-job learning mechanisms typical of lean production organizations favour self-employment ventures, and this is facilitated by
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrial Districts in Europe 171 financial guarantee mechanisms typical of the ‘extended family’, especially in rural areas (Fuà 1983). Self-employment business start-ups are considered a natural outlet for the career of the employee (Solinas 1996), and self-employment is further facilitated by the low costs of leaving the company in possible difficulties. In short, there is no waste of the OE factor in the industrial district (Garofoli 1983a, 1991). The ID organization is economic for the OE factor thanks to the ability to delegate to the market and to the external organization (via transactions with other companies) the management of functions in which the company has accumulated less knowledge and which it is therefore able to use at lower levels of efficiency. This also facilitates access to new activities because the OE capabilities needed for the highly skilled and networked company (with the local business system) are lower. The territory, through the dynamics of relationships between companies, densely intertwined social relations, the reproduction of the values of work, and the recognition of the individual success of the entrepreneur, generates social capabilities, facilitates the entry of new entrepreneurs, and increases the local system’s ability to respond to the problems of the enterprise (such as efficient governance of the labour market and effective relationships between education/training and work). Territorial relationships determine a collective learning process and therefore produce social capability and innovative dynamics through real interactive learning processes. Research in industrial districts has eliminated a widespread bias and some stereotypes— amongst both economists and policymakers—about production inefficiency and the marginalization of small business, based on an inappropriate extension of the principle of economies of scale. Analyses of IDs have highlighted not only opportunities to achieve high levels of production efficiency in small firms specialized by the specific processing phase (see Brusco 1975, 1989) but also the strong innovative potential of district enterprises. Economic efficiency and innovation can therefore be combined with the small business dimension, provided it is part of a network of relationships that allows access to a pool of specific and valuable resources (Garofoli 2003). It is important to remember a quite crucial issue often forgotten in the international literature and debate on IDs: the efficiency of ID firms. There are some works which strongly clarify the performances of firms (and the differences by firm size) in IDs. Labour productivity and profitability are quite high (very high in the ‘golden age’ of IDs), usually higher than in firms located outside the IDs, and there is no direct relationship either between firm size and labour productivity or between firm size and profitability (Garofoli 1994; Signorini 2000; Coltorti and Garofoli 2011).13 This means small firms in IDs are quite efficient and very profitable. This point, obviously, opens a large space for new interpretations. This section has underlined both the opportunity to follow different development paths and forms of industrialization, even for small firms, and the crucial difference between the interpretation of the competitive factors and the ability to maintain a sustainable development process (i.e. endurability of the model). Regarding the second question, the ID experience 13 It is quite important to remember that in the case of the research on the mechanical district of Lecco (Garofoli 1994) a balance-sheet analysis was organized in empirical research for the first time both for IDs and territorial development. This analysis accompanied field research, based on interviews with firms (to collect qualitative information on firms’ behaviour and strategies) and on existing statistical data at the territorial level.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
172 Gioacchino Garofoli explains the contradiction between short-term, traditional and long-term, dynamic competitive factors, showing the relevance of dynamic competitive factors and the crucial role of a shared vision of the future which could enable the introduction of collective actions to strengthen the territorial system.
10.4 The Rise and Decline of Industrial Districts in Europe 10.4.1 The Golden Age of Industrial Districts The rising importance of industrial districts in Italy emerged between the end of the 1970s and the beginning of the 1980s. Their positive effects on employment, wage levels, share of international markets, rate of profits, industrial investments, and number of firms spread confidence in this type of organization. It was a golden age for Italian IDs: their number increased slightly till the middle of the 1990s, with a few extensions in southern regions (see Map 10.1) and with an increase in numbers of both firms and employment. The openness of European markets was providing great opportunities, reinforcing the advantages of specialization at firm level and of division of labour amongst firms. The enlarged market allowed cooperation amongst firms, avoiding fierce competition, and increased opportunities for followers to imitate the success stories of leader and innovative firms (which often played the role of flagship firms). During this phase, the role of buyers helping SMEs access foreign markets was also important. Many district firms progressively learned their strategic position within the new markets. At the outset the export capability of SMEs in IDs was linked with static competitive advantages (lower labour costs in comparison with northern European countries) and with the flexibility of firms to rapidly change products and working processes according to buyer proposals. Quite soon the SME systems in Italian IDs started to upgrade to products characterized by higher product quality, high creative content, and higher prices. At the beginning of the 1980s the average prices of Italian exports in footwear, textiles, and furniture were already higher than those in northern European countries (Modiano 1982). The Italian districts’ competitive factors were becoming more and more linked with dynamic advantages (accumulation of knowledge and professional competences). Spain and Portugal followed, in some cases, an analogous process of development after their entry to the European Community: the great expansion of production and employment in Spain occurred during the 1980s and the beginning of the 1990s, and a bit later in Portugal. Spain and especially Portugal in those years were the countries with the cheapest wages in the European Community and the firms in IDs (thanks to agglomeration economies and the rapid learning process) were able to enter European markets especially where the use of labour both in intermediate and final products of low to medium quality was quite intensive. The positive trend in employment and in numbers of firms in Portugal during the 1990s has been astonishing in comparison with the dynamics of industrial districts in other European countries (and especially in France, Italy, and Spain), and competition on costs
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrial Districts in Europe 173
+ +
+ + +
+
+ +
+
++
+
+
+
Legends
+
- Food products - Footwear and leather products - Rubber and plastic products - Wood and furniture - Other manufacturing products - Metal-mechanics - Non metal minerals’ transformation products - Textile and clothing
Map 10.1 Map of Italian industrial districts in the 1990s Source: Garofoli (1996: table 5.1).
at European level has been determinant. Of course, the competition from developing countries at the end of the 1990s, especially during the last two decades, has dramatically changed the position of Portuguese IDs.
10.4.2 Transformation versus Crisis in Industrial District Organization The ID model represents a dynamic model of production organization and reproduction of knowledge, competences and human resources, innovation, and change (Garofoli 1983a, 1989). Change is the essence of the model that enables its resilience.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
174 Gioacchino Garofoli There are some significant examples of IDs in the nineteenth and at the beginning of the twentieth century (see, for example, the cases of the mechanical products system of Saint-Étienne and the silk district of Lyon), based on the form of social regulation and the intervention of local institutions favouring the introduction of innovation and the pursuit of special and diversified production and quality, in contrast to standardized mass production based on low production costs (Zeitlin 1985; Garofoli 1989). In Saint-Étienne in particular, at the end of the nineteenth century, the city council, which controlled the electricity supply, cut off the supply to companies after a certain time of day to prevent the companies from competing with each other, lengthening the working day and reducing the influence of the cost of labour. This forced the companies to take great care over the quality of their products and the introduction of new ones, preventing them from following strategies based on low labour costs (Zeitlin 1985). Since then, there has been progressive standardization of production or of passive imitation of the organizational systems of large companies, designed in many historical cases to stimulate small-firm systems to lower their production costs, and including them in the product cycle of the large external company, which entails progressively moving away from innovative strategies based on their own specific skills. It is important, nonetheless, to underline the risks of breakdown of the model. In recent times there have been numerous cases of industrial districts that have undergone a process of progressive restructuring due to a shift of core investment from the production side (and from the use of specific working skills and professional competences embedded in the area) to commercial control. This has driven the companies to introduce international decentralization of production in the search for lower production costs. These strategies have progressively interrupted the interrelations and linkages between local companies and the diffusion of knowledge, breaking the local productive cycle, wasting local specific resources, assets, and competences, and progressively stripping the area of its productive capacity. The case of the footwear district in Romans, the most famous and important in the French footwear industry until the end of the 1980s, is quite instructive. Between the end of the 1970s and the beginning of the 1980s, the medium-to-large firms in Romans started to decentralize the simplest, more repetitive, and more labour-intensive production phases to Italian firms which were already specialized by phase within the production cycle. After the decentralization first of the sewing of uppers and then of leather cutting, the Romans firms thought it would be convenient to decentralize even the assembly phases to the Italian firms (more and more efficient due to increasing investments). They believed they should maintain control of the crucial phases in footwear production—engineering and market control. After some years, this process caused the disappearance, and then the lack, of a dense network amongst firms and of a critical mass of technicians and specialized professionals. This network is crucial for the creation and the achievement of new ideas, projects, and technical solutions, enabling the designer’s idea to be transformed into a prototype and the final product. This means the creativity and engineering of the Romans firms was greatly reduced and the firms found it more and more convenient to use the large capability of Italian firms to introduce new products and new collections with a large variety of samples, buying their proposals and products directly, and including them in their collections (with exclusive production contracts).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrial Districts in Europe 175 The Romans firms still thought they would be able to control the trading function; but how could they compete with the largest (and specialized) trading companies that historically were able to control national and international markets? This story of the progressive de-structuring of the footwear district in Romans is a metaphor for the breaking up of this organizational model and provides a good opportunity to look ahead to the possible trajectories for IDs and for strategies and collective actions (and industrial policies) to foster ID organization in Europe. In conclusion, it is necessary to remember not only the increase in firms and employment in the industrial districts in Spain and Portugal during the 1990s but also the great re-emergence of industrial districts in Italy after the devaluation of the Italian currency in 1992–5.14 The introduction of effective anti-inflation economic policies by the Ciampi government guaranteed extra competitiveness for the Italian economy and, especially, for ID firms which had great opportunities to increase their orders. This extra competitiveness created crucial problems for other European firms, even some located in IDs. This could explain the decline of French ID firms that began in the late 1980s (see Map 10.2).
10.4.3 Learning and Innovation As the local system becomes more complex, with the ID having now extensively valorized its specific resources, the development process has taken on all the characteristics of the endogenous development model (Garofoli 1992; Vazquez Barquero 2003; Scott and Garofoli 2007). At this point the territorial system may be able to guide its development and transformation process, although this obviously does not guarantee the survival of the territorial system in the long run. Survival conditions are in fact dynamic: the ID is a model in continuous change (not a simple adaptation) in both its interrelationships within the area and its external ones (with the market, other competing areas, and other systems). All this has obvious impacts on the place of the territorial system in the international division of labour. Change and innovation are, therefore, survival conditions for the territorial system. This highlights the need for a strategy for transforming the territorial system, involving building an understanding of the relative position of the local system and predicting the medium- to long-term evolutionary scenario in order to prepare coherent private strategies and proper public interventions that would increasingly valorize the interrelations and synergy between the production system and local organizations (even education and research) and institutions15 (see also Becattini et al. 2009). The fundamental condition for the consolidation of IDs is, therefore, the adoption of an increasingly systemic production structure that strengthens economic ties between 14 The Italian ID firms even met a lot of difficulties expanding internal production and employment due to the full employment in labour markets in IDs at the time. This, in a sort of contradiction, created conditions for investment and subcontracting abroad. 15 The case of the mechanic (screw cutting for small components) district in Cluses, France, in the 1990s is quite relevant. An agency for territorial development and a Technological Antenna were built, in addition to previous long experience of governance of the mismatched labour market and the management of the relationships between training and work. These kinds of initiatives were organized through a private–public partnership (Courlet 1997).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
176 Gioacchino Garofoli
MAPPA DEI SIL
No
Tourcoing Lille r d Roubaix
Fourmies
Vimeu Viller-Outréeor Ovest Oise St.Quentin Sud Oise Fiers condé
Aube
Fougères
Neufchateau
Troyes
Epinal Remiremont Alto Reno Beasançon Morteau Mulhouse Vosgi
Cholet
Ov
e
st
Va nd
dea an tV
Es
o ur Gl St. Claude Morez
ea
Cluses Reannee Monti Oyonnex Val d’Arve del Thiers Lionese
La Rochelle Limoges
St. Etienne Romans
Millau Mazamet
Le Vigan
Castres
Legends
Lavelanet
- Textile and clothing - Footwear - Watchmaking - Wood products - Plastic products - Metal products - Other manufacturing products N.B.: I SIMBOLI IN BIANCO SI RIFERISCONO A SIL ORMAI DESPECIALIZZATI E/O DESTRUTTURATI
Map 10.2 Map of French industrial districts in the 1990s Source: De Martini and Garofoli (1993).
businesses and with their milieu, the socio-economic context and local environment, so as to produce and reproduce specific resources in the territorial system that will become the strategic localization and development factor. The variables that determine the consolidation of the local system are both endogenous (controlled internally within the area) and exogenous. Amongst the variables that can be controlled within the local system are: (a) Technological-organizational innovation: increasingly assumes the connotations of a continuous process, with cumulative and interdependent effects of a large number of technological and organizational changes, each of which is small. The innovative process in the ID is incremental (à la Rosenberg), so it doesn’t work with big jumps
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrial Districts in Europe 177 (à la Schumpeter). Technological innovation is not solely the result of diffusion processes, but very often is itself the product of the territorial system, with the introduction of new techniques, especially in cases where machine tool producers are present, even driving the shifting of the technological frontier. (b) The information system: efficient and rapid circulation of information, widespread knowledge of markets, free dissemination of information on usable technologies, inputs, and new materials are the basis of a correct assessment of the prospects for the development of local firms. (c) The ability to control the market: the strengthening of the trading capacity of the territorial production system in the national and international market is one of the determining factors for the autonomy of the local system and at the same time a continuous opportunity for the introduction of new products through production differentiation and diversification. (d) Forms of social regulation: operate outside the market, depend on a fruitful integration of local institutions and the local economy, and respond to common needs of companies. In the development and transformation process of IDs, for example, some quality control centres, technology centres, and real-service centres have taken a decisive role; the role of vocational training centres and school–work fitting institutions was also important. Finally, in some cases, the introduction of institutions regu lat ing competition between companies has facilitated the introduction of new products and processes rather than competition on production costs and, indirectly, labour costs. (Garofoli 1991, 1992) This takes us to the last topic to be discussed: the processes of change in IDs and competitive factors. The ID represents a dynamic model that must incorporate and metabolize the process of change, adapting to (or, better, anticipating) changes in technologies and markets. In this process, the construction of strategic and decision-making capabilities at the system level becomes crucial, with a potentially significant role for ‘project enterprises’ (Becattini et al. 2009: xxviii) that foster the orientation and culture of change management. By pairing these considerations of the developmental paths of IDs with analysis of the competitive factors of this organizational model, we will have strong indications on what to do in terms of political strategies to support district enterprises (see also Becattini et al. 2009: xx–xxi). It is clear throughout the Italian district literature that the competitiveness factors of district companies are identifiable in continuous innovation (especially with the introduction of new products), in their pursuit of quality production, and in their creativity and ability to solve the problems of complementary companies along the production chain. Competitiveness on prices and unit production costs must therefore be carefully avoided.
10.4.4 Lessons from the Historical Perspective: The Risks of Collapse of the Model What is striking about international comparison of IDs is the immediate perception, in the 1980s, of the processes of de-structuring and crisis that characterized famous IDs in the
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
178 Gioacchino Garofoli days of Alfred Marshall (Manchester, Leeds, Birmingham, Sheffield, Solingen, Lyon, amongst others) on the one hand, and the emergence of new small business systems, especially in the countries of southern Europe, on the other (Garofoli 1983b, 2009; Zeitlin 1985, 1995, 2007). Amongst the causes of the crisis for some famous IDs was the erroneous strategy of placing the local system in the territorial division of labour and of the uncritical transition from flexible production to mass production (Zeitlin 1985). This is historically linked to both industrial policy interventions at the national level (the most resounding case is that of the French state’s campaign for business mergers and industrial restructuring in the 1960s—see Zeitlin 1985), and the choices of individual producers (such as the abandonment of strategies for pursuing product differentiation, quality and innovation, as in the Sheffield case, in an attempt to mimic large enterprises by standardizing the product and reducing production costs, often turning into passive sub-suppliers of large enterprise—see the cases of Birmingham and Saint-Étienne; Zeitlin 1985; Garofoli 1991). The last, but not least, crisis factor for some historical IDs is the failure of (and/or failure to renew) the forms of social regulation that had previously contributed to the success of those local systems. In particular, the loss of the role of local institutions and local government in the defence of mutual interests (competition discipline to promote quality production and the introduction of innovation, quality control systems, wages stability governance, and so on—see Zeitlin 1985). In recent decades there have been numerous cases of IDs which have undergone a process of progressive de-structuring, as a result of a shift of focus from productive issues (and from the use of specific and new work skills) to commercial control. This has prompted companies to introduce international production decentralization strategies in the search for lower production costs (via the use of work in low-wage countries). These strategies have progressively disrupted the interrelationships between local companies and the dissemination of knowledge, disrupting the local production circuit with progressive dismantling of the production capacity of the area. The risks of the industrial district model breaking and collapsing may be different (Garofoli 2003) and are generally related to the adoption of strategies inconsistent with the positioning of these systems in global competition and to the lack of investment in reproducing dynamic local competitive advantages. The possible risks of implosion of the model of production organization based on ID would seem, in the light of historical experiences as well as the trends of recent years in many Italian and foreign IDs, to be the following: • failure to reproduce external economies caused by lack of private and public investment in the conditions of the economic context and in the competitive advantages of the local system, including investment in upgrading competences and know ledge in the institutions connecting firms and the research system (see also Becattini et al. 2009) • loss of ability to manage and ‘metabolize’ (at the territorial level) technologicalorganizational change, that is, to internalize technological and organizational knowledge, making it appropriate and coherent with the skills management of local businesses • decentralization of processing phases at international level, reducing capability of maintaining production interdependence at the local level, multiplier effects, and fall out mechanisms on complementary firms till these gradually disappear
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrial Districts in Europe 179 • lack of reproduction of professional, entrepreneurial, and organizational skills that represent the determinants of the social model at the heart of the ID (see the recent increasing problems with business and generation replacement, especially in family businesses, in Italian IDs) • insufficient change in the professional structure of employment (and the demand for labour) in relation to the expectations of the labour supply, especially the increasingly educated youth who are less and less attracted to traditional professions. This represents a typical socio-cultural contradiction of the small business system that matures whenever the territorial system does not perceive in time the need for change and the introduction of new professional skills and competences in local firms which are necessary for consistent strategic positioning. The model is dynamic and presents many possible transformations and bifurcations, as we shall see in section 10.5.2. This means territorial and development differentiation even in areas which belong to the same specific model. This is the role of cities and territories: they differ by nature, according to changing external conditions but even as regards strategies of territorial actors and stakeholders, through the effective interactions of challenges and opportunities.
10.5 The Forces of Globalization in the International Market: Impact on Industrial Districts 10.5.1 The International Context The last two decades have seen the entry of new competitors (especially from Asia) in the sectors and markets in which European industrial districts specialized. The phenomenon began with the entry of international leader firms to emerging industrial countries, initially with direct investments in ‘green plants’, and later with the replacement of national and local suppliers with alternative suppliers located in cheap-labour countries, with a wide international decentralization process. The emergence of the international leader firms in new industrializing and developing countries produced widespread effects both on the internal (in Western countries) and external (in developing, mainly Asian countries) restructuring of the manufacturing industry. The main consequence has been the organization of global value chains (GVCs) and the internationalization of suppliers’ markets. The competitive game changed dramatically, not only in the markets for final goods but even in suppliers’ and sub-suppliers’ markets, increasing international competitiveness amongst suppliers and the fragmentation of production in both Western and emerging countries. The negative effects in EU countries have been widespread, especially for manufacturing production of low and medium quality, with a dramatic decline in the number of existing
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
180 Gioacchino Garofoli manufacturing firms and in European manufacturing employment. Sharp international competitiveness amongst suppliers disadvantaged European countries, due to their social and historical circumstances, disrupting the maintenance of the structural organization of the division of labour in IDs. The process of production internationalization pushed even Italian and Spanish leader firms to follow a similar strategy, increasing their focus on direct investment abroad or using more and more foreign suppliers, often in those stages of production located (through GVCs) in emerging countries. In recent years, however, high-quality firms with globally recognized brands have been pursuing an opposite strategy—investing directly (often with ‘green plants’) in the territories of IDs (especially in Italy) where the quality of existing skills and competences are higher (Garofoli 2018). This represents an interesting challenge for the behaviour of SMEs and public institutions in Italian (and European) IDs, and hence in the prospects for IDs. Some positive novelties are emerging even in ID firms. First, even with some delay, there has been a transformation in the job structure of ID firms: new skills and competences have entered medium firms (e.g. several architects not only for design but even for the firm’s image and communication in furniture-district firms, tertiary jobs for services, trading offices, financial issues, direct sales organization, and monitoring of international markets). This has often produced a size increase in leader firms and a restructure of existing IDs, but it risks increasing the hierarchical organization within the ID and reducing the autonomy of SMEs and of the overall territorial system. There is a serious risk (at least at a cultural level and in the collective awareness within the district) of a return to the ‘single model’ that is the diffusion of the competitive game based on labour costs (and labour reserve), with all the consequences for regional and international competitiveness and for the reduction of autonomy in a large number of territories (‘lost in translation’), which due to naïve economic policies (for example, the European austerity policies) and conformist behaviour by local actors, were pushed into an incoherent game. The increasing role of multi-national enterprises (MNEs) is the other side of the question. Not only were they the basis of the expanding industrial employment and production in emerging countries: the new industrial regions and countries often failed to achieve autonomy in their transformation process. Moreover, since they began to take an interest in emerging industrial countries, MNEs have sought to dominate and control the great expansion of the market, especially in Asian countries, but in all emerging industrial regions generally. This new phenomenon raises two main issues for possible development strategies to be followed in European industrial districts. First, the new industrializing countries also show strong agglomerations of firms, some of them with features similar to industrial districts. In India and China (but also in other countries, from Pakistan to the Philippines) there are many clusters of specialized SMEs (Schmitz 2004; Gereffi et al. 2005). The division of labour and cooperation amongst firms tends to be more limited than in Europe. In any case export performances in these new countries are sometimes related to the advantages of industrial district organization, since the advantages of the new entrants are mainly linked with agglomeration economies. Second, the effect of the international competition introduced in the last twenty years is quite clear: European districts which were still based on low-quality production have been collapsing. Static competitive factors alone cannot enable the survival of industrial districts in Europe.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrial Districts in Europe 181
Table 10.1 Strategies, bifurcations, and transformations in industrial districts Strategy objectives
Trajectories, bifurcations, and transformations
Cost reduction
‘Low road to development’ de-structuring of IDs
Scale economies
Formation of financial groups
New markets penetration and new products
External networks and strategic cooperation
Looking for specific resources and competences
Localization of large external firms and MNEs
Valorization of external economies
‘High road to development’
Source: Garofoli (2011: table 4, p. 183).
