Intro To Clusters

Regional Industry Cluster Analysis
Edward J. Feser Department of City & Regional Planning University of North Carolina at Chapel Hill
Introduction to
Objectives
What are industry clusters? What is industry cluster policy? What are some techniques for conducting industry cluster analyses? What can you gain from industry cluster analysis? Resources and guides
Copyright 2001 Prof. Edward Feser, Department of City & Regional Planning, UNC-Chapel Hill 2
What are industry clusters?
Introduction to industry cluster

concepts
Interest in clusters coincides with focus on competitiveness Erosion of traditional sources of comparative advantage Innovation as growth engine Michael Porters framework
Macroeconomic, political, legal context
Productivity growth
Microeconomic foundations
Company operations & strategy Micro business environment
Internal
External
From Porter (May, 2000)
Copyright 2001 Prof. Edward Feser, Department of City & Regional Planning, UNC-Chapel Hill
Porters model of competitive advantage

Four major determinants of competitive advantage Clusters as one element Clusters are not necessarily regional in scope Measurement unclear Limited verification Related and Related and supporting supporting industries industries
Clusters, not industries
Firm strategy, Firm strategy, structure and structure and rivalry rivalry
Local context encourages investment and sustained upgrading
Demand Demand conditions conditions
Factor Factor conditions conditions

Factor quantity, quality, cost
Sophisticated and demanding local customers

From Porter (May, 2000; 1990) 4
institutions:
Trading sectors
A cluster consists of interdependent firms and

Related sectors
Similar technologies Share pool of labor Similar strategies
Supporting institutions
Education (univ, colleges) Training (ccs) R&D (univ, fed labs) Development agencies Regulatory agencies
Intermediate suppliers Capital good suppliers Producer services Consultants Contract R&D
Interdependence: Each member firms competitive
position depends on one, some, or all other members of the group.
What is industry cluster policy?
Why care about industry clusters?

Q: Why does success seem to breed success? Q: What do growing knowledgeintensive and high tech sectors need to survive? Q: What do traditional sectors need to survive? Q: What are you overlooking with standard industry analysis alone? Q: What can you gain by targeting efforts? A: The benefits of being the first mover. A: Opportunities and incentives to learn, create, upgrade. A: Same thing, as historical sources of advantage are ceded to overseas competitors. A: Potential strengths, underlying weaknesses, latent opportunities, new ways of tackling old problems. A: Efficient, effective use of scarce development resources.
6
Policy uses vary according to degree to

Degree of joint policy coordination around cluster concept
Low Inform existing programs Support existing or emerging clusters High Actively promote clusters
which cluster concepts inform existing initiatives or serve as the basis for wholly new strategies.
approach to cluster policymaking
The most common
Cluster definition Analysis Findings

Input to analysis
Policy implications
Output from analysis
Policy design Implementation
Example: Environmental technologies cluster

Value chain: Support supply
University-industry linkages, tech transfer, commercialization Training and trade schools Buyer-supplier conferences, networking, trade shows Venture capital Incubators Recruitment
Users: Build demand

Regulation/enforcement Resource pricing Purchasing
Environmental Equipment Sustainable goods Intermediate suppliers Environmental services
Public sector
Compliance assistance Credits and incentives
Industry
Education Marketing Recycling services
Consumers
Resources management
Arrow indicates trading linkage
approach to cluster policymaking
Another
Policy problem Cluster definition Analysis Findings Policy design Implementation
Framing Rationale Existing/potential Spatial concentration Comprehensiveness Generalizability
Input to analysis Output from analysis
Industry involvement Analysis
10
How to do a cluster analysis
types of applied cluster analyses General cluster analysis
Typical policy needs imply two major

Assessment of linkages and trends in all regional industries Comprehensive and detailed Application of generalized cluster definition or use of multiple definitions
Focused cluster analysis

Investigation of one dimension of cluster concept -and/or Investigation of one or a few key industry segments
11
Conducting a cluster study: Challenges

Relating concepts to measures Obtaining necessary data Applying data-intensive and complex analytical methods Interpreting findings Conveying findings
Begin by relating concepts to measures