10.5.2 Typologies of Transformation in Industrial Districts Within the range of evolutionary trajectories, different firm strategies can be identified that have been widely introduced in different IDs to react to the challenges of the international market, leading to a range of different bifurcations and transformations of IDs (see Table 10.1). Looking at the transformation processes in several industrial districts in Europe in recent decades, it is possible to select the following typologies of transformation: a) A progressive de-structuring of the district, due to the pursuit of a cost-reduction strategy looking for delocalization of several phases of production or looking for new suppliers abroad in poor countries with unlimited supply of labour. A process of international decentralization determines the cutting out of phases of production, reduction in the use of labour, and the progressive disappearance of specific skills and competences (cf. the cases of Romans, France, and Manzano, Italy). b) The formation of financial groups due to the pursuit of a strategy of scale economies on financial, commercial, and organizational issues. This strategy is mainly followed by ‘brand name’ firms, and it is often combined with a strategy of increasing internation alization. This produces a progressive size increase of leader firms within industrial districts. The likely consequence is often the introduction of a hierarchical organiza tion and internal restructuring, leading to the use of exclusive specialized suppliers (which operate, then, with only one client) located within industrial districts (see the cases of the spectacle-frames district in Cadore, of the leather district in Tuscany, and recently of the footwear district in Fermo). c) The pursuit of a strategy of new markets penetration and new products introduction just to escape from cost competitiveness. The strategy tends to reinforce external networks (for the introduction of external knowledge) and strategic cooperation amongst firms located in other areas (often in other industrial districts), enabling productive diversification (see the case of Lecco and, to some extent, the cases of Tüttlingen, Cholet, and Saint-Étienne).16 16 The cases of Cholet and Saint-Étienne are interesting because they were able to introduce new internal networks involving different sectors and competences amongst local firms, and transforming the economic structure of the areas. Cholet moved from production of clothing and footwear for children into an integrated local system for children’s products and services, involving sectors and competences outside manufacturing (such as child care). Saint-Étienne moved from textiles to health-textile products and health and personal care equipment industries.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
182 Gioacchino Garofoli d) The historical creation of specific resources and competences within industrial d istricts which can sometimes attract external firms looking for these specific resources (often crucial for creativity, engineering, and the introduction of new products); when this process determines the localization of external large multinational firms there is a risk of the introduction of hierarchical relationships (see the cases of the sport system in Montebelluna and the shoe district in Strà and Riviera del Brenta).17 e) The last typology is characterized by the progressive valorization of external economies, continuously introducing new competences and skills which foster creativity, quality of production, and the introduction of innovation, increasing the value added per employee and escaping from the cost pressure (see, mainly, the case of Biella at least until the end of the 1990s and partially the tile district around Sassuolo). This trajectory is a veritable ‘high road to development’. The proposed typology may accommodate the restructuring processes experienced in the last twenty to twenty-five years by many European industrial districts facing the challenges of international competitiveness. Very often, the IDs were obliged to react both to cost pressures (especially due to the entrance of new competitors) and to the need to introduce new products to move ahead with their international position. In the first case, when the process of international decentralization of production and delocalization predominates, the ‘low road to development’ is a quite likely result, with a squeeze on labour costs and increased hierarchical ties in peripheral clusters. This produces, then, a strategic and technological divide between the group of innovative (often larger) firms in the advanced industrial districts and the group of labour-intensive and traditional firms (both in IDs in advanced countries and, especially, in clusters of peripheral countries). In the second case, the introduction of external knowledge (through strategic collaboration agreements with firms located in different regions) into industrial districts could enable technological discontinuity, avoiding limitation to incremental innovations (Garofoli 1993, 2003).
10.6 Questions for Today: The Impact of the Economic Crisis and the New Internal Restructuring in Industrial Districts 10.6.1 The Governance of Transformation Processes In industrial districts the dynamic forces of the development process intervene and produce an interesting interplay amongst firms, creating multiplier effects through: • imitative effects, which determine diffusion of knowledge within the district; 17 Montebelluna represents a quite extraordinary case because, on one side, it has attracted large multinational firms looking for design and engineering competences for new products, and, on the other side, Italian firms have been moving material production abroad.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrial Districts in Europe 183 • complementary effects, which favour the clustering amongst firms and the deepening of production filières. The growing network of productive linkages and the deepening of technical and organ izational capability are based on: • the accumulation of knowledge and the progressive capability to solve technical problems • the reproduction of specific resources (especially human resources) not available elsewhere • collective efficiency (Garofoli 1983a; Schmitz 1997) • dynamic competitive advantages • the introduction of informal regulations and norms through the constitution and actions of intermediate institutions. All this explains and underlines the possible paths of development in IDs and the role of territorial development policies in fostering the reinforcement of local firms (and of the overall territorial system) facing the challenges of the changing conditions of global competition. The introduction of collective actions (a distinguishing feature of collective efficiency) is the main way of reproducing external economies on a voluntary basis, and producing new specific resources (i.e. new knowledge and competences) that can create real new public goods to reinforce local dynamic competitive advantages. Investment in external economies and in knowledge and high professional competences seems the main instrument for continuing a process of development in which competition should be based on innovation, quality, a large variety of products, and the introduction of new ones. The challenge is to restructure without hierarchy: ‘crucial to the restructuring process is the capacity of a district as a whole to permanently promote specialization as technologies develop so that operations with a high minimum efficient scale of production can continue to be centralized and distinctive competences can continue to be developed in each phase of production’ (Best 1991: 207). The role of local actors in facing these challenges could be crucial; when this happens collective entrepreneurship intervenes, creating new opportunities for the ID as a whole.
10.6.2 Collective Actions for the Reproduction of External Economies: Experiences in European Countries Intermediate institutions can assume the following roles: a) finding solutions to common problems amongst local firms within the district, and b) anticipating the needs of new skills and competences and even the needs of the market. The first role is linked with the capability to cover the traditional weak points of SMEs, especially access to strategic factors for development. The establishment of specific
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
184 Gioacchino Garofoli intermediate institutions represents the collective reaction to market failures, i.e. the limits to access to specific factors or items and the lack of capability to find solutions through market mechanisms. Many SME consortia facilitate access to credit and financial resources, services centres supply ‘real’ services to small firms, and technology centres foster the transfer of new technologies to small firms and help them in finding new technical solutions. The second role is very difficult but it is a crucial element in adapting the organization of the district towards the ‘high road to development’, especially in this phase of increasing globalization. Intermediate institutions represent the reaction of the local community to the weaknesses of SMEs, covering the structural ‘hiatus’ of the small-firm organization with respect to the large one. The opportunity to cover a high level of implicit demand for services makes for a threshold organization with the specific capabilities and professional tasks necessary to solve problems common to small firms. The very large number of cases in services c entres, in training and employment institutions, and in SMEs’ credit support consortia tells its own tale. Even external scholars can recognize it, underlining the policy lessons for developing countries (cf. Schmitz and Musyck 1994). Establishing intermediate institutions is not a permanent panacea for the weaknesses of a production system based on small firms; because the ID model is a dynamic one, continuous transformation is needed even in the organization of intermediate institutions which need to move on the ‘frontier’ of the changing implicit needs of local small firms, anticipating risks and new competitive factors. The services and technology centres must introduce new competences and skills not yet available in the market, solving new problems continuously. It is interesting to note the great ability to introduce collective actions in Italy during the 1970s and 1980s and, unfortunately, the lack of this capability after the introduction of specific ID laws and with the decentralization of state competences to the regions in the 1990s. This curious experience carries some important warnings: it is always necessary to negotiate with the national and regional state, without waiting for favours. A coherent policy approach towards IDs and, in general, towards the existing agglomerations of firms, can be identified in France and Spain in recent years, however. In France, industrial and innovation policies based on competitive poles (pôles de competitivité) and on LPSs (systèmes productifs localisés) seem clearly to have understood the main policy lessons of ID research analysis in recent decades. The state’s attention to interaction amongst local actors (particularly between research institutions and industrial firms), its role in enhancing investment in innovation, and the opportunities to organize negotiation between the nation state and local communities, are crucial to the upgrading of firms and the introduction of new skills amongst local firms and organizations (see Benko and Pecqueur 2009). A quite similar approach has been introduced in Spain, where industrial policy assumed a real territorial nature: the state, through a tender mechanism, fostered the organization of cooperative projects (consisting of firms, training centres, and research institutions) with the aim of enhancing collaborative links and networks to create knowledge, new competences, and competitive advantages (Trullen 2007, 2009). The areas eligible for this public tender were all based on the agglomeration of firms and on a local productive system (involving industrial districts, clusters, and development poles à la Perroux). The objectives of the state’s intervention have been, then, innovation and transformation as general conditions favouring the increase of national industrial competitiveness.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrial Districts in Europe 185 The governance of economic transformation in industrial districts has become more complex, involving other actors. In France, for instance, the CDIF18 was supporting industrial districts and LPSs looking for development and restructuring strategies, mainly fostering the introduction of innovation. The club was even able to negotiate with public administration at all levels, from the European level through national to regional and local state. Even innovation policies in northern European regions (see Garofoli and Musyck 2003) show how these regions learnt from the experience of industrial districts. Regional innovation policies in these countries fostered not only interactions amongst different organizations (especially inter-firm but also between research institutions and firms) but even the creation of spillover effects through the accumulation of new knowledge, collective learning, and agglomerations of interactive and innovative firms. The analysis of the productive organization of an industrial district and of its social factors allows us to shed light on important new variables in the decision process (on localization, investment, strategies) of economic actors. They condition the transformation processes of the local economy (and society) and consequently of the regional and national economy. These crucial variables include cooperative links amongst firms, relations between the production system and the socio-institutional system, the high level of workers’ competences and their involvement in the firm’s productive organization and in the wider social model of the area, and the role of specific local institutions in overcoming market failures. In other words, a social system of inter-relations, of circulation of information, of production and reproduction of values, organizes itself to permeate and characterize the mode of production. This means that many crucial factors are historically embedded in the local society and are not easily transferable to other areas: the process of development acquires its definitive character as a ‘social process’ by refusing to appear only as a technical one.
10.7 Lessons for Emerging Countries and New Industrial Regions: Some Final Conclusions Is an endogenous process of development still possible, especially in new industrial countries or in emerging industrial ones? The question is linked with territorial development (and the process of industrialization ‘on the ground’ (Scott and Garofoli 2007)), with a solid basis in society and territory, and with the transformation of territorial agglomerations of firms in emerging and developing countries (often clusters of firms). There are two kinds of problems: the first is linked with GVCs and the second with the capability of agglomerations of firms in emerging industrial countries (often clusters of firms) to move towards a real endogenous development path. 18 The French Club for Industrial Districts has recently changed its name to France Cluster.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
186 Gioacchino Garofoli As regards the first question, it is important to understand which are the firms that are able to control the GVC. It is difficult for GVCs to be controlled by firms from the new emerging industrial countries; even a country like Italy would be unable to do this. The potential for endogenous development in that situation in emerging countries is therefore quite arduous. What has been happening with GVCs is that MNEs not only control the production chain, but even the production and employment prospects of these countries, providing no opportunity of autonomy for these areas and countries, unless counterbalanced by strong initiatives involving both private and public actors in the agglomerations of firms. Moreover, it is clear from the case of Romans that controlling international production of the GVC through a functional upgrading process in engineering and market control is complex and difficult. It is necessary to allow a general upgrading process for territorial and national firms in all the phases, starting with linkages increase in production phases, favouring the attention to proper and consistent firms’ strategies which could even involve the conception and engineering of new products, and relations with the market (with an increasing trading control). Concerning the second question, Table 10.2 shows how clusters in developing countries can be moved towards an endogenous development process. In conclusion, even the clusters in developing countries need to move towards a development path very similar to that required for IDs in Europe. In this sense, the relationship between the literature and the analysis of the strategies and policies introduced in IDs and clusters of firms in developing countries needs to be reinforced. This also underlines the relevance of the literature on typologies of LPSs (for specific comments on these issues, see Garofoli 2009: 47–8). The question of orienting industrial production (within IDs and agglomerations of firms) in the direction of the progressive use of higher skills and competences is crucial; this approach could defend the quality of production and the quality of the jobs (introducing more and more workers with a wide spectrum of competences, even for the conception of new products and management positions), even guaranteeing the continuous upgrading of the firms’ strategies.
Table 10.2 Development trajectories for clusters in developing countries Self-centred cluster
Extraverted cluster
Integrated productive filière Specific skills and competences Organizational/entrepreneurial capabilities Ability to introduce collective actions Local institutions for governance Territory with an active role Some capability in trading with final markets High road to development
- Low territorial integration - Lack of technical and management competences - Competitiveness based on low wages - Low social capability - Lack of intermediate institutions - Traditional localization factors - Weak position in the international division of labour - Low road to development
Source: Garofoli (2011: table 2, p. 176).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrial Districts in Europe 187
References Becattini, Giacomo (1979) ‘Dal settore industriale al distretto industriale. Alcune considerazioni sull’unità di indagine dell’economia industriale’, Rivista di Economia e Politica Industriale V(1). Becattini Giacomo (ed.) (1987) Mercato e forze sociali: il distretto industriale. Bologna: il Mulino. Becattini Giacomo (1989) ‘Riflessioni sul distretto industriale marshalliano come concetto socioeconomico’, Stato e Mercato 25(1): 111–28. Becattini Giacomo (2000) Il distretto industriale. Torino: Rosenberg & Sellier. Becattini Giacomo and Enzo Rullani (1993) ‘Sistema locale e mercato globale’, Economia e Politica Industriale 80: 25–48. Becattini, Giacomo, Marco Bellandi, and Lisa De Propris (eds) (2009) A Handbook of Industrial Districts. Cheltenham: Edward Elgar. Benko, Georges and Bernard Pecqueur (2009) ‘Industrial Districts and the Governance of Local Economies: The French Case’, in Giacomo Becattini, Marco Bellandi, and Lisa De Propris (eds) A Handbook of Industrial Districts. Cheltenham: Edward Elgar, pp. 501–11. Bernabé Maestre, José M. (1983) ‘Industrialización difusa en la provincia de Alicante’, mimeo, Facultad de Geografía de la Universidad de Valencia. Best, Michael H. (1991) The New Competition. Cambridge: Polity Press. Brusco, Sebastiano (1975) ‘Organizzazione del lavoro e decentramento produttivo nel settore metalmeccanico’, in FLM (ed.) Sindacato e piccola impresa. Bari: De Donato. Brusco, Sebastiano (1982) ‘The Emilian Model: Productive Decentralisation and Social Integration’, Cambridge Journal of Economics 2: 167–84. Brusco, Sebastiano (1989) Piccole imprese e distretti industriali. Una raccolta di saggi. Torino: Rosenberg & Sellier. Brusco, Sebastiano and Charles Sabel (1981) ‘Artisan Production and Economic Growth’, in Frank Wilkinson (ed.) The Dynamics of Labour Market Segmentation. London: London Academic Press, pp. 99–113. Coltorti, Fulvio and Gioacchino Garofoli (2011) ‘Medium-Sized Enterprises in Europe’, Review of Economic Conditions in Italy 1: 171–202. Costa Campi, Maria Teresa et al. (1993) Cooperación entre empresas y systemas productivos locales. Madrid: IMPI. Courlet, Claude (1997) Relations interentreprises et territoire: le cas de la Vallée de l’Arve. Grenoble: IREP-D, Université Pierre Mendès France. Courlet, Claude (2008) L’économie territoriale. Grenoble: Presses Universitaires de Grenoble. Courlet, Claude and Abderraouf Hsaini (1997) Les systèmes productifs localisés et leur identification en France. Grenoble: IREPD-DATAR. Courlet, Claude and Pierre Judet (1986) ‘Nouveaux espaces de production en France et en Italie’, Les Annales de recherche urbaine 29: 95–103. Courlet, Claude and Bernard Pecqueur (1992) ‘Les sistèmes industriels localisés en France: un nouvel model de développement’, in Georges Benko and Alain Lipietz (eds) Les régions qui gagnent. Districts et réseaux: les nouveaux paradigmes de la géographie économique. Paris: Presses Universitaires de France, pp. 81–102. Courlet, Claude, Gioacchino Garofoli, Abderraouf Hsaini, and Patrick Faillenet (2000) Premières reflections pour l’élaboration d’un Livre Blanc des districts industriels et des SPL européens. BeziersCluses: Almatec-Sidemva. De Martini E. and Gioacchino Garofoli (1993) ‘Distretti industriali e sistemi produttivi locali in Francia’. Dipartimento di Scienze Economiche e Statistiche, Università di Pavia. Figueiredo, Antonio, José S. Costa, and Mario Rui Silva (1994) ‘Estratégia para a competitividade da Indústria do Norte’, A Indústria do Norte, 1. Fuà, Giorgio (1980) Problemi dello sviluppo tardivo in Europa. Bologna: Il Mulino.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
188 Gioacchino Garofoli Fuà, Giorgio (1983) ‘L’industrializzazione nel Nord Est e nel Centro’, in Giorgio Fuà and Carlo Zacchia (eds) Industrializzazione senza fratture. Bologna: Il Mulino, pp. 7–46. Ganne, Bernard (1990) Industrialisation diffuse et systèmes industriels localizés: essai de bibliographie critique du cas français. Geneva: Institut international d’études sociales, BIT. Ganne, Bernard (ed.) (1992) Développement local et ensemble de PME. Lyon: GLYSI. Garofoli, Gioacchino (1981) ‘Lo sviluppo delle aree periferiche nell’economia italiana degli anni settanta’, L’Industria II (3): 391–404. Garofoli, Gioacchino (1983a) Industrializzazione dìffusa in Lombardia. Milano: IReR—Franco Angeli. Garofoli, Gioacchino (1983b) ‘Aree di specializzazione produttiva e piccole imprese in Europa’, Economia Marche II(1): 3–46. Garofoli, Gioacchino (1989) ‘Industrial Districts: Structure and Transformation’, Economic Notes 19(1): 37–54. Garofoli Gioacchino (1991) Modelli locali di sviluppo. Milano: Franco Angeli. Garofoli, Gioacchino (ed.) (1992) Endogenous Development and Southern Europe. Aldershot: Avebury. Garofoli, Gioacchino (1993) ‘Economic Development, Organisation of Production and Territory’, Revue d’Economie Industrielle 64: 22–37. Garofoli, Gioacchino (1994) ‘The Industrial District of Lecco: Innovation and Transformation Processes’, Entrepreneurship and Regional Development 6(4): 371–93. Garofoli Gioacchino (1996) ‘I distretti industriali’, in V.A., Libro della piccola impresa. Rome: Adn kronos Libri, 119–91. Garofoli, Gioacchino (2003) ‘Distretti industriali e processo di globalizzazione: trasformazioni e nuove traiettorie’, in Gioacchino Garofoli (ed.) Impresa e territorio. Bologna: Il Mulino, pp. 539–71. Garofoli, Gioacchino (2009) ‘Regional and Local Development’, SR Scienze Regionali—Italian Journal of Regional Science 8(3): 35–58. Garofoli Gioacchino (2011) ‘Globalizzazione e sistemi produttivi locali’, in Bruno Bracalente and Marco Moroni (eds) Italia media. Milano: Franco Angeli, pp. 172–93. Garofoli Gioacchino (2018) ‘Euro fort et excédents commerciaux: le paradoxe italien’, The Conversation, 24 June. Garofoli, Gioacchino and Bernard Musyck (2003) ‘Innovation Policies for SMEs: An Overview of Policy Instruments’, in Bjorn T. Asheim, Arne Isaksen, Claire Nauwelaers, and Franz Tödtling (eds) Regional Innovation Policy for Small-Medium Enterprises. Cheltenham: Edward Elgar, pp. 119–38. Gereffi, Gary, John Humphrey, and Timothy Sturgeon (2005) ‘The Governance of Global Value Chains’, Review of International Political Economy 12: 78–104. Johannisson, Bengt (2009) ‘Industrial Districts in Scandinavia’, in Giacomo Becattini, Marco Bellandi, and Lisa De Propris (eds) A Handbook of Industrial Districts. Cheltenham: Edward Elgar, pp. 521–34. Modiano, Pietro (1982) ‘Competitività e collocazione internazionale dell’industria italiana: il problema dei prodotti tradizionali’, Economia e Politica Industriale 33. Piore, Michael J. and Charles Sabel (1984) The Second Industrial Divide. New York: Basic Books. Reis, José Joaquim Dinis (1992) Os espaços da indústria: A regulação económica e o desenvolvimento local em Portugal. Porto: Ediçoes Afrontamento. Sabel, Charles and Jonathan Zeitlin (1982) ‘Alternative storiche all’industrializzazione di massa’, Stato e Mercato, 5(August): 213–58. Sabel, Charles and Jonathan Zeitlin (1985) ‘Historical Alternatives to Mass Production: Politics, Markets and Technology in Nineteenth-century Industrialization’, Past and Present 108: 133–76. Schmitz, Hubert (1997) ‘Collective Efficiency and Increasing Returns’. IDS Working Paper No. 50. Schmitz, Hubert (ed.) (2004) Local Enterprises in the Global Economy: Issues of Governance and Upgrading. Cheltenham: Edward Elgar. Schmitz, Hubert and Bernard Musyck (1994) ‘Industrial Districts in Europe: Policy Lessons for Developing Countries’, World Development 22(6): 889–910. Scott, Allen J. (1988) New Industrial Spaces. London: Pion.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Industrial Districts in Europe 189 Scott, Allen J. and Gioacchino Garofoli (2007) ‘The Regional Question in Economic Development’, in Allen J. Scott and Gioacchino Garofoli (eds) Development on the Ground: Clusters, Networks and Regions in Emerging Economies. London and New York: Routledge, pp. 3–22. Signorini, Federico (ed.) (2000) Lo sviluppo locale. Un’indagine della Banca d’Italia sui distretti industriali. Rome: Meridiana Libri e Donzelli. Silva, Mario Rui (1988) ‘Industrialisation et Développement Local: une interprétation à partir du cas portugaise’, doctoral thesis, Université des Sciences Sociales de Grenoble, Grenoble. Silva, Mario Rui (1992) ‘Development and Local Productive Spaces: A Study on the Ave Valley (Portugal)’, in Gioacchino Garofoli (ed.) Endogenous Development and Southern Europe. Aldershot: Avebury, pp. 117–30. Solinas, Giovanni (1996) I processi di formazione, la crescita e la sopravvivenza delle piccole imprese. Milano: Franco Angeli. Trullen, Joan (2007) ‘La nueva política industrial española: innovación, economías externas y productividad’, Economía Industrial 363: 17–31. Trullen, Joan (2009) ‘National Industrial Policies and the Development of Industrial Districts: Reflections on the Spanish Case’, in Giacomo Becattini, Marco Bellandi, and Lisa De Propris (eds) A Handbook of Industrial Districts. Cheltenham: Edward Elgar, pp. 726–38. Vazquez Barquero, Antonio (2003) Endogenous Development. London: Routledge. Ybarra, Josep-Antoni (1991) Industrial Districts and the Valencia Community. Geneva: International Institute for Labour Studies, ILO. Zeitlin, Jonathan (1985) ‘Distretti industriali e struttura industriale in prospettiva storica’, in Raimondo Innocenti (ed.) Piccola città & piccola impresa. Milano: Franco Angeli, pp. 33–51. Zeitlin, Jonathan (1995) ‘Why Are There No Industrial Districts in the United Kingdom?’, in Arnaldo Bagnasco and Charles Sabel (eds) Small and Medium-size Enterprises. London: Pinter, pp. 98–114. Zeitlin, Jonathan (2007) ‘Industrial Districts and Regional Clusters?’, in Geoffrey Jones and Jonathan Zeitlin (eds) The Oxford Handbook of Business History. Oxford: Oxford University Press, pp. 219–43.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
chapter 11
The Politica l Econom y of Speci a l Economic Zon e s Pasts, Presents, Futures Patrick Neveling
11.1 Where is the Juncture? Industrialization-driven Development, SEZs, and Industrial Hubs In response to the global economic crisis that followed the real-estate bubbles bursting across Western advanced capitalist national economies in 2008, world-leading development economists have called for a revitalization of the nation-state’s role in economic development. Their call draws our attention to the realm of industry and manufacturing, in particular, and highlights the potentials of these two long-forgotten, yet vital sectors for economic growth (Stiglitz et al. 2013; Lin 2014). This new development agenda for the global economy is, in many ways, reminiscent of the long-standing agenda in many Third World nations where export-oriented industrialization has been a focal point of development policies since the 1960s or earlier (for Africa, see Neveling 2019). Yet, despite such continuities on the national scale, the renewed interest in industrialization also reminds us that inter national organizations, from the World Bank (WB) and the International Monetary Fund (IMF) to the Organisation for Economic Cooperation and Development (OECD), have shifted their attention from industrialization to other economic sectors from the 1970s onwards. The WB, for example, gradually changed course after Robert McNamara was appointed president in 1968. McNamara was an established, proto-typical Cold Warrior, who began his career as the protégé of world-leading anti-Semite Henry Ford and, after a short spell as CEO at Ford Motor Corporation, inaugurated the war of attrition in Vietnam. He led the Bank on a course that reduced funding for large-scale infrastructure projects although
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
The Political Economy of Special Economic Zones 191 these were crucial for industrialization. Instead, he championed policies that promoted so-called productive employment and veiled the persistent geographical inequalities in the global distribution of wealth during the final decades of the Cold War. That approach held the Bank in its grip well into the 2000s when the forceful opposition to Joseph Stiglitz’s short tenure as chief economist of the World Bank exemplified once more that development by sustainable industrialization was opposed to the free-market dogmas propagated under the Washington Consensus and after. Structural adjustment lending (SAL) by the Bank and the IMF since the 1980s has promoted unsustainable, short-lived industrialization at best, whereas its paramount foci have been to interfere with national sovereignty under the label ‘good governance’, and on the divestiture of national industries for privatization (Sharma 2015; Reinsberg et al. 2019; Neveling 2017c). This chapter details the historical and present-day political economy of special economic zones (SEZs). An analysis of past and present motivations for establishing particular SEZs and of similar export-oriented industrialization policies in export processing zones (EPZs), free trade zones (FTZs), and other industrial enclaves with flagship investment incentives, can advance our understanding of the pitfalls and potentials of national industrialization policies, especially in light of the changing global systemic predicaments that individual nations face when their industrial development programmes might appear detrimental to the interests of established Western imperial nations that control international development organizations with considerable leverage over national development policies. SEZs have been central to many national industrial development programmes and, at the same time, have often been imposed as development policy instruments on indebted nations as conditionalities for the WB and IMF SAL. On the surface, this makes SEZs look like counterexamples for the above diagnosis of a post-1968 policy shift in the WB and beyond. Yet, today’s well over 4,000 SEZs in more than 130 nations and the employment they provide for 100 million or more (mainly women) workers are the opposite of sustainable industrial development policies. Instead, the vast majority of SEZs, past and present, have either failed to create substantial employment or their positive effects, sometimes captured by glossy ‘miracle’ phrases, have been short-lived and to the benefit not of local populations and national economies but to the benefit of multinational corporations and their shareholders, first and foremost. This is, first, because of the zones’ crucial role in the emergence of a neoliberal division of labour within the capitalist world-system. This system that shifted manufacturing in textiles and garments, consumer electronics, toys, and later also biopharmaceutical and petrochemical industries from Western capitalist nations to so-called newly industrialized nations. For many decades, policy advisers and researchers in international organizations have praised SEZs because of their potential to stimulate manufacturing relocations and generate large-scale employment. The zones’ significant impact on declining global labour standards in manufacturing industries has been ignored, instead. Similarly, the never-ending stream of academic publications, white papers, feasibility studies, reports, newspaper articles, and investor guidelines portrays SEZs as ‘miracle makers’ or ‘engines of growth’ and ignores the short-lived nature of SEZ-booms and the damaging effects of deindustrialization at the end of such booms (Zenger 1977; Warr 1995; World Bank Country Study 1990 [1989]; Madani 1999; Zeng 2010; Dicken 2003). Second, SEZs are emblematic of the pitfalls of export-oriented industrialization policies. While most pamphlets and publications from the World Bank and other international organizations are quiet about the negative effects of the zones, human rights organizations,
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
192 Patrick Neveling international trade unions, critical political economists, and other scholars, as well as investigative journalists, are united in concluding that the proliferation of SEZs has played a crucial part in the escalation of global poverty and inequality. The relocation of industries from Western nations to such zones caused large-scale unemployment in the regions of departure, with deindustrialization followed by decades of economic decline until this day. Labour standards in ‘new’ industries in the receiving SEZ regions, however, are always lower. First and foremost, state-run and parastatal agencies advertise SEZs internationally for their cheap, abundant, and supposedly compliant workforces. This has encouraged severe labour rights violations. Women workers are especially targeted by many SEZ industries, and critical scholars have documented decades of sexist and racist abuse by shop floor managers and companies more generally that nurture orientalist discourses about non-European women workers’ ‘nimble fingers’ that allegedly make them especially suited for garment and consumer electronics production. In a contrary move, SEZ f actories often proclaim that their workers are less productive than elsewhere in the world. Yet, factories actually promote the deskilling of women workers by a preference for on-thejob-training and by tapping into stereotypes of local patriarchal and religious groups so that young women workers are stigmatized if they speak up against exploitation and labour rights violations (Neveling 2015c; Ong 1987; Fernández-Kelly 1983; Shakya 2018; Campbell 2018). Third, SEZ promotion literature from the World Bank and other international organizations commonly ignores the fact that national governments have to cover health care, social welfare, and pensions because SEZ workers earn so little that they can hardly ever accumulate savings. Added to this is the short lifespan of SEZ employment and the fact that investment incentives mean low to zero taxation on capital gains, no customs duties on imports, and instead significant state expenditure on infrastructure and other fixed capital amenities. On the one hand, this leaves national budgets with little income from ongoing SEZ operations and exposes the state and its agencies to the volatility of international trade with the need to recover upfront expenses in fixed assets like industrial land and infrastructure. On the other hand, the fact that multinationals are liberated from initial investment means that SEZs encourage ‘runaway’ shops that relocate whenever there is a rise in the collective bargaining power of workers, when preferential export quota ceilings are met, or when wages increase (Neveling 2015a, 2015b, 2017a, 2017b; Fernández-Kelly 1983; Fröbel et al. 1981; Mezzadri 2017; Campbell 2018; Ngai 2005; Orenstein 2011; Schrank 2003). Thus, the contemporary condition of SEZs as much as that condition’s historical forging in a capitalist world-system riven with the injustices of Western-led super-exploitation during the Cold War and after has run counter to many postcolonial nations’ efforts at economic independence during the era of decolonization and non-alignment as much as it runs counter to contemporary efforts at industrial development that is sustainable in economic, ecological, and social terms. This is the juncture where industrial hubs enter the picture and where past and present efforts to establish such hubs gain paramount importance in the mundane terms of economic development policies and in the lofty terms of academic development economics analysis. Could such hubs, if successful, capture the commonplace runaway capital that roams the world’s SEZs and keeps it tied in a given geographical space? To answer this question, section 11.2 of this chapter offers an extended critical political- economy analysis of the historical predecessors of contemporary SEZs. This is developed in
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
The Political Economy of Special Economic Zones 193 critical conversation with neoliberal development historiographies that misrepresent SEZs as extensions of so-called free ports and thus as devices that should facilitate imperial control over free trade and increased capital mobility at reduced tax and customs rates. Such n eoliberal historiography replaces the social, political, and economic ambitions of post-1945 national development programmes—to create lasting growth, to secure mass employment, and to offer deprived and unemployed populations longer-term perspectives for a better life—with empirically flawed and analytically absurd praises of ‘freedom’ and ‘growth’. Section 11.3 details the emergence of the contemporary SEZ model in the post-World War II US dependency Puerto Rico and discusses the proliferation of SEZs from the 1940s to the present with a focus on the political and economic pressure groups behind this spread. The final section suggests a working programme for industrial hubs that could reverse the negative global effects of SEZs to limit capital mobility and implement standards for sustainable investment.
11.2 Free Trade or Industrial Districts? Predecessors of Special Economic Zones The global spread of special economic zones in recent decades has created a widespread interest in the historical predecessors and origins of this remarkable, world-making model for economic development. Definitions are fairly far-reaching and depend on a given author’s political orientations and whether they are employed in research departments of international organizations (Farole 2011; Farole and Akinci 2011; Akinci et al. 2008; Currie and Economist Intelligence Unit 1985; United Nations Centre on Transnational Corporations and International Labour Organization 1988; Romero 1995) or whether they have the opportunity to conduct independent research at universities and other publicly funded, comparatively independent and non-biased institutions (Neveling 2017a; Fernández-Kelly 1983; Fröbel et al. 1981; Mezzadri 2017). Yet, in sum, all definitions identify SEZs as demarcated in one way or another—by physical or legal boundaries or by a combination of both. Definitions further agree that SEZs attract investors because of a distinct set of incentives that offer reductions and waivers in taxation—first and foremost capital gains and income taxes, customs duties for the import of machinery, raw materials, and part-processed goods, and for the export of semi-finished or finished goods. SEZ regulations have expanded from industrial manufacturing to banking and finance, real estate, agriculture, and tourism in recent decades. Critical scholars and trade unions also include laws that prohibit collective bargaining and related trade union activities as well as very low wages and employment standards in the defining features of SEZs. Neoliberal development economists further fail to recognize another defining feature of the zones, which is the significant state investment in fixed capital, principally in the cultivation of land for industry, which may involve anything from flattening slopes to providing sewerage and roads, telecommunications connections, fences, factory buildings, and capital for all the feasibility studies and reports required to create industrial estates and zones (Neveling 2017a, 2017c). Further differences in definitions of the SEZ phenomenon derive from misunderstandings of the zones’ origins and their impact on global capitalism. Recent publications from
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
194 Patrick Neveling the research department of the International Finance Corporation (IFC), which is the public–private partnership subsidiary of the World Bank Group, founded in 1955, have emphasized and expanded on earlier narratives, dating back to the 1950s, that propagate free ports in Mediterranean antiquity as the birthplaces of today’s SEZs (Baissac 2011; Thoman 1956). The problem with such comparisons is that they fail to understand the political and economic purpose of such ports, which was the opposite of free trade. As a predatory imperial formation, the Roman Empire used the (often mistakenly praised) ‘free’ port of Delos to destroy the economy of Rhodes’ harbour because that island refused to succumb to Roman imperial rule and economic orders. Similarly, several IFC-authored publications further misrepresent ports of the Hanseatic League and in Italian medieval city-states as genuinely ‘free’ when all those ports were incorporated into clientelist and cartel-like trading alliances that sought to control regional markets and oust competitors (Neveling 2015d). Misunderstandings within the realm of neoliberal development economics extend to representations of European colonial free ports such as Singapore and Hong Kong as forerunners of SEZs. Again, such ports and their respective trading regimes served the purpose of enforcing monopolies, such as the Dutch monopoly in the nutmeg trade from the Moluccas for as long as it lasted (Chaudhuri 1985; Frank 1998). Further, the British Empire’s ports in Hong Kong and Singapore captured and channelled existing regional trade, attracting trading diasporas of the Eastern Indian Ocean and of the South China Sea, which were subjected to so-called tax farming practices via opium dens and harbour levies throughout the second half of the nineteenth and the first half of the twentieth century. If manufactur ing industrialization in both ports emerged via migrant entrepreneurial networks that may have been attracted by Singapore’s and Hong Kong’s position in global trading networks, colonial free-port administrations devoted zero attention to supporting such native industrialization efforts. Instead, as in many other European colonies, British administrations were eager to prevent thriving local manufacturing industries so as to protect export markets for European produce (Wong 1988; Ken 1978). The above highlights the shortcomings of mainstream deliberations on SEZ predecessors. The fact that most recent such publications come from a network of pro-SEZ activists associated with the World Export Processing Zones Association (WEPZA), whose advisory board members also hold crucial positions in various World Bank organizations and in private-sector SEZ development consulting, points to the political agenda that informs such historiography.1 Researchers and consultants for international organizations are here cladding a salient and escalating industrial development practice with a big-history narrative that embeds SEZs in a lineage of Western-centric, mythical imperial and proto-imperial success stories (Farole and Akinci 2011; Farole 2011; Zeng 2010; Madani 1999; Bräutigam and Xiaoyang 2011). Yet, such stories have little relation to the economic development pol icies and ambitions of postcolonial developing nations in the twentieth and twenty-first centuries. What is more, the historical cases offered as predecessors of contemporary SEZs all lack a crucial feature of those contemporary zones, which is large-scale employment in manufacturing and industrial enterprises that are firmly embedded in global commodity chains and global value chains. If this is the central feature of SEZs, European free ports of the late nineteenth and early twentieth century in Hamburg, Copenhagen, and elsewhere resemble them much more 1 See http://www.wepza.org/officers for advisory board member affiliations (last visited 3 June 2019).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
The Political Economy of Special Economic Zones 195 closely than do eighteenth- and nineteenth-century free ports in other world regions—where Kingston in the Caribbean or Mauritian Port Louis in the Western Indian Ocean served as havens of illicit trade where privateering letters could be obtained for capturing vessels of enemy merchant navies and selling their merchandise. However, in those nineteenth- century European free ports where manufacturing enterprises part-processed or finished part-processed goods, this was for imports from European colonies and not for re-export. Instead, if European free ports—and from the 1930s onwards also US free ports—were sites of tax- and customs-free manufacturing, this involved the final steps in production processes before those goods were sold on national markets (for empirical evidence on this analysis see Anderson 1934; Harreld 2015; Marzagalli 1996; Fenner 2013). Thus, whereas contemporary SEZs are firmly embedded in export-oriented industrialization policies, earlier free-port manufacturing industries can be regarded as subtypes of import-substitution policies that were established in competition with export-substitution industries such as sugar-cane mills and other processing plants for colonial agricultural and other raw material exports. Tellingly, such free ports that qualify as predecessors of contemporary SEZs were located in Western late-industrializing nations, where governments had a vested interest in protecting native industries against foreign competition. In many ways, part-processing importsubstitution industries for coffee beans, cane sugar, groundnuts, and other colonial produce in free-port territories in Hamburg, Copenhagen, Gothenburg, and a number of smaller northern European port cities therefore worked along the lines of industrial districts, a planning and policy device for accelerated industrialization that gained traction across Europe and the United States in the late nineteenth century and that was based on studies by the neoclassical economist Alfred Marshall, who had assessed in detail the rise of English small firm-driven industrialization in Yorkshire and Lancashire during the eighteenth century (Marshall and Whitaker 1975a, 1975b). Marshall’s conclusions about the factors that helped industrialization take off in those regions were summarized during the 1990s globalization wave, when the concept of industrial districts was revived in regional planning: Agglomeration is said to offer a series of advantages upon which a system of vertically disintegrated production can draw. Echoing the factors first identified by Alfred Marshall in his work on small-firm districts in Lancashire and Yorkshire during the nineteenth century, these advantages are said to include the build-up of a local pool of expertise and know-how and a culture of labour flexibility and cooperation resulting from dense social interaction and trust; lowered transport and transaction costs; and the growth of a local infrastructure of specialized services, distribution networks and supply structures. Via the consolidation of particular product specialisms in different regions a federation of self-contained regional economies is anticipated, each with its own cumulative causation effects drawing upon strong external economies of agglomeration. (Amin and Thrift 1992: 572)
In fact, efforts to replicate the rise of England to the world’s first powerhouse of the Industrial Revolution were widespread across late colonial and postcolonial nations in the economic development drive during the first decades after 1945. India, Singapore, and many other nations set up steel industries, for example, and sought to group related industries around these centres of heavy industry (Strümpell 2014). Not all late-colonial territories and postcolonial regions had the capacity for heavy industries of the size of steel plants, of course, and it is in these regions where the concept of contemporary SEZs was born.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
196 Patrick Neveling
11.3 Legacies of Imperialism? The Global Spread of Special Economic Zones After World War II, the Caribbean island of Puerto Rico radically changed its positioning in the global economy. Whereas many other regions fought for and gained political independence from Western colonial rulers, the United States never allowed Puerto Ricans to enjoy the right to national self-determination, even though this had been proclaimed as a cornerstone of a new world order by US president Woodrow Wilson in 1919. Instead, the United States held onto the island they had acquired after a war with the crumbling Spanish empire in 1898 and quashed protests from the Puerto Rican National Party and progressive, often socialist, pro-independence movements. Puerto Ricans had certain US citizenship rights, just enough to be conscripted as soldiers for World War I, but during the global recessions of the 1920s and 1930s, the island turned into a ‘stricken land’, as Rexford Tugwell, the last appointed US governor during the 1940s, described it (Tugwell 1947). Although Tugwell conjured a set of variables composed of rapid population growth, high unemployment, political upheaval, and economic backwardness, later analysts agree that Puerto Rico was stricken because of the political economic structure that US colonial capit alism had established on the island. Certainly, Spanish rule had created huge problems, in part because there was actually very little rule during the nineteenth century, and the rules that the Spanish made helped establish a hacienda system that saw few families accumulate great wealth and the lion’s share of the population struggle for everyday survival. From 1898 onward, the United States incorporated Puerto Rico into the mainland economy with the dollar as currency and without tariff barriers. This encouraged the expansion of US agricultural trusts, principally in the Puerto Rican sugar industry, and at the same time this destroyed the economic prospects of smallholders in sugar, tobacco, and other agricultural industries. After World War I, hundreds of thousands of Puerto Ricans migrated to US cities where they entered the lowest strata of the population and worked in sweatshops in the garment and textile industries. At the same time, such industries began outsourcing embroidery and other works to Puerto Rico where middlemen allocated piecework to small workshops and to households. During the 1920s and 1930s, the needlework industry outpaced the sugar sector in terms of growth rates and employment numbers. Yet when the Great Depression hit Puerto Rico, neither sector was sufficiently crisis-proof to save the island from further impoverishment and starvation (Ríos 1995; Dietz 1986; García-Colón 2009; Boris 1996). The Roosevelt administration injected significant sums of money into Puerto Rican infrastructure after 1934, in response to massive strikes and riots across all economic sectors on the island. In particular, the local administration benefited from rebates on imports to the mainland, with rum exports after the end of prohibition garnering millions of US dollars. In return, the island’s infrastructure was pushed to a standard that was, for two or more decades, much higher than anywhere else in the Caribbean and possibly any other developing region in the world. A road network connected all districts with refurbished harbours in San Juan in the north and Ponce in the south, and an airport was built near San Juan to international standards. From the 1940s onward, with further support from the US
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
The Political Economy of Special Economic Zones 197 government to keep Puerto Ricans happy and on the side of the anti-fascist war effort, Puerto Rico set up three crucial institutions: a planning board that, amongst other tasks, oversaw the implementation of industrial standards, a development bank that had money to guarantee external investments and supply new industries with capital, and a parastatal development corporation that managed new import-substitution industries such as a cement plant, a textile mill, and factories for pulp and paper, glass bottles, and shoes. However, the import-substitution phase in Puerto Rico was to be short-lived. Five years after those factories had opened in 1942, Puerto Rico changed course and, in the words of the Wall Street Journal at the time, ‘lured’ mainland manufacturers with tax holidays and financial subsidies to relocate their production to Puerto Rico (Diefenderrer 1946; Ross 1966; Neveling 2015b). The Puerto Rican programme differed significantly from the historical policies that had applied in free ports since Roman times. The focus of the local government was manufac turing and employment rather than trade promotion. This was because Puerto Rico’s own produce was already widely traded, but brought little revenue, and because there was never an option to turn the island into a nodal point for global trade similar to Singapore or Hong Kong, for example. Instead, Puerto Rico was able to embark on an export-oriented manu facturing industrialization drive because it was already an integral part of the then largest domestic market in the world, the US economy. No trade barriers to the mainland meant unlimited, customs-free movement of goods. The US dollar as local currency meant that no investor worried about currency fluctuations. US rule meant that from the early days of the Cold War onward, there was little to no chance that a socialist or communist party could ever win local elections and nationalize industries. In fact, violent crackdowns on nationalists and anti-colonial fighters raged throughout the 1950s in Puerto Rico with some freedom fighters managing an attack on then US president Harry S. Truman and a second gun attack in Congress (Neveling 2015b, 2017b). The 1950s were a boom decade for Puerto Rico, with the industrial development corporation, PRIDCO, celebrating the opening of the 400th manufacturing enterprise in 1955. As the Puerto Rican export-oriented industrialization programme coincided with the beginning of US international development aid for pro-capitalist allies across the Third World, the Truman administration’s Point Four programme made extensive use of the island’s success story. Thousands of officials from postcolonial nations were flown in and other European colonial powers in the Caribbean were pushed to embrace similar programmes (Neveling 2015b). The Caribbean Conference, an exchange forum of the various Western colonial powers in the Caribbean islands, was increasingly controlled by the United States. Local governments in Jamaica and Trinidad initiated W. Arthur Lewis’s studies of the Puerto Rican development model that chiefly informed his famous tractate ‘Economic development with unlimited supply of labour’ (Lewis 1954; Whitham 2013). The United States also sent Puerto Rican politicians on missions across its Latin American embassies, where they served as native witnesses of the benefits and benevolence of US rule. A Bostonbased marketing corporation, Arthur D. Little Inc. (ADL), was chiefly involved in the Puerto Rico policy shift from import substitution to export orientation and in the promotion of that programme across the mainland. Accordingly, ADL managed to win several US government contracts under Point Four and later programmes and also advised early postcolonial governments in Jamaica, Mexico, Ireland, and many other nations on exportoriented development promotion in the style of Puerto Rico (Neveling 2015b).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
198 Patrick Neveling At the time, no standard term existed for ADL’s increasingly standardized policy promotion package. In Panama, Colon was to become a ‘free zone’. Jamaica and the Philippines set up Puerto Rican-style programmes in the 1950s, but it was not until the 1960s and 1970s that they were called ‘export processing zones’. This is perhaps one of the reasons why World Bank and other publications mistakenly claim that the Shannon Free Zone, set up by the Shannon Free Airport Development Corporation (SFADCo) in 1959, was the first modern zone (Baissac 2011). In fact, Shannon emerged from reconnaissance visits by its then chairman, Brendan O’Reagan, to Puerto Rico and Panama. Despite gaining international fame as the world’s first zone and as the model on which China built its SEZ programme from 1979 onwards, Shannon was in fact a small and fairly unsuccessful zone that never achieved disproportionate growth or contributed significantly to employment and foreign exchange earnings in the Republic of Ireland. If anything, the fame of Shannon is the product of an awkward shift in the promotion of the EPZ and later SEZ model. From the late 1960s, the United Nations Industrial Development Organization (UNIDO) took over from the United States and spearheaded EPZ-promotion even though UNIDO had been founded in 1966 to support the industrial development needs of the non-aligned nations. UNIDO began with surveys on existing industrial development strategies and the institution’s archival record shows that several questionnaires came back with details on Puerto Rican-style export-oriented development programmes in Malta, Cyprus, the Philippines, South Korea, Malaysia, India, and, notably, the Republic of China, i.e. Taiwan. As a result, one of UNIDO’s fourteen working groups (WGs) was tasked with promoting export-oriented industrialization. UNIDO’s funding structure invited national governments to donate for particular purposes, so that the United States and other Western nations with a vested interest in the proliferation of SEZs could make sure that this particular WG was flush with cash. William Tanaka, the Japanese head of the WG, initiated a massive EPZ promotion drive from 1968 onwards that saw UNIDO run the first feasibility study for an EPZ in Mauritius and also embark on a large-scale publication, technical assistance and fellowship programme for SEZs (Neveling 2015a, 2015b, 2017b). At the time, Taiwan’s Kaohsiung Export Processing Zone, set up as part of the city’s container harbour in 1965, was the most successful zone in the world and certainly also the best organized. UNIDO intended to run various international workshops and a fellowship programme through Kaohsiung, but plans fell through when the PRC joined the UN and Taiwan walked out in response in the early 1970s. Shannon’s O’Reagan was quick to spot the opportunity and turned the ailing Shannon zone into a hotspot for international EPZ establishment training. Shannon housed dozens of workshops and hundreds of UNIDO fellows, amongst them later PRC leader Jiang Zemin in 1981 (Neveling 2015b). UNIDO also helped standardize the term EPZ in the mid-1970s. What were initially called ‘export-free zones’ (Kelleher 1976) became known as ‘export processing zones’ in the following years—not least through the seminal study ‘The New International Division of Labour’ by German macrosociologists Folker Fröbel, Jürgen Heinrichs, and Otto Kreye (Fröbel et al. 1981). If the 1970s were the decade when the EPZ concept was consolidated, the 1980s and 1990s were the decades when it took off (see Table 11.1). This was mainly because the WB and the IMF made EPZs an integral part of the many structural adjustment programmes (SAPs) that postcolonial nations had to sign up to during the so-called Third World debt crisis. Far from being limited to the Third World, this was a global crisis that initially got a grip on the United Kingdom (bailed out by the IMF in 1976) and various US cities such as New York and
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
The Political Economy of Special Economic Zones 199
Table 11.1 The global spread of EPZs/SEZs since 1975 Year
1975
1978
1984
1986
1997
2002
2006
Number of countries with EPZs Number of EPZs Employment (millions) —of which PR China —of which other countries with figures available Share of PR China in %
25 79 0.725 – 0.725
28 N/A 0.694 – 0.694
35 N/A 0.837 0.015 0.822
47 176 1.97 0.07 1.9
93 845 22.5 18 4.5
116 3,000 43 30 13
130 3,500 66 40 26
0
0
1.79
3.55
80
69.77
60.60
Source: Neveling (2015d).
Detroit that defaulted on debts and were subject to their own variant of SAPs. Mauritius was possibly the first Third World nation to enter a period of near-default that lasted from 1979 to the late 1980s; subsequently many other nations across the world would sign SAPs and be forced to privatize national industries, establish tripartite institutions of the Puerto Rican kind set up, with national standard boards that now oversee compliance with the International Organization for Standardization’s ISO norms series, national development banks, and development organizations with private-sector majorities on executive boards (Neveling 2017c). Notably, the PRC also needed IMF support in the mid-1980s. The IMF conditionally consolidated important modifications to the SEZ programme established in 1978–9 which at the time suffered from severe capital shortages as numerous investors failed to inject the capital they had pledged and left the PRC high and dry in terms of state funding for vital infrastructure in Shekou, Shenzhen, and the wider Guangdong and Fujian region. In sum, the global spread of SEZs since the 1940s was driven by the Cold War efforts of the United States and other Western nations to propagate export-oriented capitalist development for the world market and to do so in competition with other development models that championed import substitution or incorporation into an alternative socialist world economic system. Most nations that became successful with EPZs and SEZs did so after initial efforts to substitute imports proved insufficient for national development. However, it is crucial to note that those nations that managed to develop their zones into a lasting success never entirely abandoned import substitution, as is evidenced by the many steel plants that helped industry thrive in India, Singapore, and South Korea, for example. Importantly, successful SEZs not only created new employment opportunities in Third World regions that sometimes became known as ‘newly industrialized nations’, but also created massive unemployment in the regions where capital left for SEZs. So, why did Western nations promote SEZ establishment in developing nations even though heartlands of textile and garment production in the United States, the United Kingdom, Germany, and elsewhere went into rapid decline with mass unemployment and deindustrialization? Such regions had become hotspots of organized labour with strong bargaining power in wage negotiations. Emerging multinational corporations in the textile and garment sector and later also in consumer electronics, automobile production, and so forth could use their increased and low-risk capital mobility to destroy established communities of workers and play off manufacturing
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
200 Patrick Neveling locations against one another by forcing trade unions to accept wage cuts in exchange for relocation waivers (Chomsky 2008; Koistinen 2000; Kasmir and Gill 2018; Neveling 2017b). Yet, the question that emerged in many SEZs across the world was how to avoid similar deindustrialization as ever more zones sprang up and a global race to the bottom in corpor ate taxation, wages, and labour standards developed that further escalated already abject industrial relations politics that included the widespread murder of trade unionists, child labour, sexual harassment, and super-exploitation on a global scale (Marhoz and Szymanski 1996).
11.4 From Zones to Hubs? Learning to Curb Runaway Capital in SEZs Some time in the 2000s, political scientists and mainstream development economists became interested in the concept of learning, which manifested in the SEZ field in publications with titles like ‘Learning from Shenzhen’ or ‘Special Economic Zones in Africa: Comparing Performance and Learning from Global Experience’ (O’Donnell et al. 2017; Farole 2011). The above overview of the birth and global spread of SEZs reveals that ‘learning’ or a variant thereof has been integral to the global spread of SEZs from their inception. Yet, such learning has little to do with postmodern euphemisms about policy diffusion. Instead, learning is part and parcel of the maintenance of global hegemony by capital over labour, with the postcolonial nation-state often reduced to the role of conduit for the interests of multinational capital even when national budgets pay for the lion’s share of investment in fixed capital from infrastructure to industrial estates and factories. Indeed, the transmission of the SEZ policy model from Puerto Rico initially happened through Western government promotion and private-sector marketing agencies like ADL, moving on into the realm of international organizations in the 1960s and after. In the process, gendered exploitation became a mainstay of SEZ manufacturing enterprises. Reports of grave violations of the human right to decent work in SEZs continue to pile up, with large-scale dispossession of subsistence farmers in India and elsewhere to make way for SEZ land development creating additional hardship (Seneviratne 2019; Shakya 2018; Campbell 2018; Mezzadri 2017; Sampat 2015; Reyes 2015; Kleibert 2015; Doucette and Lee 2015). At the same time, my interviews with national development agency officials in several postcolonial nations over the past ten years have revealed that nowadays most investors in manufacturing and in a growing number of other sectors relocate their production only if they are offered incentives, whether genuine SEZ treatment or other subsidies ranging from direct monetary payments by national governments to massive tax waivers that often equal tens of thousands of US dollars per job created. At the same time, research on the effects of FDI on national economies ignores the impact of SEZs and their negative impact on national social welfare programmes (Rudra 2008; Mosley 2010). Yet, SEZs are, initially, a massive financial risk for any government, which is why the question arises how SEZ manufacturing enterprises can be persuaded to stay in a given place at least until national governments have recouped their upfront expenses for fixed capital assets. However, this
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
The Political Economy of Special Economic Zones 201 requires a global agreement on minimum taxation for industrial ventures and a ban on tax incentives in SEZs. Further, decent work standards and minimum wages are needed to enable workers in the zones to recover the reproduction costs for their labour and their households and to escape unhealthy factory regimes. In this regard, it is important to highlight that the many decades of spreading SEZs, refining and developing the concept, and learning from existing zones hardly ever tackled the problem of escalating competition within the global SEZ sphere and the deindustrialization and crisis that have set in, even in those nations and regions with successful SEZs. In Puerto Rico, for example, the 1950s boom came to an abrupt ending when zones with similar benefits for capital opened in the Philippines, Mexico, South Korea, and elsewhere. Likewise, from the late 1980s, the successful Mauritian EPZ came under threat from relocations to neighbouring countries with cheaper and larger labour forces and the zone only managed to survive until the early 2000s because the government was able to m aintain preferential export quotas to Europe and North America. Meanwhile, SEZs in neighbouring countries went from crisis to crisis due to a series of civil wars and waves of strikes (Neveling 2015a). Industrial hubs, and variants of the concept, such as industrial clusters and vertically integrated industries, have surfaced in several national and SEZ promotion policies since the 1970s. In fact, a survey of The Economist’s archives reveals that some of the regions in Western capitalist nations that suffered from relocations to early SEZs tried to rebrand themselves as industrial hubs. The textile and garment heartlands of the North-eastern United States, notably New Hampshire, had all the qualities that Alfred Marshall would have wanted to see from the industrial districts he praised as hotbeds of industrialization in the late nineteenth century. Minneapolis, in the Midwest, tried to promote itself with the backing of the local city council and numerous local businesspeople as an ‘industrial hub’ in the 1960s (The Economist 1961) and South Africa advertised itself as Southern Africa’s industrial hub once global protests against the apartheid regime began to bite in the late 1960s (The Economist 1968). Crucially, the early stages of development in all SEZs from the 1940s to the present create locational factors that fundamentally contradict the principles of industrial districts and their twenty-first-century extension to industrial hubs. Initial investment in the so-called sweatshop phase of SEZ labour does not focus on skilled workers and, if that phase involves sophisticated distribution networks, cooperation, and trust, those networks commonly lack backward and forward linkages into local economies. Even though investors may regularly lament low productivity, absenteeism, and low work morale, relocations happen because wages are low and most factories have an interest in on-the-job training, which allows them to maintain a good turnover of young, often predominantly female workers who have little reason to demand higher wages and who lack the social cohesion needed to run successful wildcat strikes or form trade unions. In Puerto Rico, for example, there was a limited interest in establishing industrial districts and clusters when the local government supported industrial estates in the 1950s (Lebrón 1955). The Shannon Free Zone represented a very small cluster of factory buildings that had 8,000 workers in its peak years with a much larger number of workers spread out over several other factories in the wider Irish Mid-West region. In Mauritius, the government set up two industrial estates in Plaine Lauzun and in Coromandel in the early 1970s. Both projects ran over budget and the government struggled to service World Bank loans incurred to shoulder construction
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
202 Patrick Neveling expenses, so that 18.5 per cent of Mauritian sovereign debt at the time of near-default in the late 1970s was to the World Bank for those projects. The success years from 1983 instead saw small and medium-sized factories mushrooming across the island, often in neighbourhoods so that factories had no transport costs and local women workers could manage working days of over fourteen hours on the shop floor and in the home (Neveling 2015c, 2017c). Likewise, early PRC efforts to boost SEZs were highly uneven with different agencies in charge of the genuine industrial estate planned in Shekou near Shenzhen and of the rather randomly emerging factories in the wider Shenzhen region. Ambitions to turn SEZs into something akin to today’s industrial hubs have certainly been in place at planning stage or emerged during early years of operation in many nations. Yet, those ambitions have always suffered from the failure of SEZs to deliver on the promise of technology transfer. As bilateral and multilateral trading agreements until the late 1990s regulated quotas on the basis that the zones would part-process goods (also under US tariff clauses 806.30 and 807.00 applied in Mexican maquilas), there was no option to establish a full production process in SEZs so that the capital-intensive production steps were located in research and design departments in advanced capitalist nations. Vertical integration, for example by way of setting up spinning mills and dyeing factories alongside garment factories, is often hindered by the fact that raw materials such as cotton come from far-away places even when international agreements such as the African Growth and Opportunities Act (AGOA) support the concentration of manufacturing steps in a wider region. The question thus remains: how could SEZs turn into industrial hubs and establish a political economy that captures multinational runaway capital in space? The interest in such ‘hubbing’ is certainly driven by a new awareness of the potentially unsustainable nature of SEZs which, although they may benefit from the much-cited ‘flying geese’ model, also prod uce their own flying geese as soon as labour costs rise or competition emerges in other countries with ever-lower taxation and wage levels.
References Akinci, Gokhan, James Crittle, and FIAS/The World Bank Group (2008) Special Economic Zones: Performance, Lessons Learned, and Implications for Zone Development. Washington, DC: World Bank. Amin, Ash and Nigel Thrift (1992) ‘Neo-Marshallian Nodes in Global Networks’, International Journal of Urban and Regional Research 16(4): 571–87. Anderson, S. Axel (1934) ‘Location Factor in the Choice of Free Port Sites’, Economic Geography 10(2): 147–59. Baissac, Claude (2011) ‘Brief History of SEZs and Overview of Policy Debates’, in Thomas Farole (ed.) Special Economic Zones in Africa: Comparing Performance and Learning from Global Experiences. Washington, DC: World Bank, pp. 23–60. Boris, Eileen (1996) ‘Needlewomen under the New Deal in Puerto Rico, 1920–1945’, in Altagracia Ortiz (ed.) Puerto Rican Women and Work: Bridges in Transnational Labor. Philadelphia, PA: Temple University Press, pp. 33–54. Bräutigam, Deborah and Tang Xiaoyang (2011) ‘African Shenzhen: China’s Special Economic Zones in Africa’, The Journal of Modern African Studies 49(01): 27–54. Campbell, Stephen (2018) Border Capitalism, Disrupted: Precarity and Struggle in a Southeast Asian Industrial Zone. Ithaca, NY: Cornell University Press. Chaudhuri, Kirti Narayan (1985) Trade and Civilisation in the Indian Ocean: An Economic History from the Rise of Islam to 1750. Cambridge: Cambridge University Press.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
The Political Economy of Special Economic Zones 203 Chomsky, Aviva (2008) Linked Labor Histories: New England, Colombia, and the Making of a Global Working Class, American Encounters/Global Interactions. Durham, NC: Duke University Press. Currie, Jean and Economist Intelligence Unit (1985) ‘Export Processing Zones in the 1980s: Customs-free Manufacturing’, EIU Special Report No. 190.’ London: Economist Intelligence Unit. Dicken, Peter (2003) Global Shift: Reshaping the Global Economic Map in the 21st Century. 3rd edition. London: Sage. Diefenderrer, M. M. (1946) ‘Puerto Rican Lure’. Wall Street Journal, 7 June. Dietz, James L. (1986) Economic History of Puerto Rico: Institutional Change and Capitalist Development. Princeton, NJ: Princeton University Press. Doucette, Jamie and Seung-Ook Lee (2015) ‘Experimental Territoriality: Assembling the Kaesong Industrial Complex in North Korea’, Political Geography 47: 53–63. Farole, Thomas (2011) Special Economic Zones in Africa: Comparing Performance and Learning from Global Experiences. Washington, DC: World Bank. Farole, Thomas and Gokhan Akinci (2011) Special Economic Zones: Progress, Emerging Challenges, and Future Directions. Washington, DC: World Bank. Fenner, Justus (2013) ‘Configurando la cadena de café: casas mercantiles alemanas y consumo de café guatemalteco en Alemania, 1889–1929’, América Latina en la historia económica 20(3): 28–55. Fernández-Kelly, María Patricia (1983) For We Are Sold, I and my People: Women and Industry in Mexico’s Frontier, SUNY Series in the Anthropology of Work. Albany, NY: State University of New York Press. Frank, Andre Gunder (1998) ReOrient: Global Economy in the Asian Age. Berkeley, CA: University of California Press. Fröbel, Folker, Jürgen Heinrichs, and Otto Kreye (1981) The New International Division of Labour: Structural Unemployment in Industrialised Countries and Industrialisation in Developing Countries. Cambridge: Cambridge University Press. García-Colón, Ismael (2009) Land Reform in Puerto Rico: Modernizing the Colonial State, 1941–1969. Gainesville, FL: University Press of Florida. Harreld, Donald J. (ed.) (2015) A Companion to the Hanseatic League, Brill’s Companions to European History. Leiden: Brill. Kasmir, Sharryn and Lesley Gill (2018) ‘No Smooth Surfaces: The Anthropology of Unevenness and Combination’, Current Anthropology 59(4): 355–77. Kelleher, Tom (1976) Handbook on Export Free Zones. Industrial Development Abstracts. Vienna: United Nations Industrial Development Organization Archives. Ken, Wong Lin (1978) ‘Singapore: Its Growth as an Entrepot Port, 1819–1941’, Journal of Southeast Asian Studies 9(1): 50–84. Kleibert, Jana Maria (2015) ‘Islands of Globalisation: Offshore Services and the Changing Spatial Divisions of Labour’, Environment and Planning A 47(4): 884–902. Koistinen, David (2000) ‘Dealing with Deindustrialization: Economics, Politics, and Policy during the Decline of the New England Textile Industry, 1920–1960’, The Journal of Economic History 60(2): 501–4. Lebrón, Luis M. Rodríguez (1955) ‘Planned Industrial Districts as a Tool for Development in Puerto Rico’. MA thesis, Massachusetts Institute of Technology. Lewis, W. Arthur (1954) ‘Economic Development with Unlimited Supplies of Labour’, The Manchester School 22(2): 139–91. Lin, Justin Yifu (2014) ‘Industrial Policy Revisited: A New Structural Economics Perspective’, Revue d’économie du développement 22(1): 51–70. Madani, Dorsati (1999) ‘A Review of the Role and Impact of Export Processing Zones (English)’. Policy Research Working Paper No. WPS 2238: 114. Marhoz, Jean-Paul and Marcela Szymanski (1996) ‘Behind the Wire—Anti-Union Repression in Export Processing Zones’. Bruxelles: International Confederation of Free Trade Unions. Available at http://actrav.itcilo.org/actrav-english/telearn/global/ilo/frame/epzicftu.htm. Marshall, Alfred and John K. Whitaker. (1975a) The Early Economic Writings of Alfred Marshall, 1867–1890, Vol. 1. London: Macmillan for the Royal Economic Society.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
204 Patrick Neveling Marshall, Alfred and John K. Whitaker (1975b) The Early Economic Writings of Alfred Marshall, 1867–1890, Vol. 2. London: Macmillan for the Royal Economic Society. Marzagalli, Silvia (1996) ‘Port Cities in the French Wars: The Responses of Merchants in Bordeaux, Hamburg and Livorno to Napoleon’s Continental Blockade, 1806–1813’, The Northern Mariner/Le Marin du nord 6(4): 65–73. Mezzadri, Alessandra (2017) The Sweatshop Regime: Labouring Bodies, Exploitation, and Garments Made in India. New York: Cambridge University Press. Mosley, Layna (2010) Labor Rights and Multinational Production, Cambridge Studies in Comparative Politics. Cambridge: Cambridge University Press. Neveling, Patrick (2015a) ‘Export Processing Zones and Global Class Formation’, in James Carrier and Don Kalb (eds) Anthropologies of Class: Power, Practice, Inequality. Cambridge: Cambridge University Press, pp. 164–82. Neveling, Patrick (2015b) ‘Export Processing Zones, Special Economic Zones and the Long March of Capitalist Development Policies during the Cold War’, in Leslie James and Elisabeth Leake (eds) Negotiating Independence: New Directions in the Histories of the Cold War and Decolonisation. London: Bloomsbury, pp. 63–84. Neveling, Patrick (2015c) ‘Flexible Capitalism and Transactional Orders in Colonial and Postcolonial Mauritius: A Post-Occidentalist View’, in Jens Kjaerulf (ed.) Flexible Capitalism: Exchange and Ambiguity at Work. Oxford: Berghahn, pp. 207–34. Neveling, Patrick (2015d) ‘Free Trade Zones, Export Processing Zones, Special Economic Zones and Global Imperial Formations 200 bce to 2015 CE’, in Immanuel Ness and Zak Cope (eds) The Palgrave Encyclopedia of Imperialism and Anti-imperialism. Basingstoke: Palgrave Macmillan, pp. 1007–16. Neveling, Patrick (2017a) ‘Capital over Labor: Health and Safety in Export Processing Zones Garment Production since 1947’, in Rebecca Prentice and Geert de Neve (eds) After Rana Plaza: Rethinking the Health and Safety of Global Garment Workers. Philadelphia, PA: University of Pennsylvania Press, pp. 123–46. Neveling, Patrick (2017b) ‘The Global Spread of Export Processing Zones and the 1970s as a Decade of Consolidation’, in Knud Andersen and Stefan Müller (eds) Changes in Social Regulation—State, Economy, and Social Protagonists since the 1970s. Oxford: Berghahn Books, pp. 23–40. Neveling, Patrick (2017c) ‘The Political Economy Machinery: Toward a Critical Anthropology of Development as a Contested Capitalist Practice’, Dialectical Anthropology 41(2): 163–83. Neveling, Patrick (2019) ‘Industry and Manufacturing’, in Stefano Bellucci and Andreas Eckert (eds) The New General Labour History of Africa, Vol. 2: 1900–present. Basingstoke: Palgrave, pp. 177–94. Ngai, Pun (2005) Made in China: Women Factory Workers in a Global Workplace. Durham, NC: Duke University Press. O’Donnell, Mary Ann, Winnie Wong, and Jonathan Bach (2017) Learning from Shenzhen: China’s Post-Mao Experiment from Special Zone to Model City. Chicago, IL: University of Chicago Press. Ong, Aihwa (1987) Spirits of Resistance and Capitalist Discipline: Factory Women in Malaysia, SUNY Series in the Anthropology of Work. Albany, NY: State University of New York Press. Orenstein, Dara (2011) ‘Foreign-Trade Zones and the Cultural Logic of Frictionless Production’, Radical History Review 109: 36–61. Reinsberg, Bernhard, Alexander Kentikelenis, Thomas Stubbs, and Lawrence King (2019) ‘The World System and the Hollowing Out of State Capacity: How Structural Adjustment Programs Affect Bureaucratic Quality in Developing Countries’, American Journal of Sociology 124(4): 1222–57. Reyes, Victoria (2015) ‘Legacies of Place and Power: From Military Base to Freeport Zone’, City & Community 14(1): 1–26. Ríos, Palmira N. (1995) ‘Gender, Industrialization, and Development in Puerto Rico’, in Christine E. Bose and Edna Acosta-Belén (eds) Women in the Latin American Development Process. Philadelphia, PA: Temple University Press, pp. 125–48. Romero, Anna T. (1995) ‘Export Processing Zones: Addressing the Social and Labour Issues’ (draft), 28. Available at www.ilo.org/public/english/85multi/research/epzad.htm.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
The Political Economy of Special Economic Zones 205 Ross, David F. (1966) The Long Uphill Path: A Historical Study of Puerto Rico’s Program of Economic Development. San Juan: Talleres Graficos Interamericanos. Rudra, Nita (2008) Globalization and the Race to the Bottom in Developing Countries: Who Really Gets Hurt? Cambridge: Cambridge University Press. Sampat, Preeti (2015) ‘The “Goan Impasse”: Land Rights and Resistance to SEZs in Goa, India’, Journal of Peasant Studies 42(3–4): 765–90. Schrank, Andrew (2003) ‘Foreign Investors, “Flying Geese”‚ and the Limits to Export-led Industrial ization in the Dominican Republic’, Theory & Society 32(4): 415–43. Seneviratne, Prajna (2019) ‘Estranged Labour on the Global Assembly Line? “Sewing Girls” of Postcolonial Garment Factories in Sri Lanka’, Journal of International Women’s Studies 20(2): 151–65. Shakya, Mallika (2018) Death of an Industry: The Cultural Politics of Garment Manufacturing during the Maoist Revolution in Nepal. New Delhi: Cambridge University Press. Sharma, Patrick (2015) ‘Between North and South: The World Bank and the New International Economic Order’, Humanity: An International Journal of Human Rights, Humanitarianism, and Development 6(1): 189–200. Stiglitz, Joseph E., Justin Yifu Lin, and Célestin Monga (2013) The Rejuvenation of Industrial Policy. Washington, DC: World Bank. Strümpell, Christian (2014) ‘The Politics of Dispossession in an Odishan Steel Town’, Contributions to Indian Sociology 48(1): 45–72. The Economist (1968) ‘South Africa’, advertisement. 9 November 1968, 10. Available at http://tinyurl. galegroup.com/tinyurl/9mzj85. The Economist (1961) ‘Twin City Reborn’, 9 September. Available at http://tinyurl.galegroup.com/ tinyurl/9mzpm9. Thoman, Richard S. (1956) Free Ports and Foreign Trade Zones. Cambridge, MD: Cornell Maritime Press. Tugwell, Rexford G. (1947) The Stricken Land, the Story of Puerto Rico, 1st edition. Garden City, NY: Doubleday. United Nations Centre on Transnational Corporations and International Labour Organization (1988) Economic and Social Effects of Multinational Enterprises in Export Processing Zones. Geneva: International Labour Office. Warr, Peter G. (1995) ‘Explaining the Thai Miracle: Dragons, Planners, and Other Myths’. Economics Division Working Papers Southeast Asia. Research School of Pacific Studies, ANU, Canberra. Whitham, Charlie (2013) ‘The Committee for Economic Development, Foreign Trade and the Rise of American Corporate Liberalism, 1942–1948’, Journal of Contemporary History 48(4): 845–71. Wong, Siu-lun (1988) Emigrant Entrepreneurs: Shanghai Industrialists in Hong Kong. Hong Kong: Oxford University Press. World Bank Country Study (1990 [1989]) Mauritius: Managing Success. Washington, DC: The International Bank for Reconstruction and Development/World Bank. Zeng, Douglas Zhihua (ed.) (2010) ‘Building Engines for Growth and Competitiveness in China: Experience with Special Economic Zones and Industrial Clusters’, Vol. 56, 447. Directions in Development—Countries and Regions. Washington, DC: The International Bank for Reconstruction and Development/World Bank.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
chapter 12
Cluster Dy na mics a n d R egiona l N et wor ks New Argonauts, Silicon Valley, and Route 128 AnnaLee Saxenian
12.1 Introduction Regional clusters and their competitive advantages have attracted growing research and policy focus. This chapter reviews cluster dynamics, regional advantages, and the talent flow amongst the leading innovation hubs, notably Silicon Valley, and emerging hubs in newly industrializing economies. In exploring cluster dynamics, cultures, and networks, the chapter compares Silicon Valley and Route 128, the leading innovation hubs in the United States. It provides an alternative network approach to examine and explain the dynamics of industrial hubs and cluster dynamics. It also shows that economies of the periphery (such as Hsinchuu Science Park in Taipei) can rely on strategies of talent circulation and brain gain through collaboration with leading innovation hubs. Through a comparative study of firms in Silicon Valley and Route 128, the chapter explains the uneven performance between the two regional hubs, and provides new insights on regional clusters. This has implications for countries exploring effective policy mechanisms for cluster dynamics to develop technological capabilities.1 Entrepreneurs and their far-flung networks now play an important role in the global expansion of the technology industries—and are an increasingly important determinant of developmental outcomes here and abroad. Ventures like July Systems and VeriSilicon are amongst the thousands of start-ups that have already contributed to the creation of dynamic technology clusters in locations ranging from Israel and Taiwan to India and China. These investments may be small by comparison with total foreign direct investment—but by supporting indigenous entrepreneurship they create a visible alternative to the state-led or family-owned companies that dominate many developing economies. They also contribute 1 See Porter (1990).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Cluster Dynamics and Regional Networks 207 to the creation of local ecosystems that support entrepreneurial experimentation and cap ability building—which in turn can attract subsequent rounds of investment. Such a process is most evident in Silicon Valley, where networks of foreign-born e ngineers and entrepreneurs are transferring technical and institutional know-how between distant regional economies faster and more flexibly than most multinationals. The main actors in this process are not large corporations but rather the ‘new Argonauts’; these networks of foreign-born engineers, entrepreneurs, managers, and professional service providers (such as lawyers and bankers) have the linguistic and cultural capabilities as well as the institutional knowledge to collaborate with their home-country counterparts. While systematic data on these highly decentralized two-way flows of skill, technology, and capital is scarce, their impacts are arguably as important as more easily measured multinational investments. The migration of talented youth from developing to advanced countries was viewed in the postwar decades as a ‘brain drain’ that exacerbated international inequality by enriching already wealthy economies at the expense of their poor counterparts. Data on these trends is hard to find, but the UN has estimated a total of 300,000 highly skilled emigrants from all developing countries moved to the West during the 1960s; the 1990 census showed 2.5 million highly skilled immigrants residing in the United States, excluding students. This chapter is organized as follows. We first review the ‘new Argonauts’ and the rise of cross-regional entrepreneurship, followed by a discussion of firm dynamics, talent gain, and cross-regional start-ups. After reviewing agglomeration economies and regional advantages, with a special focus on Silicon Valley and Route 128, we explore the network approach to regions and industrial hubs, and the synergy between regional networks and firm production systems respectively. The focus of the chapter is a specific and significant period in the transformation of the two clusters (Silicon Valley and Route 128) that occurred between the 1970s and 1990s.
12.2 The New Argonauts and the Rise of Cross-regional Entrepreneurship Like the Argonauts of Greek mythology who ventured with Jason centuries ago, these US-educated entrepreneurs undertake the risky but often rewarding search for wealth in distant lands. Armed with deep knowledge of technology markets and global rolodexes, the new Argonauts are ideally positioned to mobilize the expertise and capital needed to start successful global ventures. Their success is forcing us to think afresh about how nations and regions grow. In the late 1990s immigrants were at the helm of nearly 30 per cent of all Silicon Valley start-ups. These entrepreneurs, overwhelmingly from peripheral economies like India and China, started technology businesses in the region at the same rate as their US-born counterparts and have become global entrepreneurs, based in Silicon Valley while tapping low-cost technical talent and financing in their home countries. Others return home to start businesses but continue working with customers and partners in Silicon Valley (Saxenian and Hsu 2005). As these cross-regional collaborations multiply and deepen they are bringing benefits to both the US and foreign economies.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
208 AnnaLee Saxenian This globalization of entrepreneurial networks reflects dramatic changes in global labour markets. The falling cost of transportation and communications has facilitated dramatic increases in the mobility of high-skilled workers, while digital technologies have accelerated the formalization and exchange of vast amounts of information across long distances. International migration, historically a one-way process, has become a reversible choice, particularly for those with scarce technical skills, and it is now possible to collaborate in real time, even on complex tasks, with counterparts located at great distances. Scientists and engineers from developing countries—once forced to choose between settling abroad and returning home to far less attractive professional opportunities—are contributing to their home economies while maintaining professional and business ties in more technologically advanced economies.
12.2.1 From Brain Drain to Brain Circulation California’s Silicon Valley has benefited disproportionately from this process. The region’s technology producers grew very rapidly from the 1970s through to the 1990s, absorbing technical skill voraciously, irrespective of their national origin. Tens of thousands of talented immigrants from developing countries who initially came to the United States for graduate engineering education accepted jobs in Silicon Valley rather than return to their home countries where professional opportunities were limited. By the end of the 1990s over half of Silicon Valley’s 200,000 scientists and engineers were foreign born, primarily from Asia, and only a small proportion planned to return to their home countries. These immigrants, often excluded from established networks, quickly created ethnic social and professional networks that have supported career advancement and entrepreneurial success in Silicon Valley’s open labour markets (Saxenian and Hsu 2005). The successes of high-profile startups like Sabeer Bathia’s Hotmail, Jerry Yang’s Yahoo, and Min Zhu’s Webex are only the most visible reflections of the extent to which Silicon Valley’s immigrant engineers have mastered the region’s entrepreneurial business system. The same individuals who left poor countries like India and China for better professional and economic opportunities abroad are now increasingly reversing the brain drain, transforming it into ‘brain circulation’ as they return home to establish business relationships or start new companies while maintaining their social and professional ties to the United States. Foreign-educated engineers-turned-venture capitalists often lead this process by investing in their home countries. As experienced engineers, managers, and investors return home, either temporarily or permanently, they transfer first-hand knowledge of US capital markets and business models to peripheral regions. These individuals also often serve as advisers to domestic policymakers anxious to promote technology growth. These crossregional technical communities have the potential to jump-start local entrepreneurship— but they succeed over the long term only if they build alliances with technical professionals, businesses, and policymakers in their home countries. In the early 1980s foreign-born engineers transferred the Silicon Valley model of early-stage high-risk investing to Taiwan and Israel—locations that US venture capitalists typically had neither the interest in nor the ability to serve. Native-born investors provided the cultural and linguistic know-how needed to operate profitably in these markets. In addition to capital, they brought technical and operating experience, knowledge of new
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Cluster Dynamics and Regional Networks 209 business models, and networks of contacts in the United States. Israel and Taiwan today boast the largest venture capital industries outside North America (US$4 billion is invested annually in Israel and US$1.3 billion in Taiwan). Both have high rates of new firm formation, innovation, and growth. Israel is now known for software and Internet firms like Mirablis (an instant-messaging programme developer) and Checkpoint (security software); Taiwan has become a centre for leading-edge personal computer (PC) and integrated circuit (IC) manufacturing with firms like Acer (PCs and components) and TSMC (semiconductor foundry). Immigrants from India and China with experience in Silicon Valley have started to influence economic development in their home countries as well by transferring technology and know-how when they return home to work or start businesses, as well as indirectly by influencing the formation of policy and other aspects of the institutional environment. By 2004, venture capital and private equity firms were investing more than US$1 billion annually in enterprises located in China and a comparable amount in India. While this is a fraction of the venture capital invested annually in the United States or even the amount of FDI in these economies, it is contributing to the development of local ecosystems that support indigenous entrepreneurship and an alternative, increasingly competitive, trajectory to the development opportunities provided by both the established domestic firms and the multinational corporations in these nations. Not surprisingly, this was a significantly easier process in the small nations like Israel and Taiwan than in the complex political economies of China and India. Nonetheless, the long-term impacts of returning entrepreneurs and their communities will likely be more far-reaching in the latter countries.
12.2.2 Entrepreneurship in the Periphery Traditional accounts of economic development have assumed that new products and technologies emerge in industrialized nations that combine sophisticated skill and research capabilities with large, high-income markets—and that mass manufacturing is shifted to less costly locations once the product is standardized and the process stabilized. Success in this view builds on success in advanced economies, while peripheral economies remain followers. This divide is perpetuated by both the strategies of multinational corporations and the tendencies towards agglomeration created by the economics of increasing returns. This model leaves little room for the development of independent technological capabilities in the periphery. At best, foreign investment from the core might contribute to the incremental mastery of manufacturing techniques and upgrading of local suppliers.2 Even the most successful newly industrializing countries are destined to remain imitators in this model because leading-edge skill and technology reside in the corporate research labs and universities in the core. The primary route to development in the periphery, in this view, is the mobilization by the state, in conjunction with local banks and industry, of the resources to either develop or import the mass manufacturing capabilities that were perfected in the core. Transformations in the world economy have undermined the power of this core–periphery model, however. The increasing mobility of highly skilled workers and information on 2 See Whiting (1987), Piore and Sabel (1984).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
210 AnnaLee Saxenian the one hand, and the fragmentation of production in information and communications technology sectors on the other, provide unprecedented opportunities for formerly peripheral economies. Regions that missed the postwar economic boom, in particular, have provided fertile environments for a decentralized growth based on entrepreneurship and experimentation. The key actors in this process are neither policymakers nor multinational corporations in isolation, although both certainly play a role, but rather communities of technically skilled immigrants with work experience and connections to Silicon Valley and related technology centres. US-educated and trained engineers are increasingly transferring know-how and market information and helping to jump-start local entrepreneurship, allowing their home economies to participate in the information technology revolution. Because of their experience and professional networks, these cross-regional entrepreneurs can quickly identify promising new market opportunities, raise capital, build management teams, and establish p artnerships with other specialist producers—even those located far away. The ease of communication and information exchange within ethnic professional networks accelerates learning about new sources of skill, technology, and capital, as well as about potential collaborators. It also facilitates the timely responses that are essential in a highly competitive environment. This decentralized responsiveness is an advantage that few multinationals can claim. This is not a one-way process. As recently as the 1970s, only large, established corpor ations had the resources and capabilities to grow internationally, and they did so primarily by establishing marketing offices or manufacturing branch plants overseas. Today the fragmentation of production and the falling costs of transportation and communication allow even small firms to build partnerships with foreign producers to tap overseas expertise, cost savings, and markets. Start-ups in Silicon Valley today are often global actors from their first day of operations; many raise capital, subcontract manufacturing or software development, and market their products or services outside the United States. The scarce resource in this environment is the ability to locate foreign partners quickly and to manage complex business relationships and teamwork across cultural and linguistic barriers. This is particularly challenging in high-tech industries in which products, markets, and technologies are continually redefined—and where product cycles are often nine months or less. First-generation immigrants like the Chinese and Indian engineers in Silicon Valley who have the language, cultural, and technical skills to function well in the United States as well as in their home markets have a commanding advantage here (Saxenian and Hsu 2005). They have created institutions and social structures that enable even the smallest producers to locate and maintain mutually beneficial production and markets. Late-developing economies typically face two major disadvantages: they are remote from the sources of leading-edge technology, and they are distant from developed markets and the interactions with users that are crucial for innovation. Firms in peripheral locations use a variety of mechanisms to overcome these disadvantages, from joint ventures and technology licensing to foreign investment and overseas acquisitions. However, a network of technologists with strong ties to global markets and the linguistic and cultural skills to work in their home country is arguably the most efficient and compelling way to overcome these limitations. Cross-regional entrepreneurs and their communities can facilitate the diffusion of technical and institutional know-how, provide access to potential customers and partners, and help to overcome reputational as well as informational trade barriers for isolated economies.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Cluster Dynamics and Regional Networks 211 The increasing sophistication of information and communications technologies and the liberalization of global markets have accelerated this process. It is now quick, simple, and inexpensive to communicate internationally and to transfer information between distant locations. Information systems that facilitate the formalization of knowledge are dramatically expanding the volume as well as the variety of possible forms of information exchange. However, information technology alone cannot ensure successful coordination or efficient transfers of technical and institutional knowledge. Long-distance collaborations still depend heavily upon a shared social context and language that ensures mutual intelligibility between partners, particularly as speed and responsiveness are essential in today’s technology competition. Market liberalization has been equally important to the economic transformation of countries like China and India. However, the reduction of trade barriers and bureaucratic intervention alone does not create the institutional and social context, let alone the domain knowledge, required to sustain entrepreneurial success in global industries. Technology entrepreneurship remains highly localized even in the most advanced economies and it cannot be created by fiat, as evidenced by decades of failed attempts to ‘grow the next Silicon Valley’. Efforts to jump-start entrepreneurship by mobilizing researchers, capital, and a modern infrastructure cannot replicate the shared language and trust of a technical community that permits open information exchange, collaboration, and learning (often by failure) alongside intense competition in places like Silicon Valley. The new technology centres differ significantly from one another, and from Silicon Valley, in their technological sophistication as well as the specializations of local producers. Cross-regional entrepreneurs rarely compete head-on with established US producers; instead they build on the skills and the technical and economic resources of their home countries. Israeli entrepreneurs, for example, have successfully applied the findings of the nation’s advanced military research to innovations in the Internet security and telecommunications arenas. Indian entrepreneurs, by contrast, recognized the opportunity to mobilize the thousands of underemployed English-speaking Indian engineers to provide software development services for American corporations. Returning entrepreneurs are ideally positioned to identify appropriate market niches, mobilize domestic skill and knowledge, connect to international markets, and work with domestic policymakers to identify and devise strategies to overcome obstacles to further growth.
12.3 Firm Dynamics: Cross-Regional Start-ups and Talent Gain A new generation of cross-regional start-ups combine Silicon Valley’s new product vision, technology architecture, marketing, and research and development coordination with the technical capabilities in distant regions, including R&D implementation and the advanced manufacturing and production logistics available in Taiwan and China, and the sophisticated management of large-scale software development, IT services provision, and research and development implementation in India. While investors and entrepreneurs in crossregional start-ups may initially be motivated by the availability of low-cost skill in the periphery, over time they remain because of the quality and productivity advantages in
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
212 AnnaLee Saxenian these locations (Saxenian and Hsu 2005; Saxenian 1994). In fact, the clustering of technology production has generated rapidly rising wages and intensifying congestion in places like Bangalore and Shanghai, where engineering salaries are now amongst the highest in their nations, yet new and established firms cluster there rather than seeking lower-cost locations. Cross-regional entrepreneurs and their communities are not replicating Silicon Valley. Variations in national economic and political institutions, themselves the products of diverse histories and cultures, ensure divergent economic trajectories.3 It is more appropriate to see the emerging regions as hybrids, combining elements of the Silicon Valley industrial system with inherited local institutions and resources. Returning entrepreneurs typically seek (with varying success) to transfer venture capital, merit-based advancement, and corporate transparency to economies with traditions of elite privilege, government control, and corruption. They seek to reproduce the team-based firm with a limited hier archy in an environment dominated by family-run businesses or state-owned enterprises. And they seek to influence policy because the national institutions that support the Silicon Valley system—efficient capital markets, an independent judiciary, regulatory oversight, and sophisticated education systems, research institutions, and physical infrastructure—are rarely present in these peripheral economies. Technology entrepreneurship is nonetheless a growing presence in each of these econ omies. Returning entrepreneurs have developed a variety of adaptations to the challenges created by conditions in their home countries. In India, entrepreneurs rely on private telecommunications facilities and power supplies rather than on the nation’s costly and unreliable infrastructure, while in China, returning entrepreneurs have learned how to negotiate the complex bureaucratic rules and policies that regulate private companies. They also rely on US institutions: in addition to graduate training in the United States, many establish headquarters or research labs in Silicon Valley, locate venture capital, professional services, and managerial and technical talent from the United States, and even raise funds on US capital markets. These cross-regional start-ups still face significant challenges. Venture investment is still in the early stages in most of the world: there has not yet been a complete cycle of investments and reinvestments in a second generation of entrepreneurs. There are shortages of experienced managerial talent and ongoing difficulties coordinating distant activities, particularly in developing organizational synergy and persistent, consistent communication. Entrepreneurshipled growth, with highly competitive, specialized technology producers in high-skill regions connecting to, and collaborating with, counterparts elsewhere is only one possible future for these regions. They could forgo the opportunity to upgrade local skills and capabilities, and instead remain suppliers of low-cost labour to global (or domestic) corporations. China and India have the labour supply to do this for a relatively long time. However, many cross-regional entrepreneurs have maintained close ties to the technology and markets of Silicon Valley, and are constructing firms committed to an alternative, high-value-added trajectory.
12.3.1 Looking Ahead Is brain circulation between technology regions making the world ‘flat’? Hardly. The new Argonauts cluster tenaciously in the leading technology centres, w hich is why Palo Alto 3 Bulkeley (1987).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Cluster Dynamics and Regional Networks 213 now has more in common with Taipei or Tel Aviv than it does with Fresno, a three-hour drive away. Likewise, residents of Bangalore enjoy Western standards of living, while those in the nearby rural areas remain mired in poverty. Economic geographers have documented this phenomenon of increasing returns in specific locations—where the advantages of locating in a crowded and costly place outweigh the increasing costs, resulting in a pronounced clustering process. The rise of entrepreneurship-led growth suggests that the regional cluster may be replacing the nation as the locus of economic growth. Overseas technology investments are not motivated solely by low labour costs, as critics suggest. The leading destinations for cross-border technology investment today are regions like Bangalore and Shanghai, where wages and other costs are significantly higher than their surrounding economies, and rising more rapidly. Even when low wages attract initial investments, local enterprises distinguish themselves from other low-cost regions by collaborating with Silicon Valley-based partners. This allows local producers to develop specialized skills, expertise, and relationships that ensure a regional advantage that compensates for their high costs: Israel in sophisticated Internet and security technologies, Taiwan in global logistics and design, China in efficient IT manufacturing, and India in managing complex software services and consulting projects. The old pattern of one-way flows of technology and capital from the core to the periphery is being replaced by a far more complex and decentralized two-way flow of skill, capital, and technology between differently specialized regional economies. Silicon Valley is now at the core of this rapidly diversifying network of economies because it is the largest and most sophisticated market as well as a leading source of new technologies. However, this too could change: the relationships between these emerging technology regions are multiplying and new markets are opening up that promise to further transform the world economy. The fast-growing market for wireless communication in Asia, for example, has created oppor tunities for firms in China and India to contribute to the direction of the technology and its applications—even if they do not define the leading edge of the technology. Over time, producers in developing regions can build independent capabilities and define entirely new specializations and markets. The new regional economies are not replicas of Silicon Valley—although institutions and professional service providers from that region are fast expanding into these new locations. These regions have co-evolved with Silicon Valley. Firms in these regions do not typically seek to compete directly with Silicon Valley producers; they focus instead on developing capabilities in areas that US producers are not pursuing and over time they are transforming activities once regarded as mundane and low tech into more efficient and dynamic sectors. Taiwan was known in the 1980s for its cheap PC clones and components; today it is recognized for the flexibility and efficiency of its IC and electronic systems producers. China was known in the 1990s for ‘me-too’ Internet ventures; today Chinese producers are poised to play a lead role in developing wireless technology. In the 1990s, India was a provider of labour-intensive software coding and maintenance; today local companies are managing large-scale software services projects for leading global corporations. Israel was a low-cost location for research in the 1980s; since then local entrepreneurs have pioneered sophisticated Internet and security technologies. These developments explain why even the largest Silicon Valley corporations are active participants in all of these regions as investors and as partners, not simply as competitors. An established firm like Cisco designs and sources critical parts of its operating system
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
214 AnnaLee Saxenian software from India and application-specific integrated circuits from Israel for its high-end routers, and the manufacturing of most of its hardware from Taiwan and China (Saxenian and Hsu 2005; Saxenian 1994). Like Intel and Acer, it also invests in start-ups with promising technologies in these locations. On the other hand, a start-up (or ‘micro-multinational’) like July Systems obtained venture capital from firms based in the United States, Taiwan and China, and India, and its products will likely incorporate components from all of these locations as well as being targeted at all their markets. US technology producers now look to their counterparts in Taiwan and China, India, and Israel not simply for low-level implementation but increasingly for co-development and co-architecting of products and components. In addition, firms in the new technology regions are increasingly partnering with one another as well as with firms from Silicon Valley, as when a Taiwanese semiconductor firm invests in Israeli start-ups specialized in digital speech-processing chips, or when an Israeli company contributes intellectual property components to a chip design firm based in India. These collaborations deepen the capabilities of each of the partners and over time can support a process of reciprocal innov ation and upgrading in the respective regions.
12.3.2 A Model for Others? This is not to suggest that all developing economies are positioned to reap the benefits of brain circulation and peripheral entrepreneurship. This opportunity is benefiting countries that have invested heavily in higher education, typically technical education, and are sufficiently politically and economically stable for migrants to consider returning home. Some of the largest technically skilled immigrant groups in Silicon Valley have not built business or professional connections to their home countries for political reasons. Most of the region’s Iranian and Vietnamese immigrants, for example, are political refugees and hence not inclined to return to countries that, in any case, lack the economic stability needed for technology investment or entrepreneurship. This criterion applies in varying degrees to many of the developing economies that have technically skilled communities in the United States and at home, including Russia, parts of Eastern Europe, and Latin America. It is possible that urban areas like Saint Petersburg or Buenos Aires will become more attractive to returning entrepreneurs in the future as their economies develop and eventually provide greater professional opportunities for returnees.4 However, large parts of Africa and Latin America lack the skill base or political openness necessary to become attractive environments for technology entrepreneurship. The creation of a transnational community is a two-way process. While policymakers and planners can encourage cross-regional connections, they cannot create or substitute for transnational entrepreneurs and their decentralized networks. Government agencies from foreign countries regularly sponsor networking events for their expatriates in the Bay Area as a way to recruit return entrepreneurs and investments. However, the absence of entrepreneurial collaborators at home means that these agencies can provide incentives and information but not access to partners with an interest in jointly transforming the home environment. Governments cannot by themselves ensure the preconditions for return entre4 See Scott (1988a), Vernon (1960), Castells (1989).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Cluster Dynamics and Regional Networks 215 preneurship; this is inherently a process of collaborative institution building that requires both local knowledge and understanding of global technology markets and networks. Cross-regional networks develop only when skilled migrants are both willing and able to return to their home countries for business in large enough numbers to create close links to the technical community in the home country. The receptiveness of the home country depends upon factors such as political stability, economic openness, and level of economic development. It often builds on multinational investments in research and development that have contributed to a developing local skill base and infrastructure that supports entrepreneurship. The critical variable is the possession of political leaders willing to collaborate with returning entrepreneurs to develop a shared vision and remove institutional obstacles to entrepreneurship-led growth. Technology markets are shifting quickly, however, with demand growing rapidly outside the United States. In the early 2000s, the West and Japan produced more than half of global output but accounted for less than 15 per cent of the world’s population. This will change decisively in coming decades. Goldman Sachs predicts that customer demand from India and China will dominate global markets within a decade and that these two economies will be bigger than the US economy by 2050. Producers in other peripheral economies will surely develop the capabilities to participate in global networks as well. They will likely share with their predecessors a history of investments in education and research, as well as an institutional openness that ensures both competitive intensity and long-distance collaborations. Silicon Valley’s role as the dominant technology centre will continue to diminish. This does not mean that the region will decline, only that it will become one of many nodes in a more open and distributed global network of differently specialized and complementary regional economies.
12.4 Agglomeration Economies and Industrial Clusters Regional clusters have gained both scholarly and policy attention by virtue of their distinct competitive advantages. An extensive body of literature across several disciplines attributes the advantages arising from the clustering of economic activity primarily to external econ omies of scale. The concept of external economies has its limitations. It sees each industrial firm as an autonomous production unit with defined boundaries separating what happens inside and outside it. In doing so, the concept neglects the complex and evolving relations amongst the internal organization of firms and their linkages to one another and to local social and institutional dynamics. The work of Alfred Marshall (1920) advanced the concept of ‘external economies of scale’ or the productivity gains driven by factors beyond each firm. He theorized that when factors of production external to the firm such as infrastructure, services, labour, and knowledge are geographically concentrated, additional productivity benefits or ‘agglomeration economies’ occur. Once established in a locality, such an advantage becomes self-reinforcing through a dynamic process of increasing returns (Arthur 1990; Krugman 1991; Scott 1988b; Storper 1989).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
216 AnnaLee Saxenian The development trajectories of Silicon Valley and Route 128 are seen as typical cases of the Alfred Marshall-type external economies arising from industrial localization. Both clusters benefited from similar agglomeration effects resulting from spatially concentrated technical skill, venture capital, specialized input suppliers and services, infrastructure, and spillovers of knowledge associated with proximity to universities and informal information flows (see, for example, Castells 1989; Hall and Markusen 1985; Krugman 1991; Porter 1990; Scott 1988b). However, the performance of the two regional clusters diverged in the 1980s and 1990s. The external economies approach fails to explain this divergence (see Appendix: Figures A12.1 and A12.2). Despite their shared origins, Silicon Valley and Route 128 evolved differently in response to growing international competition. The growth of both clusters was associated with postwar military spending and a deep-rooted history of university-based research. While the two regions faced economic downturns in the 1980s, Silicon Valley regained its position as a lead technological cluster propped by a new surge of start-up businesses and the renewed dynamism of existing companies such as Intel and Hewlett-Packard (HP). Route 128 computer producers, the Digital Equipment Corporation (DEC) and other minicomputer companies, on the other hand, experienced sustained decline conceding their former dominance to Silicon Valley by the 1980s.5 While levels of employment in technology were similar in the two regions in 1975, between 1975 and 1990 Silicon Valley generated 150,000 new technology jobs, three times as many as Route 128. By 1990, Silicon Valley producers exported a third of the nation’s total electronics products and the cluster was home to 39 of the country’s fastest-growing electronics companies. Some of the fastest-growing firms at this time, such as Apple Computers and HP, were also based in Silicon Valley.6 The notion of external economies alone does not explain why a similar spatial concentration of specialized technical skills, suppliers, and information led to exponential growth in Silicon Valley and relative decline along Route 128.7 The spatial proximity of factors of production generating external economies is not indicative of the capacity of firms to respond to and thrive amidst rapidly evolving markets, technologies, and competitors. The complexity of institutional and social relationships that connect industrial firms is critical in understanding regional cluster dynamics, and why some thrive while others do not, despite similar initial conditions. Given the limits of external economies in explaining variations in cluster performance, an alternative network approach is proposed here to analyse regional economies and highlight the role of local drivers of competitive advantage. In a networked approach to regional clusters, the distinction between what is internal and external to industrial firms is diluted. The broadest interpretations of technological external economies recognize that firms learn from one another through the flow of information, ideas, and know-how (Storper 1989), but they do so only by denying the theoretical distinction between internal and external economies, between what is inside and outside the firm. 5 See Florida and Kenny (1990), Edmondson (1988), Greene (1990), Tuller (1988). 6 On HP, Sun, Appolo, see Nash and Hayes (1993), Nee (1991), Furlong (1991). See also Beam and Frons (1985), Bell and Corliss (1989), Basche (1991), Comerford (1992), CorpTech (1993), DeNucci (1991), Sheff (1989), McWilliams (1992), Kalb (1991). 7 See Scott (1988a).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Cluster Dynamics and Regional Networks 217
12.5 A Network Approach to Regions and Industrial Hubs Firms do not operate in isolation. They are embedded in networks of social and institutional relationships and are often shaped by their strategies and structures (Granovetter 1985). A network approach enables us to understand the evolution of relationships amongst firms and between firms and the social structures and institutions of their particular localities (Nohria and Eccles 1992b; Powell 1987). It helps, for example, to illuminate the rationale for the evolution of distinct industrial systems in Silicon Valley and Route 128 in the postwar period, despite the similarities of these two regions in origin and access to technologies. Unfortunately, the differences in productive organization have often been overlooked by economic analysts or treated simply as superficial differences between ‘laid-back’ California and the more ‘buttoned-down’ East Coast. On the contrary, the differences demonstrate the importance of local social and institutional determinants of industrial adaptation and help to explain why these two regions have responded so differently to the same external forces. Silicon Valley has a regional, network-based industrial system that promotes learning and mutual interest amongst producers of related technologies. The region’s dense social networks and open labour markets also encourage entrepreneurship and experimentation. Despite the intense competition, firms in the region learn from each other about changes in market conditions and new technologies through informal networking and communication. The informal team structures also encourage horizontal communication within the firm and between firms and suppliers and customers. In short, in a network-based system, the functional boundaries within firms are porous, as are the boundaries amongst firms and between firms and local institutions, such as trade associations and universities. In contrast, the Route 128 region is dominated by relatively large and autarkic corpor ations that internalize a wide range of productive activities. Secrecy and corporate loyalty are paramount for these firms and they determine the relationship amongst firms and between firms and their customers, suppliers, and competitors, creating a regional culture that is less open and interactive. Corporate hierarchies within firms ensure that authority remains centralized, and information tends to flow vertically. In this self-reliant and inward-looking system, social and technical networks are largely internal to the firm, and the boundaries amongst firms and between firms and local institutions remain far more distinct.
12.5.1 Regional Networks and Shifting Industrial Systems From the discussion above, two distinct types of regional economies are identified: a network-based system and an autarkic firm-based system. Similarly, two models of industrial systems are distinguished: a regional network-based system and an independent (self-reliant) firm-based system. Understanding regional economies from these perspectives helps to illuminate the divergent trajectories of the Silicon Valley and Route 128 economies during the 1980s. For example, it would be erroneous to attribute Silicon Valley’s superior performance to differentials in real-estate costs, wages, or tax levels. In fact, during the 1980s, land and office space were significantly more costly in Silicon Valley than in the Route 128 region.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
218 AnnaLee Saxenian Likewise, the wages and salaries of production workers, engineers, and managers in Silicon Valley were higher (Sherwood-Call 1992), although there was no significant difference in tax rates between California and Massachusetts (Tannenwald 1987). One possible explanation for the differences in regional performance could be the pattern of defence spending and the spillover effects on firms in the region. Although import ant, it seems, however, that the regional differences cannot be attributed to defence-related spending. Historically, Route 128 has relied more heavily on military spending than Silicon Valley and, therefore, in principle, is more vulnerable to defence cutbacks. However, the downturn in the Massachusetts electronics industry began in 1984, which means that defence spending cannot account for the timing of the downturn in the region’s technology industry, although no doubt military spending cutbacks that began in the late 1980s may have contributed to the difficulties of an already troubled regional economy. Finally, it is sometimes tempting to attribute the prosperity of a region to the ability of local firms to shift low-wage jobs elsewhere. However, the relocation of low-wage jobs elsewhere alone cannot account for the differential performance of the two regions. Technology firms from both Silicon Valley and Route 128 have, since the 1960s, moved their routine manufacturing operations to lower-wage regions of the United States and the Third World (Scott 1988b; Saxenian 1985). A plausible explanation for Route 128’s difficulties lies in the rigidities of the local industrial system. The independent, firm-based system flourished in an environment of market stability and slowly changing technologies, because extensive integration offered the advantages of scale economies and market control (Chandler 1977). Route 128 has been overwhelmed, however, by changing competitive conditions. Corporations that invest in dedicated equipment and specialized worker skills find themselves locked into obsolete technologies and markets, and their self-sufficient structures limit their ability to adapt in a timely fashion. The surrounding regional economy, in turn, is deprived of resources for self-regeneration, because large firms tend to internalize most local supplies of skill and technology. In contrast, regional, network-based industrial systems such as Silicon Valley’s are well suited to conditions of technical and market uncertainty. Producers in these systems deepen their capabilities by specializing while engaging in close—but not exclusive—relations with other specialists. Network relations promote a process of technology diffusion and reciprocal innovation that reduce the distinctions between large and small firms and between industries and sectors (DeBresson and Walker 1991). Evidence from the industrial districts of Europe suggests that the localization of know-how and information encourages the pursuit of diverse technical and market opportunities through spontaneous regroupings of skill, technology, and capital. In such circumstances, the region, if not all of the firms in the region, is organized to innovate continuously, which is a precondition for firm-level growth and competitiveness (Best 1990, 2018; Sabel 1988). The competitive advantages of network organizational forms are reflected in the experience of Japanese industry as well. Japanese producers of electronics and automobiles, for example, rely on extensive networks of small and medium-sized suppliers, to which they are linked through ties of trust and partial ownership.8 Although Japan’s large firms may have exploited subcontractors in the past, many of these firms are increasingly collaborating with suppliers, encouraging them to expand their technological capabilities and organizational 8 See Piore and Sabel (1984), Fruin (1992), Imai (1989).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Cluster Dynamics and Regional Networks 219 autonomy (Nishiguchi 1989). Like their Silicon Valley counterparts, these firms tend to be geographically clustered and to depend heavily on informal information exchange as well as more formal forms of cooperation (Friedman 1988; Imai 1989). Of course, all economic activity does not cluster within a single regional economy. Firms in network systems serve global markets and collaborate extensively with distant customers, suppliers, and competitors. Producers of new electronics and computing technologies, in particular, are highly international. The key lesson from the successful regional econ omies is that most strategic relationships are often local, because timeliness and face-to-face communications are very important in complex, uncertain, and fast-changing industries (Nohria and Eccles 1992a).
12.6 Synergy Between Regional Networks and Firm Production Systems The differing organizational and adaptive capacities of Silicon Valley’s regional network and Route 128’s independent, firm-based industrial systems become evident through a comparative analysis of selected firms. A comparison of Apollo Computers and Sun Microsystems, both established in the 1980s to enter the emerging workstation market, shows how small firms benefit from the open flow of information, technology, and expertise in a network system. A comparison of the two leading computing systems producers in the two regions—DEC and HP—illustrates how regional networks can facilitate the reorganization of large firms. While these cases cannot represent the complexity of the two regional economies, or portray the myriad decentralized relationships in a regional, network-based system, they do illustrate the social and institutional dimensions of productive organization often disregarded by the concept of external economies.
12.6.1 Start-ups and Spin-out Firms In the late 1970s, Silicon Valley experienced its largest surge of start-ups, which continued to accelerate during the 1980s. The start-ups specialized in a range of computer products, resulting in the diversification of the regional economy beyond its original concentration of semi-conductors. In contrast, the emergence of start-ups slowed down along Route 128 in the 1980s, and the rates of high-technology firm formation were lower than the rest of the country (Kirchoff and McAuliffe 1988). The remarkable performance of Silicon Valley companies such as Sun Microsystems in the 1980s could not be matched by those along Route 128 (see Appendix: Figures A12.1 and A12.2). By the end of the 1980s, public companies that had started in Silicon Valley during that decade collectively accounted for more than US$22 billion in sales, whereas their Route 128 counterparts had generated only US$2 billion (Standard & Poor’s 1992). By 1992, 113 technology enterprises located in Silicon Valley reported revenues exceeding US$100 million, compared with seventy-four companies along Route 128. Investment flows into northern California were also significant in driving Silicon Valley’s growth. Annual venture capital investments
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
220 AnnaLee Saxenian in northern California during the 1980s were double or triple those in Massachusetts, and Silicon Valley companies were consistently awarded at least one-third of the nation’s total venture capital pool during the 1980s and early 1990s (Venture Capital Journal, 1980–92). A comparison of Apollo Computer and Sun Microsystems shows how the autarkic structures and practices of Route 128’s independent, firm-based system created disadvantages for start-ups in a technologically fast-paced industry. Apollo was initially highly successful when it pioneered the engineering workstation in 1980, a product considered far superior to that of its Silicon Valley counterpart, Sun Microsystems (started two years after Apollo, in 1982). While the two firms competed closely during the mid-1980s, Apollo began to fall behind and never recovered its lead. By the time Apollo was sold to HP in 1989, it was fourth in the industry and Sun was at the top (Bell and Corliss 1989). Apollo’s initial strategy and structure reflected the model of corporate self-sufficiency of the Route 128 region’s large minicomputer companies. The firm adopted proprietary standards and chose to design and fabricate its own central processor and specialized integrated circuits, making its products incompatible with other machines. Sun, in contrast, pioneered open systems, making the specifications for its systems widely available to suppliers and competitors. This challenged the proprietary and highly profitable approach of industry leaders, such as IBM, DEC, and HP, which restricted customers to a single vendor of hardware and software. Sun also purchased virtually all of its components from external vendors and subcontracted the manufacture and assembly of its printed circuit boards. This strategy allowed it to specialize—to focus on designing the hardware and software for its workstations and to limit manufacturing to prototypes, final assembly, and testing. Relying on outside suppliers greatly reduced Sun’s overheads and ensured that the firm’s workstations contained state-of-the-art hardware. This focus on purchasing equipment also allowed Sun to introduce complex new products rapidly and to alter its product mix continuously, keeping it well ahead of competitors and clone-makers. Sun’s innovative computing strategy was inseparable from the firm’s location in the sophisticated and diversified technical infrastructure of Silicon Valley. Apollo, in contrast, responded sluggishly to industry changes, in part because of a more limited regional infrastructure. Although it became apparent that customers preferred the cheaper, non-proprietary Sun workstations, Apollo, like the Route 128 minicomputer producers, was slow to abandon its proprietary operating system and hardware. The commitment to formality, hierarchy, and long-term stability at Apollo—which typified most large Route 128 companies—could not have offered a greater contrast to the ‘controlled chaos’ that characterized Sun (Weiss and Delbecq 1987). Like many Silicon Valley companies, Sun developed decentralized organizational forms as a way to preserve the flexibility and enthusiasm of a start-up company even as it grew. Corporate strategy was generated by discussions amongst representatives of autonomous divisions rather than being dictated by a central committee, and Sun’s culture encouraged informal communications, participation, and individual initiative (Levine 1988). In the late 1980s, when Sun surpassed Apollo in both sales and profitability, more than a dozen Apollo managers defected to their West Coast rival. There they joined other experienced and ambitious engineers from ailing Route 128 companies who recognized that opportunities to join or start technologically exciting new ventures lay not in New England but along the increasingly crowded freeways of northern California. As skilled engineers moved west, the advantages of Silicon Valley’s network-based industrial system multiplied.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Cluster Dynamics and Regional Networks 221
12.6.2 Big Businesses and Cluster Dynamics The wave of successful start-ups of the 1980s was a driving force of Silicon Valley’s faster adaptation to changes in the computer industry. Large firms were also critical in this transition. Decentralization of operations of established firms in Silicon Valley built and expanded inter-firm production networks based on the region’s social and technical interdependencies, strengthening its industrial system. Long-standing practices of informal cooperation and exchange were institutionalized, formalizing the process of collective learning in the region. Local firms repositioned themselves within local production networks, and new markets and sectors emerged within the regional cluster. By contrast, the Route 128 regional cluster adapted more slowly, constrained by the aut arkic organization and practices of its leading producers. The region’s large minicomputer firms adjusted very slowly to the new market dynamics given their inward focus and lack of dynamic start-ups from which to draw technological and organizational innovations. By the end of the 1980s, they were struggling to survive in a computer industry that they had once dominated (see Appendix: Figures A12.1 and A12.2). A look at two large firms, DEC and HP, along Route 128 and Silicon Valley respectively, illustrates how the relationship of large firms to their region differed in the network and firm-based industrial systems. Both DEC and HP were vertically integrated producers of proprietary minicomputers with shared origins in an earlier era of computing. Yet the companies responded differently to comparable competitive challenges. HP gradually opened up by building a network of local alliances and subcontracting relationships while strengthening its global reach. DEC, in spite of its formal commitment to decentralization, retained a substantially more autarkic organizational structure and corporate mindset. The transformations in the computer industry during the 1980s forced producers to develop new products and bring them to market faster than ever before; the cost of developing new products increased, as they became more technologically complex and innovations in all segments of the industry made it difficult for a single firm to produce all of the components. This increasingly competitive environment posed a challenge for established computer makers such as DEC and HP and their vertically integrated production systems. Variations in corporate performance always have multiple causes, but the firms’ organiza tional structures and their relationships to their respective regions help explain these differences. DEC maintained clear boundaries between itself and other companies or institutions in the region. This was, in part, a result of extensive vertical integration. Moreover, DEC’s corporate culture rewarded secrecy and corporate loyalty. Thus, departed employees typically were treated like pariahs and cut off from the corporate ‘family’ (Rifkin and Harrar 1990). As a result, the technical and social networks that mattered were all internal, and there were few opportunities for collaboration, learning, and exchange with other local firms. HP was both less dominant in Silicon Valley and more open to the surrounding economy. DEC dominated the Route 128 region. With more than 30,000 Massachusetts employees in 1990, DEC accounted for almost 20 per cent of regional high-technology employment, whereas HP’s 20,000 Silicon Valley employees were only 8 per cent of the regional total. HP benefited from a long history of participation in the region’s rich associational life and fluid labour markets. Continuous and open exchange about everything from the latest start-up companies to technical breakthroughs allowed local engineers to stay at the leading edge of new computing technologies and market trends (Vedoe 1990).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
222 AnnaLee Saxenian HP’s decentralized divisional structure also offered an ideal training ground for general managers. Former HP executives were responsible for starting more than eighteen firms in Silicon Valley between 1974 and 1984, including such notable successes as Rolm, Tandem, and Pyramid Technology (Mitchell 1989). In contrast, DEC’s matrix organization—which represented only a partial break from traditional functional corporate hierarchies—stifled the development of managerial skill and initiative in the Route 128 region. The matrix demanded continuous negotiations to reach consensus, and despite the addition of cross-functional relations amongst product groups, final authority remained highly centralized (Schein 1985). As a result, it is difficult to identify any successful spin-offs from DEC other than Data General. HP also built alliances with local companies that offered complementary technologies. In 1987 HP allowed Weitek, the leading designer of ultra-high-speed ‘number-crunching’ chips that were used in complex engineering problems, to use its state-of-the-art fabrication facility as a foundry, hoping to improve the performance of the Weitek chips in its workstations. The arrangement assured HP a steady supply of Weitek’s chips and allowed HP to introduce its new workstation faster than if it had designed the chip in-house. It provided Weitek with a market, the legitimacy of a close association with HP, and access to a state-of-the-art foundry.9 Moreover, the final product represented a significant advance over that which either firm could have produced independently. This partnership allowed each company to draw on the other’s distinctive and complementary expertise to devise a novel solution to a shared problem. HP opened itself to other outside influences during the 1980s, creating a model of a large firm that is internally decentralized and horizontally linked to networks of other specialists. DEC’s dominant and isolated position along Route 128, by contrast, hindered its efforts to shift to new technologies or a new corporate form. Saddled with an autarkic organizational structure and located in a region that offered little social or technical support for a more flexible business model, DEC’s difficulties worsened. Ultimately, by 1990, HP had successfully managed the transition from minicomputers to workstations with open systems, whereas DEC remained dependent on its proprietary VAX line of minicomputers. As a result, even though both enjoyed 1990 revenues from electronics products of US$13 billion, HP earned US$771 million and DEC lost US$95 million.
12.7 Conclusions This chapter has reviewed the importance of regional advantages and cluster dynamics, and how the network approach can contribute to a better understanding of regional hubs and the competitive advantage of firms. Industrial hubs significantly shape regional and national economies, and promote innovation and increasing returns to scale. The recent phenom enon shows that new emerging economies in the global South can benefit by linking with leading technology hubs to circulate talents and bring brain gain. This focus of the chapter is a specific and important period in the transformation of the two clusters (Silicon Valley and Route 128) between the 1970s and 1990s, and the circulation of talent in the 1980s and 1990s. This comparison of Silicon Valley and Route 128 industries highlights the analytical leverage gained by treating regions as networks of relationships rather than as collections of 9 Yoder (1991).
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Cluster Dynamics and Regional Networks 223 atomistic firms. By transcending the theoretical distinction between what lies inside and outside the firm, this approach offers important insights into the structure and dynamics of regional economies. It directs attention to the complex networks of social relationships within and between firms and between firms and local institutions. The Silicon Valley experience also suggests that the network form of organization flourishes in regional agglomerations. Proximity facilitates the repeated, face-to-face interaction that fosters the mix of competition and collaboration required in today’s fast-paced technology industries. Yet the case of Route 128 demonstrates that geographic clustering alone does not ensure the emergence of regional networks. Competitive advantage derives as much from the way that skill and technology are organized as from their presence in a regional environment.
appendix
Employment (in 10 thousands)
30 25
21
20
17
15
12
10 5 0
27
26.5
9
7
6
15.5
13.5 10
6
2.5
2 1959
1965
1970
1975
Silicon Valley
1980
1985
1990
Route 128
Figure A12.1 Total high technology employment, Silicon Valley and Route 128, 1959–90
45
39
40 33
Companies
35 30 25 20 15
25
22 14
26
25 15
14
10
10
7
5 0
1985
1986
1987 Silicon Valley
1988
1989
4 1990
Route 128
Figure A12.2 Fastest-growing electronics firms, Silicon Valley and Route 128, 1985–90
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
224 AnnaLee Saxenian
References Arthur, Brian (1990) ‘Positive Feedbacks in the Economy’, Scientific American 262(2): 92–9. Basche, Todd (1991) Vice President, Sparcstation Group, Sun Microsystems, personal communication. Beam, A. and M. Frons (1985) ‘How Tom Vanderslice is Forcing Apollo Computer to Grow Up’, Business Week, 25 March, pp. 96–8. Bell, A. and E. Corliss (1989) ‘Apollo Falls to the West’, Mass High Tech 1: 9. Best, Michael (1990) The New Competition: Institutions of Industrial Restructuring. Cambridge, MA: Harvard University Press. Best, Michael (2018) How Growth Really Happens: The Making of Economic Miracles through Production, Governance, and Skills. Princeton, NJ: Princeton University Press. Bulkeley, William (1987) ‘Culture Shock: Two Computer Firms with Clashing Styles Fight for Market Niche’, Wall Street Journal, 6 July. Castells, Manuel (1989) The Informational City: Information Technology, Economic Restructuring, and Urban-Regional Process. Oxford: Basil Blackwell. Chandler, Alfred (1977) The Visible Hand: The Managerial Revolution in American Business. Cambridge, MA: Belknap. Comerford, Richard (1992) ‘How DEC Developed Alpha’, IEEE Spectrum 29(7): 26–31. CorpTech (1993) Technology Company Information: Regional Disks. Woburn, MA: Corporate Technology Information Services. DeBresson, C. and R. Walker (eds) (1991) ‘Special Issue on Networks of Innovators’, Research Policy 20: 5. DeNucci, Joe (1991) Vice President, Entry Systems Group, MIPS Computer Systems, personal communication. Edmondson, Harold (1988) Vice President of Corporate Manufacturing, Hewlett Packard Corporation, personal communication. Electronic Business, ‘The Top 100 Exporters’, pp. 4–42. March 16, 1992. Florida, Richard and Martin Kenney (1990) ‘Silicon Valley and Route 128 Won’t Save Us’, California Management Review 33(1): 68–88. Friedman, David (1988) The Misunderstood Miracle: Industrial Development and Political Change in Japan. Ithaca, NY: Cornell University Press. Fruin, Mark (1992) The Japanese Enterprise System. Oxford: Oxford University Press. Furlong, Tom (1991) RISC Workstation Manager, DEC Palo Alto, personal communication. Granovetter, Mark (1985) ‘Economic Action and Social Structure: The Problem of Embeddedness’, American Journal of Sociology 91(3): 481–510. Greene, T. (1990) ‘Can HP Find the Right Direction for the ’90s?’, Electronic Business, pp. 26–9, 22 January. Hall, Peter and Ann Markusen (1985) Silicon Landscapes. Boston, MA: Allen and Unwin. Herrigel, Gary (1993) ‘Large Firms, Small Firms, and the Governance of Flexible Specialization: The Case of Baden-Wurtemmberg and Socialized Risk’, in Bruce Kogut (ed.) Country Competitiveness. New York: Oxford University Press, pp. 15–35. Imai, Ken-ichi (1989) ‘Evolution of Japan’s Corporate and Industrial Networks’, in Bo Carlsson (ed.) Industrial Dynamics. Boston, MA: Kluwer, pp. 123–56. Kalb, Jeff (1991) President, MasPar Computer Corporation, personal communication. Kirchoff, B. and R. McAuliffe (1988) Economic Redevelopment of Mature Industrial Areas. Report prepared for Technical Assistance and Research Division, Economic Development Administration, US Department of Commerce. Krugman, Paul (1991) Geography and Trade. Cambridge, MA: MIT Press. Levine, J. B. (1988) ‘Sun Microsystems Turns on the Afterburners’, Business Week, pp. 114–18, 18 July. Marshall, Alfred (1920) Industry and Trade. London: Macmillan.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
Cluster Dynamics and Regional Networks 225 McWilliams, G. (1992) ‘Crunch Time at DEC’, Business Week, pp. 30–3. Mitchell, J. (1989) ‘HP Sets the Tone for Business in the Valley’, San Jose Mercury News, pp. 1D–2D, 9 January. Nash J. and M. Hayes (1993) ‘Key DEC Project Moving to Palo Alto’, Business Journal (San Jose and Silicon Valley), 19 July, pp. 1, 17. Nee, E. (1991) ‘Back to Basics at HP’, Upside, June/July, pp. 38–78. Nishiguchi, T. (1989) ‘Strategic Dualism: An Alternative in Industrial Societies’. PhD dissertation, Nuffield College, Oxford University. Nohria, Nitin and Robert G. Eccles (1992a) ‘Face-to-Face: Making Network Organizations Work’, in Nitin Nohria and Robert G. Eccles (eds) Networks and Organizations: Structure, Form, and Action. Boston, MA: Harvard Business School Press. Nohria, Nitin and Robert G. Eccles (1992b) Networks and Organizations: Structure, Form, and Action. Boston, MA: Harvard Business School Press. Piore, Michael and Charles Sabel (1984) The Second Industrial Divide: Possibilities for Prosperity. New York: Basic Books. Porter, Michael (1990) The Competitive Advantage of Nations. New York: The Free Press. Porter, P. (1993) ‘Executive Interview: HP’s Gary Eichorn Tackles Enterprise Computing’, Mass High Tech, 23 August, p. 3. Powell, W. (1987) ‘Neither Market nor Hierarchy: Network Forms of Organization’, in B. Staw (ed.) Research in Organizational Behavior. Greenwich, CT: JAI Press. Rifkin, Glenn and George Harrar (1990) The Ultimate Entrepreneur: The Story of Ken Olsen and Digital Equipment Corporation. Rocklin, CA: Prima Publishing. Sabel, Charles (1988) ‘Flexible Specialization and the Reemergence of Regional Economies’, in Paul Q. Hirst and Jonathan Zeitlin (eds) Reversing Industrial Decline? Industrial Structure and Policy in Britain and her Competitors. Oxford: Berg. Saxenian, AnnaLee (1985) ‘Silicon Valley and Route 128: Regional Prototypes or Historic Exceptions?’, in M. Castells (ed.) Reversing Industrial Decline? Industrial Structure and Policy in Britain and her Competitors. Beverly Hills: Sage. Saxenian, AnnaLee (1994) Regional Advantage: Culture and Competition in Silicon Valley and Route 128. Cambridge, MA: Harvard University Press. Saxenian, AnnaLee and Jinn-Yuh Hsu (2005) ‘The Silicon Valley-Hsinchu Connection: Technical Communities and Industrial Upgrading’, in Stefano Breschi and Franco Malerba (eds) Clusters, Networks, and Innovation. Oxford: Oxford University Press, pp. 235–60. Schein, Edgar (1985) Organizational Culture and Leadership. San Francisco, CA: Jossey-Bass. Scott, Allen (1988a) Metropolis: From the Division of Labor to Urban Form. Berkeley, CA: University of California Press. Scott, Allen (1988b) New Industrial Spaces: Flexible Production Organization and Regional Development in North America and Western Europe. London: Pion. Sheff, D. (1989) ‘A New Ballgame for Sun’s Scott McNealy’, Upside, November/December, pp. 46–54. Sherwood-Call, C. (1992) ‘Changing Geographic Patterns of Electronic Components Activity’, Economic Review (Federal Reserve Board of San Francisco), pp. 2, 25–35. Standard & Poor’s (1992) Compustat PC+ database. Storper, Michael (1989) ‘The Transition to Flexible Specialization in the US Film Industry: External Economies, the Division of Labor, and the Crossing of Industrial Divides’, Cambridge Journal of Economics 13: 273–305. Tannenwald, R. (1987) ‘Rating Massachusetts’ Tax Competitiveness’, New England Economic Review, November/December pp. 33–45. Tuller, D. (1988) ‘HP Plans to Buy 10 per cent Stake in Octel’, San Francisco Chronicle, 12 August, p. B1. Vedoe, C. (1990) Manager, Workstation Marketing, Sun Microsystems, personal communication. Venture Capital Journal, 1980–92. Selected issues. Wellesley, MA: Venture Economics. Vernon, Raymond (1960) Metropolis 1985. Cambridge, MA: Harvard University Press.
OUP CORRECTED PROOF – FINAL, 22/06/20, SPi
226 AnnaLee Saxenian Weiss, Joseph and Andre Delbecq (1987) ‘High Technology Cultures and Management: Silicon Valley and Route 128’, Group and Organization Studies 12(1): 39–54. Whiting, C. (1987) ‘For Flexible, Quality Manufacturing Don’t Do It Yourself ’, Electronic Business, 15 March, pp. 46–7. Yoder, S. (1991) ‘A 1990 Reorganization at Hewlett-Packard Already is Paying Off ’, Wall Street Journal, 22 July, p. A1.
OUP CORRECTED PROOF – FINAL, 24/06/20, SPi
chapter 13
Ag gl om er ation of Eu ropea n I n dustr ie s Michael A. Landesmann and Joris M. Schröder
13.1 Introduction This chapter analyses the remarkable dynamic of relocation of industrial activity across European regions and countries that happened over the past few decades. We shall focus specifically on the developments in and across three groups of EU member countries: the Western and Northern EU member states plus Norway (EU–NW+NO), the Southern EU member states (EU–SE, comprising Portugal, Spain, Italy, and Greece) and the new EU member states in Central and Eastern Europe (EU–CEE11).1 The emphasis on relocation of industrial activity in Europe’s economic geography is somewhat different from the main focus of the existing literature on agglomeration. This literature concentrates in particular on the tensions between ‘efficiency gains’ from agglomeration vs. ‘equity costs’ (see, for example, Petrakos et al. 2005; Crozet and Koenig 2007; Martin 2008; Gardiner et al. 2010; Rodriguez-Pose 2018; Storper 2018; Iammarino et al. 2019). We shall introduce in section 13.2 shortly the direction of our analysis against the background of the available literature on agglomeration in Europe. In this chapter the focus on agglomeration of European industry is motivated by an issue that is of great macroeconomic significance, that of persistent external imbalances amongst member countries of the European Union. This can be considered to be one of the core issues that destabilizes the construction of the European Monetary Union and hence the European integration process as a whole. We shall argue in section 13.3 why we think that this is a core issue for the survival of the European integration process and why agglomeration of industrial activity is a crucial aspect of external imbalances in the European Union. Section 13.4 will analyse industrial agglomeration phenomena at the regional level in the European Union. It will point to tendencies of deindustrialization, but also of reindustrialization that can be observed in subsets of low-, medium-, and high-income regions of the 1 At times, we shall also include in our analysis countries, such as those of the Western Balkans, that are not yet members of the European Union.
OUP CORRECTED PROOF – FINAL, 24/06/20, SPi
228 Michael A. Landesmann and Joris M. SchrÖder European Union. Section 13.5 will refer particularly to the role that is played by regional production networks (RPNs) that are at the centre of the relocation processes of industrial activity that we observe in the European economy. Section 13.6 concludes.
13.2 Agglomeration of European Industries: The Focus of this Chapter The focus of this chapter deviates somewhat from the voluminous literature on agglomeration phenomena which followed the lines of enquiry originally explored by Alfred Marshall in the 1890s and 1910s (see Marshall 1898, 1919) and which emphasized the ‘efficiency gains’ that can be reaped from locational agglomerations of industrial activity. Marshall enumerated a number of external economies (external, that is, to the individual firm) as explanation for such efficiency gains: in summary, Marshall emphasized the advantages of a pooled market for workers with differently specialized skills, the availability of a wide array of specialized input suppliers, and knowledge or technological spillovers across firms. On top of that there is the feedback on the demand side, in that a range of increasingly specialized suppliers also attracts demand that benefits from this variety and low search and travel costs. A large literature developed from Marshall’s early insights: in the context of modern labour market analysis, some authors emphasized efficiency gains derived from reduced search costs and easier job-matching in pooled labour markets (Simpson 1992; Wheeler 2001); others concentrated on the faster learning processes lifting firm productivity and the development of human capital (Boschma 2005; Duranton and Puga 2001); increased information flows were added to reduced transport and logistics costs (Gordon and McCann 2013; Egeraat and Jacobson 2006); and the increasing importance of informational and knowledge externalities which result from the concentration of related firms was emphasized (Storper 1995; Malmberg et al. 2000; Malmberg and Maskell 2002). Further, the public interest (and lobbying power) of agglomerated business activity also leads to better provision of public infrastructure and facilities (Martin 1999; ESPON 2004; Fujita et al. 1999; Fujita and Thisse 2002; Crescenzi and Rodriguez-Pose 2012). The other side are negative externalities, such as congestion, rising rental rates etc. However, the empirical studies (see, for example, Ciccone and Hall 1996; Ciccone 2002; Baldwin and Martin 2004) find that the negative externalities are outweighed by the positive efficiency impact of agglomeration. Thus, a number of studies estimate the impact of agglomeration on productivity and on innovation activity, either using firm-level data or at the regional level (Ellison and Glaeser 1997; Rosenthal and Strange 2004; Moreno et al. 2005; Rice et al. 2006; Oerlemans and Meeus 2005; Artis et al. 2009; Brülhart and Mathys 2008; Brülhart and Sbergami 2009; ESPON 2013). Our focus in this chapter is a more specific one, i.e. analysing the recent (last two decades) developments with regard to changes in the main locations of industrial activity in the European Union (and some associated countries). We consider industrial capacity as particularly important to solve the issue of external imbalances, which many low- and medium-income countries (LMICs) face as they attempt to catch up in income levels with more advanced countries and regions. Hence our focus is on that group of countries and
OUP CORRECTED PROOF – FINAL, 24/06/20, SPi
Agglomeration of European Industries 229 the regions therein. However, in order to obtain an overview of the relative strengths and weaknesses of different European countries and regions we shall also present information regarding the geographic agglomeration and specialization in another group of industries, namely that of ‘advanced tradable services’ (ATS) activities.2 Our analysis will show the patterns of specialization of low-, medium-, and high-income regions in Europe with regard to these two groups of tradables, i.e. manufacturing and the group of ‘advanced tradable services’.3
13.3 External Imbalances in the European Union and the Role of Industrial Activity As mentioned in the introduction, this chapter emphasizes the link between industrial agglomeration and issues of external imbalances in the European economy (see also Landesmann 2015; Landesmann and Hanzl-Weiss 2016). The argument goes as follows: low- and medium-income countries (LMICs) are more dependent on manufacturing activity as the principal tradable sector which can avoid persistent trade account deficits, while advanced economies have the further option of specialization on the relatively fast developing areas of ‘advanced tradable services’ (ATS). This has much to do with the importance of ‘tacit’, and ‘non-codifiable’ knowledge in ATS activities which is less easily transferable to LMICs than are process and product technologies in manufacturing (see Bekkers et al. 2017, where this hypothesis has been tested). Hence, for LMICs it is particularly important to build up sufficient export capacities in manufacturing over time, if they are to avoid persistent and—over time—unsustainable deficits to develop in the current account. Strong agglomeration of industrial activity in the European economy might therefore lead to the build-up of unsustainable current account positions of significant groups of LMICs. Figure 13.1 shows the developments of the current accounts and its components in a number of European economies from 2002 onwards (thus covering the pre-crisis and crisis periods). The figure limits itself to the group of lower- and middle-income countries (LMICs). What it shows is that over the pre-crisis period many of the LMICs experienced very severe current account deteriorations that had been mostly driven by the trade accounts, i.e. the value of imports exceeding the value of exports. This was true for the Southern European economies (with the exception of Italy), the Baltic states, Romania and Bulgaria, and the Western Balkan countries (WB-6). There is an exceptional 2 The ‘advanced tradable services’ (ATS) group of industries includes NACE industry categories J and K to N without L (real estate activities), i.e. J: information and communication; K: financial and insurance activities; M and N: professional, scientific, technical, administration, and support service activities. 3 We shall leave out other (also tradable) services activities, such as transport services and travel/tourism, as these are less strongly associated with specialization options linked to levels of economic development than are the ATS group of service activities. The link between specialization on ATS activities and levels of economic development will emerge clearly from our analysis.
OUP CORRECTED PROOF – FINAL, 24/06/20, SPi
230 Michael A. Landesmann and Joris M. SchrÖder 20.0 10.0 0.0 –10.0 –20.0 –30.0 –40.0
CZ
HU
PL
SK
SI
20.0 10.0 0.0 –10.0 –20.0 –30.0 –40.0
BG
EE
LT
LV
RO
ES
GR
IE
IT
PT
AL
BA
HR
ME
MK
20.0 10.0 0.0 –10.0 –20.0 –30.0 –40.0
20.0 10.0 0.0 –10.0 –20.0 –30.0 –40.0
Goods & services (Trade)
Secondary income
Primary income
Current account
Figure 13.1 Composition of the current account of the balance of payments, 2002–18, in per cent of GDP Source: wiiw Database incorporating national and Eurostat statistics.
group, namely the group of Central European economies (Czech Republic, Hungary, Poland, Slovakia, and Slovenia; EU–CEE5), most of which achieved balanced trade before the start of the financial crisis in 2008 and some even a surplus. This is in stark contrast to the other LMICs that experienced deficits in the trade accounts of between 10 and 25 per cent of GDP per annum.
OUP CORRECTED PROOF – FINAL, 24/06/20, SPi
Agglomeration of European Industries 231 goods
Service
trade balance: goods and services
100 80 60 40 20 0 –20 –40
CEE5
BG, RO
Baltics
ES, GR, PT
WB-6
UA
Figure 13.2 External trade, 2003–17, in per cent of GDP Note: WBC-6 refers to Western Balkan countries. Source: wiiw Annual Database incorporating national and Eurostat statistics.
Persistent deficits in the current accounts lead to the build-up of external debt positions. With the outbreak of the financial crisis financial markets were no longer prepared to finance further deficits in the current accounts in those economies where the path of external debt build-up was no longer seen as sustainable. The consequence was that the e conomies were cut off from further external finance (a ‘sudden stop’), and they had to retrench dramatically their spending on imports. Exports could not be expanded in the short term given that devaluation of the exchange rate was not an option as a member of the Monetary Union or—as is the case for most of the WB-6 countries—being closely linked to the Union through pegged or quasi-fixed exchange rate regimes. Hence imports had to severely contract which happened (in countries that had persistently high current account deficits in the pre-crisis period) by moving into deep (and, in some countries prolonged) recessions. If we look more closely at the reasons for such persistent trade imbalances, what emerges is that many of the LMICs had and still have a relatively small export sector (see Figure 13.2). The average shares of exports in GDP in the case of some very small and open economies did not exceed 20–35 per cent (Western Balkans, Greece); for others it remained in the 35–45 per cent range, which is still very small. The exception is again the group of Central European economies (EU–CEE5) that show export-to-GDP ratios of 80–90 per cent which is a sign of a healthy tradable sector. Finally, we want to refer the reader to Map 13.1 which shows (at the country level) the contributions of the different European economies to—what the new literature on inter national value chains (see, for example, Baldwin 2013; Baldwin and Lopez-Gonzalez 2015)— calls value-added exports (VAX). The calculation of such VAX utilizes the World Input–Output Database (see Timmer et al. 2015) in order to add all value-added contributions which a country makes to exporting activity in Europe, both through its own exports directly, but also—directly and indirectly—contributing through the supply of intermediate inputs to the exports of other European economies. On the right-hand side of this map are the contributions of the different European economies to manufacturing exports and on the left-hand side to total exports. What we see clearly in this map, is that there is a ‘Central
OUP CORRECTED PROOF – FINAL, 24/06/20, SPi
232 Michael A. Landesmann and Joris M. SchrÖder
Value-added export shares divded by GDP shares 2014 Total economy
Category –0.91 – –0.57 –0.57 – –0.20 –0.20 – –0.01 –0.01 – 0.21 –0.21 – 0.62 –0.62 – 1.40
Value-added export shares divded by GDP shares 2014 Manufacturing
Category –0.91 – –0.57 –0.57 – –0.20 –0.20 – –0.01 –0.01 – 0.21 –0.21 – 0.62 –0.62 – 1.40
Map 13.1 Countries’ shares in ‘value-added exports’ (VAX)—in manufacturing (rhs) and in total European exports Note: Value-added exports (VAX) have been calculated from the World Input–Output Database (WIOD) that analyses contributions to European exports in value-added terms taking into account cross-border trade linkages.
European manufacturing core’ (see also Stehrer and Stöllinger 2015) encompassing Germany, Switzerland, Austria, and neighbouring Central–Eastern European economies (Czech Republic, Hungary, Slovakia, Slovenia) that accounts for contributions to European manufacturing exports which are more than 50 per cent above what their shares in European GDP would indicate. This high regional concentration of manufacturing export capacities in Central Europe indicates the importance of cross-border production linkages in current European agglomerations of industrial activity. We shall come back to this in section 13.5. The additional feature we want to emphasize with this map is that countries—with some exceptions—that are doing badly in contributing (directly or indirectly) to European manufacturing exports (right-hand map) are also doing badly in contributing to overall exporting activity (left-hand map). Countries that fall into this category are the Southern European economies (Greece, Portugal, Spain, Italy), but also France and Ukraine. The exceptions we referred to earlier are the United Kingdom and Norway in that they contribute less than 50 per cent (directly and indirectly) to manufacturing exports in Europe compared to what the size of their economies would indicate. However, they compensate through above-average contributions to other European exports (Norway through its oil and gas exports, and the United Kingdom—only partially—through its strong contributions in financial and business services exports) and hence their overall contributions of European exports are significantly higher than that to manufacturing. We shall now move to discussing agglomeration phenomena and their dynamic at the regional (sub-national) level.
OUP CORRECTED PROOF – FINAL, 24/06/20, SPi
Agglomeration of European Industries 233
13.4 The Agglomeration of Industrial Activity in the ‘Central European Manufacturing Core’ In the following, a few maps and figures will illustrate the agglomeration of industrial activity across European regions. Map 13.2 shows the shares of manufacturing activity in total employment at the regional level; we present a three-year average 2014–16.4 From Map 13.2 we see the relatively strong position of manufacturing in Germany (particularly Southern Germany), parts of Austria, as well as in the regions of adjoining Central European economies (the Czech Republic, Slovakia, and regions in Poland, Hungary, and Slovenia). This industrial belt also includes by now the Western regions of Romania and then makes the circle back to Northern Italy. Switzerland, for which we do not have the regional statistics in the Eurostat dataset, would also be included in this industrial agglomeration. Further to the North we can see the importance of manufacturing in Southern Sweden and also Southern Finland. Let us now examine the dynamic that lies behind the agglomeration of industrial activity with some figures (Figure 13.3 and later Figure 13.5) that examine ‘convergence’ or ‘divergence’ of regions in Europe, i.e. we check whether regions that in the initial period (2000/2) start with a high/low share of manufacturing, experience an increase or a decrease in that share over the period 2000/2–14/16. Regions that start with a high share and experience a fall in that share are ‘deindustrializing’, while those regions starting with a low share and increase the share of manufacturing over the period could be seen as ‘industrializing’ (sometimes ‘re-industrializing’ if they experienced an earlier collapse of industry). In Figure 13.3 we show such regional convergence and divergence patterns for the three groups of countries. We identify these in different shades of grey (respectively icons): the group of EU Northern and Western European countries (EU–NW+NO, i.e. including Norway), the group of Southern EU economies (EU–SE that includes Greece, Italy, Portugal, and Spain) and the group of Central and Eastern European EU member countries (EU–CEE11). On the horizontal axis we measure the shares of manufacturing (this time we chose shares in gross value added rather than in employment) in the initial period (2000/2) and on the horizontal axis the change in the shares over the period 2000/2–14/16. We also plot a ‘regression line’ (Lift) through the regional data points for each of the groups of countries. A ‘downward sloping’ regression line indicates that regions that started with a high share of manufacturing in its GDP lost (most strongly) in the weight of its manufacturing sector as compared to regions that started with a lower share. Hence this would indicate the strongest deindustrialization in those regions which were originally more highly industrialized. This, for example, was the case for the Southern EU countries’ regions (EU–SE). The EU Central and Eastern European economies (EU–CEE11), on the other hand, present a different picture. There we see an upward sloping regression line, which indicates that regions that started 4 The information in the maps and following figures has been calculated on the basis of three-year averages: base period 2000–2 up to 2014–16 (the last year for which comparable regional data were available at the time of writing this chapter). This is in order to smooth over single-year outliers in the statistical data. In order to shorten the reference to these three-year averages, we shall refer at times to the mid-years 2001 and 2015 as abbreviations for these three-year averages.
OUP CORRECTED PROOF – FINAL, 24/06/20, SPi
234 Michael A. Landesmann and Joris M. SchrÖder
Shares of employment in manufaturing in % avg. 2014 – 2016
0.0 – 5.0 5.0 – 10.0 10.0 – 15.0 15.0 – 20.0 20.0 – 25.0 25.0 – 30.0 30.0 – 35.0 35.0 – 40.0 no data
Map 13.2 Shares of employment in manufacturing in total employment by NUTS-2 regions
with a higher share of manufacturing in GDP gained further rather than lost in weight in the regions’ GDP. For the more advanced Northern and Western EU economies’ regions (EU–NW+NO) we observe a horizontally sloped regression line, hence no particular positive or negative relationship between high (low) initial GDP share and following gain (or loss). However, looking at the data points for this group more closely, we do see quite a few regions that had initially a high share of manufacturing in GDP also gaining in weight over the subsequent period, but there is also a bunching of regions with a middle position in initial share of manufacturing losing shares over the period (i.e. an indication that they are ‘deindustrializing’). Figures 13.4a and 13.4b assess for the latest period for which regional statistical data were available whether there is a link between income levels of a region (calculated as GDP per capita) and the shares of employment in manufacturing (Figure 13.4a), on the one hand, and in ‘advanced tradable services’ (ATS) (Figure 13.4b), on the other. Take Figure 13.4b first: there we can see a clear (‘linear’) relationship between the importance of advanced tradable services industries and income per capita. ATS activities are prominently represented in the high and highest income European regions. A few EU–CEE11 regions also feature in that group (diamond icons) and these are the high-income capital city regions in the new EU member states. As regards the importance of manufacturing (Figure 13.4a) the situation is more complex: on the one hand, we see the bunching of regions with a high share of manufacturing
OUP CORRECTED PROOF – FINAL, 24/06/20, SPi
Agglomeration of European Industries 235
Change in shares of GVA in manufacturing 2001–15 in pp
10
5
0
–5
–10
0
10
40
20 30 Shares of GVA in manufacturing 2001 in %
EU–CEE11 Lfit EU–CEE11
EU–NW+NO Lfit EU–NW+NO
EU–SE Lift EU–SE
Figure 13.3 Shares in GVA of manufacturing 2000–2 vs. changes 2000/2–14/16 Source: Eurostat; calculations by authors.
amongst the low income regions of Central and Eastern Europe (diamond icons). On the other hand, there is also a group of middle to high income regions (mostly in EU–NW) also holding a strong position in manufacturing. This is further evidence of the complementarity we already saw in Map 13.2 between the rich regions in Germany, Austria, Northern Italy, Southern Sweden, and Finland, etc. and the low-income regions of Central and Eastern Europe through the formation of manufacturing cross-border production networks. On the other hand, if we leave out this group of higher-income regions of NW Europe involved in regional production networks with Central and Eastern Europe, the rest of EU middle- and high-income regions follow an inverse U-shaped curve, whereby the importance of manufacturing rises up to a certain level of income and then falls off. This inverse U-shaped pattern is well known in the development economics literature (see, for example, Chenery and Syrquin, 1975). In Europe, however, this pattern is somewhat modified through the strong involvement of (by European standards) low-income regions in Central and Eastern Europe in regional production networks together with neighbouring high-income regions of Western and Northern European countries. The final set of interesting diverging trends is depicted in Figure 13.5 where we show again ‘convergence’/‘divergence’ pictures at the regional level,5 but this time looking separately at high-, middle- and low-income regions. It is interesting to see again the rather diverging patterns of Central and Eastern European regions (EU–CEE11) from those in 5 Notice that in Figure 13.5 we look at shares of manufacturing in total gross value added (GVA) rather than in employment.
OUP CORRECTED PROOF – FINAL, 24/06/20, SPi
236 Michael A. Landesmann and Joris M. SchrÖder (a)
Shares of employment in manufacturing avg. 2014–16 in %
40
30
20
10
0 0
20000
40000
60000
80000
GDP per inhabitant 2016 in PPS
Figure 13.4a Employment shares in manufacturing (NUTS-2 regions) and GDP per capita levels Source: Eurostat; calculations by authors.
(b)
Shares of employment in adv. tradable services avg. 2014–16 in %
40
30
20
10
0 0
20000
40000
60000
80000
GDP per inhabitant 2016 in PPS EU–CEE11
EU–NW+NO
EU–SE
Figure 13.4b Employment shares in advanced tradable services (NUTS 2 regions) and GDP per capita levels Source: Eurostat; calculations by authors.
OUP CORRECTED PROOF – FINAL, 24/06/20, SPi
Agglomeration of European Industries 237 Europe’s South (EU–SE) and Europe’s North West (EU–NW+NO). While we observe for both low- and middle-income regions in Europe’s South and North West a clear picture of ‘de-industrialization’ (i.e. regions in each of these groups that had a relatively high share of manufacturing in 2000/2 experienced a rather strong decline in that share over the subsequent period), the Central and Eastern European regions did not—in the majority— experience such a decline in manufacturing shares. Rather the opposite, quite a few regions with a high initial share experienced further increases in their shares. Amongst the high-income regions, the situation is again somewhat different, but compatible with the story told so far. Quite a few of the high-income regions of North Western Europe maintain the strong position in manufacturing and even increase its importance (upward sloping regression line) which refers again to those regions that are involved in cross-border production networks in Central Europe. There are very few high-income regions in Central and Eastern Europe and these are the metropolitan regions which are moving away from manufacturing and towards advanced services (as we already saw in Figures 13.4a and 13.4b).
13.5 The Role of Cross-border Production Networks and Patterns of Vertical Specialization The role of cross-border production networks to explain the agglomeration patterns of manufacturing activity in Europe can now be further elucidated. The EU–CEE11 countries were major destination points for international investors from the mid-1990s onwards (see, for example, Hunya 2019). In quite a few of the countries, FDI moved quite strongly into so-called non-tradable sectors (banks, wholesale and retail trade, real estate), thus generating serious problems for the current accounts. However, in the EU–CEE5, in particular, FDI played a major role to significantly build up export capacities in the manufacturing sector. Particularly, the motor vehicles industry developed rapidly but so did other industries of the mechanical and electrical engineering sector. More recently, such production sites have also been set up in the Baltics (Estonia, Latvia, Lithuania) and the Balkans (Romania, Serbia, Macedonia, etc.). When we look at the characteristics of the regional production networks (RPNs) that were organized by Western European, particularly German (but also Austrian, Swiss, Italian, Swedish) companies, they had strong features of cross-border specialization on different production stages, so-called ‘vertical differentiation’ (see, for example, Arndt and Kierzkowski 2001). Thus, the availability of relatively low-wage industrial workers in Central and Eastern Europe led to a pattern whereby initially lower- and medium-skill activities were transferred to the candidate and then new member countries of the EU. Particularly, the regions close to the Western borders were chosen as production sites for such activities. Over time, also more advanced production stages were located in these countries and networks and domestic supplier networks emerged. In the context of the recently developed ‘smile curve economics’ (see, for example, Stöllinger 2019), one can observe in the EU–CEE5 a typical specialization on production
OUP CORRECTED PROOF – FINAL, 24/06/20, SPi
238 Michael A. Landesmann and Joris M. SchrÖder
Change in shares of GVA in manufacturing 2001–2015 in pp
(i) Low income regions: 10
5
0
–5 0
10
20 30 Shares of GVA in manufacturing 2001 in %
40
10
20 30 Shares of GVA in manufacturing 2001 in %
40
10
20 30 Shares of GVA in manufacturing 2001 in %
40
Change in shares of GVA in manufacturing 2001–15 in pp
(ii) Middle income regions: 10 5 0 –5 –10 0
Change in shares of GVA in manufacturing 2001–15 in pp
(iii) High income regions: 10 5 0 –5 –10 0
EU–CEE11 Lfit EU–CEE11
EU–NW+NO Lfit EU–NW+NO
EU–SE Lfit EU–SE
Figure 13.5 Shares in 2000/2 and change in shares of GVA of manufacturing 2000/2–14/16 Source: Eurostat; calculations by authors.
OUP CORRECTED PROOF – FINAL, 24/06/20, SPi
Agglomeration of European Industries 239
1
5 Support services
4 Logistics
0
3 Production
.5
2 R&D
5 Support services
0
4 Logistics
.5
1.5
1 HQ
Relative functional specialization
1
2 R&D
5 Support services
4 Logistics
3 Production
2 R&D
.5
DE
2
1.5
1 HQ
Relative functional specialization
1
1 HQ
Relative functional specialization
1.5
0
AT
2
3 Production
SK
2
Figure 13.6 Complementarities in functional specialization within the Central European manufacturing core Note: SK - Slovakia, AT - Austria, DE - Germany Source: FDI markets database, wiiw calculations.
activity while Germany and other Western European producers show a specialization profile in the direction of pre- and post-production stages. An example for such complementary specialization structure can be seen in Figure 13.6.6 The information in this graph has been compiled from detailed information provided by the FDI markets database7 which contains data on greenfield investments by international companies across the globe and which allows a classification of such investments by ‘functions’ such as production activities and other activities (headquarter, R&D, logistics, marketing, etc.) Recent analysis pointed to the importance of tracking whether countries/regions remain ‘stuck’ in relatively ‘low valueadded capturing’ stages in the value chain or which factors are important for countries/ regions to move upstream towards ‘higher value added capturing’. The analysis in this respect with respect to European RPNs is still ongoing.
13.6 Conclusions This chapter has analysed the agglomeration of industrial activities across the European economy. A particular emphasis was the uneven dynamic we observe in the development of 6 This figure has been supplied by Roman Stöllinger, wiiw. 7 For details see Stöllinger (2019).
OUP CORRECTED PROOF – FINAL, 24/06/20, SPi
240 Michael A. Landesmann and Joris M. SchrÖder manufacturing capacities across European regions. For low- to medium-income regions in Europe, maintaining or building up sufficient manufacturing capacities was emphasized as particularly important to counteract the tendencies towards sustained ‘external imbalances’ which has proved to be a serious strain for the European integration process. Relative to other advanced economies (such as the United States, but even Japan), Europe has overall maintained a strong position in manufacturing, but this has gone along with strong agglomeration tendencies over the past two to three decades. We emphasized in this chapter the emergence of a strong ‘Central European manufacturing core’ which has developed into a potent cross-border network of manufacturing and ancillary activities involving highly developed, high-income regions in Germany, Austria, Switzerland, Sweden, and the low- to middle-income regions in Central and Eastern Europe (CEE). The CEE countries and regions that participate in these regional production networks (RPNs) have avoided structural external imbalances problems and have, so far, successfully embarked on a path of up grading their positions in these RPNs. However, we also observe a rather wide belt of regions in Europe’s South that have lost out in terms of participating in these newly developing RPNs and have been deindustrializing without building up sufficient strengths in other areas of tradable activities. Advanced economies (such as the United Kingdom, Netherlands, Belgium) and also large conurbations more generally have the option of building up comparative advantages in the fast expanding areas of ‘advanced tradable services’ (ATS) but these are less accessible to less advanced countries and regions. Hence to avoid serious macroeconomic imbalances the location of manufacturing industries (or of ‘fragments’ or ‘stages’ of manufacturing processes) remains a very important issue to avoid major macroeconomic imbalances in the European economy. Future research will explore the further evolution of agglomeration of industrial activity in Europe and to which extent specialization patterns and imbalances get frozen or, alternatively, we might see a wider dissemination in the participation of RPNs. Thereby blockages in the development processes of significant regions of the European economy might be overcome. This is an eminently important not only economic, but also political question (on this see also, Dijkstra et al. 2015; Rodriguez-Pose 2018; and Storper 2018).
References Arndt, Sven W. and Hendryk Kierzkowski (2001) Fragmentation: New Production Patterns in the World Economy. Oxford: Oxford University Press. Artis, Michael J., Ernest Miguelez, and Rosina Moreno (2009) ‘Assessing Agglomeration Economies in a Spatial Framework with Endogenous Regressors’. CEPR Discussion Paper No. 7267, London. Baldwin, Richard (2013) ‘Global Supply Chains: Why They Emerged, Why They Matter, and Where They Are Going’, in Deborah K. Elms and Patrick Low (eds) Global Value Chains in a Changing World. Geneva: World Trade Organization, pp. 13–59. Baldwin, Richard and Javier Lopez-Gonzalez (2015) ‘Supply-chain Trade: A Portrait of Global Patterns and Several Testable Hypotheses’, The World Economy 38(11): 1682–721. Baldwin, Richard and Philippe Martin (2004) ‘Agglomeration and Regional Growth’, in Vernon Henderson and Jacques-Francois Thisse (eds) Handbook of Regional and Urban Economics, Vol. 4. Amsterdam: Elsevier Science, pp. 2671–711. Bekkers, Eddy, Michael Landesmann, and Indre Macskasi (2017) ‘Trade in Services versus Trade in Manufactures: The Relation between the Role of Tacit Knowledge, the Scope for Catch-up, and
OUP CORRECTED PROOF – FINAL, 24/06/20, SPi
Agglomeration of European Industries 241 Income Elasticity’. Working Paper No. 139. The Vienna Institute for International Economic Studies (wiiw), Vienna. Boschma, Ron (2005) ‘Proximity and Innovation: A Critical Assessment’, Regional Studies 39(1): 61–74. Brülhart, Marius and Nicole A. Mathys (2008) ‘Sectoral Agglomeration Economies in a Panel of European Regions’, Regional Science and Urban Economics 38(4): 348–62. Brülhart, Marius and Federica Sbergami (2009) ‘Agglomeration and Growth: Cross-country Evidence’, Journal of Urban Economics 65(1): 48–63. Chenery, Hollis B. and Moises Syrquin (1975) Patterns of Development: 1950–1970. New York: Oxford University Press for the World Bank. Ciccone, Antonio (2002) ‘Agglomeration Effects in Europe’, European Economic Review 46(2): 213–27. Ciccone, Antonio and Robert E. Hall (1996) ‘Productivity and the Density of Economic Activity’, American Economic Review 86(1): 54–70. Crescenzi, Riccardo and Andres Rodriguez-Pose (2012) ‘Infrastructure and Regional Growth in the European Union’, Papers in Regional Science 91(3): 487–513. Crozet, Matthieu and Pamina Koenig (2007) ‘The Cohesion vs. Growth Tradeoff: Evidence from EU Regions (1980–2000)’, Research Paper, University of Paris I. Dijkstra, Lewis, Enrique Garcilazo, and Philip McCann (2015) ‘The Effects of the Global Financial Crisis on European Regions and Cities’, Journal of Economic Geography 15(5): 935–49. Duranton, Gilles and Diego Puga (2001) ‘Nursery Cities: Urban Diversity, Process Innovation, and the Life Cycle of Products’, American Economic Review 91(5): 1454–77. Egeraat, Chris van and David Jacobson (2006) ‘The Geography of Linkages in the Irish and Scottish Computer Hardware Industry: The Role of Information Exchange’, Journal of Economic and Social Geography 97(4): 45–81. Ellison, Glenn and Edward Glaeser (1997) ‘Geographic Concentration in US Manufacturing Industries: A Dartboard Approach’, Journal of Political Economy 105(5): 889–927. ESPON (2004) Transport Services and Networks: Territorial Trends and Basic Supply of Infrastructure for Territorial Cohesion. Luxembourg: ESPON. ESPON (2013) The Case for Agglomeration Economies in Europe: Targeted Analysis 2013/2/1; Appendix C1: Investigating Agglomeration Economies in a Panel of European Cities and Regions. Luxembourg: ESPON. Fujita, Masahisa and Jacques-Francois Thisse (2002) Economics of Agglomeration: Cities, Industrial Location and Regional Growth. Cambridge: Cambridge University Press. Fujita, Masahisa, Paul Krugman, and Anthony J. Venables (1999) The Spatial Economy: Cities, Regions and International Trade. Cambridge, MA: The MIT Press. Gardiner, Ben, Ron Martin, and Peter Tyler (2010) ‘Does Spatial Agglomeration Increase National Growth? Some Evidence from Europe’, Journal of Economic Geography 11(6): 1–28. Gordon, Ian R. and Philipp McCann (2013) ‘Industrial Clusters: Complexes, Agglomeration and/or Social Networks?’, Urban Studies 37(3): 513–32. Hunya, Gabor (2019) ‘Outward FDI and Intra-regional Integration of EU-CE11 Countries’, Monthly Report, The Vienna Institute for International Economic Studies (wiiw), March, Vienna, pp. 13–21. Iammarino, Simona, Andres Rodriguez-Pose, and Michael Storper (2019) ‘Regional Inequality in Europe: Evidence, Theory and Policy Implications’, Journal of Economic Geography 19(2): 273–98. Landesmann, Michael (2015) ‘The New North–South Divide in Europe—Can the European Convergence Model Be Resuscitated?’, in Jan Fagerberg, Staffan Laestadius, and Ben R. Martin (eds) The Triple Challenge for Europe: The Economy, Climate Change and Governance. Oxford: Oxford University Press, pp. 60–90. Landesmann, Michael and Doris Hanzl-Weiss (2016) ‘Correcting External Imbalances in the European Economy’, in Marek Belka, Ewald Nowotny, Pawel Samecki, and Doris Ritzberger-Gruenwald (eds) Boosting European Competitiveness: The Role of the CESEE Countries. Cheltenham: Edward Elgar, pp. 14–36. Malmberg, Anders and Peter Maskell (2002) ‘The Elusive Concept of Localisation Economies: Towards a Knowledge-based Theory of Spatial Clustering’, Environment and Planning A 34(3): 429–49.
OUP CORRECTED PROOF – FINAL, 24/06/20, SPi
242 Michael A. Landesmann and Joris M. SchrÖder Malmberg, Anders, Bo Malmberg, and Per Lundquist (2000) ‘Agglomeration and Firm Performance: Economies of Scale, Localisation and Urbanisation among Swedish Export Firms’, Environment and Planning A 32(2): 305–21. Marshall, Alfred (1898) Principles of Economics, 4th edition. London: Macmillan. Marshall, Alfred (1919) Industry and Trade. London: Macmillan. Martin, Philippe (1999) ‘Public Policies, Regional Inequalities and Growth’, Journal of Public Policies 73(1): 85–105. Martin, Ron L. (2008) ‘National Growth versus Spatial Equality? A Cautionary Note on the New “Trade-off ” Thinking in Regional Policy Discourse’, Regional Science, Policy and Practice 1(1): 3–13. Moreno, Rosina, Raffaele Paci, and Stefano Usai (2005) ‘Geographical and Sectoral Clusters of Innovation in Europe’, Annals of Regional Science 39(4): 715–39. Oerlemans, Leon and Marius Meeus (2005) ‘Do Organisational and Spatial Proximity Impact on Firm Performance?’, Regional Studies 39(1): 89–104. Petrakos, George, Andres Rodriguez-Pose, and Antonius Rovolis (2005) ‘Growth, Integration and Regional Disparities in the European Union’, Environment and Planning A 37(10): 1837–55. Rice, Patricia, Anthony J. Venables, and Eleonora Patacchini (2006) ‘Spatial Determinants of Productivity: Analysis for the Regions of Great Britain’, Regional Science and Urban Economics 36(6): 727–52. Rodriguez-Pose, Andres (2018) ‘The Revenge of the Places That Don’t Matter (and What To Do about It)’, Cambridge Journal of Regions, Economy and Society 11(1): 189–209. Rosenthal, Stuart and William Strange (2004) ‘Evidence on the Nature and Sources of Agglomeration Economies’, in Vernon Henderson and Jacques-Francois Thisse (eds) Handbook of Economic Growth, Vol. 4. Amsterdam: Elsevier, pp. 2119–71. Simpson, Wayne (1992) Urban Structure and the Labour Market: Worker Mobility, Commuting and Underemployment in Cities. Oxford: Clarendon Press. Stehrer, Robert and Roman Stöllinger (2015) ‘The Central European Manufacturing Core: What Is Driving Regional Production Sharing?’, FIW-Research Report 2014/15(2), Vienna. Stöllinger, Roman (2019) ‘Testing the Smile Curve: Functional Specialisation in GVCs and Value Creation’. Working Paper No. 163. The Vienna Institute for International Economic Studies (wiiw), Vienna. Storper, Michael (1995) ‘The Resurgence of Regional Economies, Ten Years Later: The Region as a Nexus of Untraded Interdependencies’, European Urban and Regional Studies 2(3): 191–221. Storper, Michael (2018) ‘Separate Worlds? Explaining the Current Wave of Regional Economic Polarization’, Journal of Economic Geography 18(2): 247–70. Timmer, Marcel P., Erik Dietzenbacher, Bart Los, Robert Stehrer, and Gaaitzen J. de Vires (2015) ‘An Illustrated User Guide to the World Input–Output Database: The Case of Global Automotive Production’, Review of International Economics 23(3): 575–605. Wheeler, Christopher (2001) ‘Search, Sorting and Urban Agglomeration’, Journal of Labor Economics 19(4): 879–99.
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
chapter 14
The Gl oba l Ex per ience i n Speci a l Econom ic Zon e s A History and Review Thomas Farole
14.1 Introduction For more than fifty years since the establishment of the first modern special economic zones (SEZs),1 countries have put their faith in this policy instrument as a source of economic transformation. In some countries, the model has delivered spectacularly, with SEZs playing a catalytic role in growth and structural transformation. This includes East Asia’s ‘tiger economies’ and of course China, which has used SEZs as a platform to support the development of export-oriented manufacturing. In Latin America, the Dominican Republic, El Salvador, and Honduras, amongst others, used free zones to take advantage of preferential access to US markets and have generated large-scale manufacturing sectors in economies previously reliant on agricultural commodities. In the Middle East and North Africa, SEZs have played an important role in catalysing export-oriented diversification in countries like Egypt, Morocco, and the United Arab Emirates. And in sub-Saharan Africa, Mauritius is an example where SEZs were a central policy tool supporting a highly s uccessful process of economic diversification and industrialization. In addition, SEZs have played an important political economy role. In many countries, they have supported partial exposure to global markets while maintaining protective bar riers in a ‘stepwise’ approach to reform. SEZs have helped pilot experimental new policies before rolling them out to the broader economy; and in the absence of political will to undertake reforms, have acted as ‘second-best environments’ and ‘pressure valves’ to absorb excess labour. 1 The term ‘special economic zone’ is used generically in this chapter to encompass a wide range of spatially defined interventions.
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
244 Thomas Farole But despite high-profile success stories like China, SEZs have a decidedly mixed record in supporting sustainable industrialization. In many locations, SEZs have created employment in the short term but failed to sustain competitiveness in the face of rising wages or eroding trade preferences. In other cases, zones failed to attract investors, leaving ‘white elephants’ that did both fiscal and political damage. This chapter proceeds as follows.2 In section 14.2, we flesh out the definition of SEZs and alternative approaches, along with a discussion on the way in which the instrument has traditionally been employed by policymakers. Section 14.3 is a brief historical overview of SEZs and discussion of how current changes in trade and technology are reshaping traditional approaches to zones. This is followed in section 14.4 by key lessons learned for the design and implementation of effective zones and, finally, by a brief conclusion in section 14.5.
14.2 Definitions, Objectives, and Forms of SEZs SEZs are demarcated geographical areas within a country’s national boundaries where the rules of business are different—generally more liberal—from those that prevail in the national territory. Specifically, most economic zones create a ‘special’ regime that confers four main advantages to investors relative to what they could normally receive in the domestic environment:3 1) infrastructure (including serviced land, factory shells, and utilities) that is easier to access and more reliable than is normally available domestically; 2) a special customs regime including efficient customs administration and (usually) access to imported inputs free of tariffs and duties; 3) an improved regulatory and administrative regime, including streamlined procedures for company set-up, licensing, and operations; and usually; 4) an attractive fiscal regime, including reduction or elimination of corporate taxes, VAT, other taxes, labour contributions (e.g. pension/social security), and sometimes training or other subsidies. SEZs are designed as instruments of trade, investment, and ultimately of spatial industrial policy. They are generally established with four specific (although by no means exclusive) policy goals (FIAS 2008). First, and foremost, SEZs are designed to facilitate trade and attract foreign direct investment—virtually all SEZ programmes from traditional export processing zones (EPZs) to China’s large-scale SEZs have as one of their main objectives the promotion of trade, normally exports, and most aim to do this primarily by attracting FDI. Beyond this common theme, the objectives of SEZ programmes depend very much on how the
2 This chapter draws heavily on Farole (2011) and Farole and Akinci (2011). 3 See Farole (2011: chs 1 and 2).
OUP CORRECTED PROOF – FINAL, 23/06/20, SPi
The Global Experience in Special Economic Zones 245 government views economic reform in the context of wider development and structural transformation of the country’s overall economy. Thus, zones may be used to:4 • Unlock agglomeration economies: By concentrating economic infrastructure and public goods in one geographic area, SEZs enable industries to overcome minimum size thresholds and begin to leverage scale economies (Collier and Page 2009). This happens through exploiting backward and forward linkages (Ottaviano and Puga 1998); through labour pooling which facilitates matching between firms and workers (Combes and Duranton 2006); and through technology spillovers (Rodríguez-Pose and Crescenzi 2008). • Pilot the application of experimental new policies and approaches: China’s wide-area SEZs are classic examples—financial, legal, labour, and even pricing reforms were introduced and tested first within the SEZs before being extended to the rest of the economy. • Support a wider economic reform strategy: In this approach, SEZs are a simple tool permitting a country to develop and diversify exports. Zones are a way of reducing anti-export bias while keeping protective barriers intact. The EPZs of Taiwan–China, Mauritius, and the Republic of Korea follow this pattern. • Serve as ‘pressure valves’ to alleviate large-scale unemployment: In the absence of political will to undertake reforms, zones have still been used effectively as enclaves to absorb excess labour. Tunisia and the Dominican Republic are examples of robust, job- creating SEZ programmes that had limited impact beyond the walls of the SEZs. SEZs meet these objectives by overcoming barriers that hinder investment in the wider economy, such as restrictive policies, poor governance, inadequate infrastructure, and problematic access to land, amongst others. By doing so, and more broadly by signalling a spatial concentration in investment, SEZs may overcome coordination failures that prevent the private sector from investing in a location. SEZs, and spatial industrial policy initiatives more broadly, take many different forms. Form follows function, and so the physical form of the spatial intervention, in theory at least, will match its policy objective, that is, the problem it is trying to solve. Table 14.1 summarizes the most common forms of these spatial regimes, ranging from (typically natural) agglomerations5 and clusters6 through to simple industrial parks and ultimately large-scale SEZs. Figure 14.1 provides a basic hierarchy of the package of amenities that may exist in an SEZ programme. Starting with the most basic aspects of an SEZ programme—land and valueadding infrastructure—where the policy environment is attractive for investors but land and industrial infrastructure access is binding, a standard industrial park model is normally sufficient. If land is accessible and infrastructure is of good quality, but business facilitation is problematic, the establishment of an effective one-stop shop may obviate the need for an 4 Farole (2011) based on FIAS (2008). 5 An agglomeration is simply a description of the fact that economic activity is not distributed evenly over space; we observe an agglomeration where economic activity concentrates. It may but does not necessarily have any relationship with specialization or clusters. 6 A cluster describes a specific form of agglomeration related to specialization. To observe a cluster, we should not only see a spatial concentration of firms in a specific sector but also a concentration of firms in related industries and in supporting industries (e.g. business services)—generally we will see firms from across a value chain located in a relatively narrow spatial area.
Co-location of firms in the same sector
Co-location of related and supporting industries and institutions
Agglomeration
Maybe, but not necessarily
Maybe, but not necessarily
Cluster
Yes
Industrial park
Typical size
Activities
Special customs regime
No
Varies—typically several sq kms
Varies
No
No
Yes
Not usually
Varies from 10ha to large regions
Varies
Not usually
Maybe, but not necessarily
Maybe, but not necessarily
Not usually
Yes