Three dimensions of measurement: Linkage
Time
Geography
Linkage: Innovation, labor, inputs Geography: Localized, non-localized Time: Existing, declining, emerging or potential
Linkage dimension
What binds the firms in the cluster together?
Value-chain clusters: Firms that are members of the
same extended value (product) chain
Labor-based clusters: Firms that share similar labor

requirements (utilize a shared labor pool)
Innovation clusters: Firms that exchange key
information, knowledge that leads to technological improvements or changes
14
Geographic dimension
Are member firms co-located in specific regions?
Localized: Clusters in which firms are co-located in
specific regions of state (or regions that span state borders)
Non-localized: Linked firms that are not localized in any

particular region but exist across the state or that are concentrated in multiple regions
15
Time dimension
At what stage of development are identified clusters in?
Existing: Clusters with significant presence and exemplify
particular linkages that are the focus of study; usually presumes expansion expanding
Declining: Existing clusters that are declining rather than Emerging: Clusters that appear to gaining a significant
presence but havent achieved a critical mass yet
Potential: Clusters that might emerge given trends, luck,

or policy stimulation
General Features of different methods

Method
Expert opinion
Advantages
Easy Inexpensive Detailed contextual info Easy, inexpensive Can supplement methods Only major source of data on interdependence in U.S. Comprehensive and detailed Key measure of interdependence Can be used to identify labor affinities Visualization aids interpretation and analysis Flexibility to collect ideal data; current
Pitfalls
Not generalizable No matter how much you believe, its still opinion Focus is on sectors, not clusters May be dated Industry definitions imperfect Neglects supporting institutions Data not available in U.S. (see OECD activities) Occupation rather than skills based; fairly aggregate Methods, software still limited Costly Difficult to implement properly
17
Specialization indicators (LQs) Input-output: Trade Input-output: Innovation Staffing patterns: Labor Graph theory/ network analysis Surveys
Framing helps determine approach

Identify and frame the policy problem first, i.e., KNOW WHAT YOU WANT Cluster definition Existing? Potential or emerging? Spatial or non-spatial? Type of linkage? Methodology Comprehensive? Generalizable? Qualitative? Quantitative? Implementation Industry buy-in? Analysis only? Marketing only?
18
Policy problem: What will you use the study for?
Determines:
Focus on value-chain and labor clusters

No applicable data to identify innovation-based clusters Value-chain and labor based clusters can provide good first approximation to innovation-based clusters Qualitative methods can be used to identify innovation linkages within VC or labor clusters
19
Value-chain clusters: Identify interdependencies

Tradebased Current, local Supporting institutions
Regional input-output Simple descriptive indicators Survey (of existing linkages) Expert opinion Survey Interviews Expert opinion National input-output Available studies of i-o relations Simple descriptive indicators Survey (of buyer-supplier needs) Survey (of other needs) Secondary data Expert opinion
Tradebased Emerging/ potential Supporting institutions
20
Two methods of VC cluster analysis

Bottom-up method:
Metal stampings, n.e.c. Logging Petroleum refining Veneer and plywood Sawmills and planing mills
Identify value chain for a single sector by finding key first, second, third tier suppliers
Non-upholstered wood furniture
If an intermediate goods industry, also find key buying sectors Regional input-output approximates local trading patterns inputNational input-output approximates potential trading patterns inputUse other methods to identify related sectors and supporting institutions
21
Two methods of VC cluster analysis

Top-down method:
Identify an exhaustive set of industry clusters through comprehensive analysis of interdependence between all sectors
Detailed interindustry transactions data: For each sector, sales to and purchases from every other sector
Result: Cluster Templates

Cluster 1 Cluster 2 . . . Cluster n
Apply data reduction techniques (factor analysis, statistical cluster analysis, etc.)
22
Benchmark U.S. manufacturing clusters

Metalworking Vehicle manufacturing Brake & wheel products Chemicals & rubber Concrete, cement, brick Petroleum refining Nonferrous metals Aluminum Wood products Printing & publishing Electronics & computers Platemaking & typesetting Earthenware products Knitted goods Fabricated textiles Leather goods Aerospace
Heavy manufacturing
Light manufacturing
Independent industries
Packaged foods Canning & bottling Feed products Dairy products Meat products Tobacco
Food goods & tobacco

23
Features of benchmark clusters

Amounts to value chain-based rather than product-based manufacturing classification system 3- and 4-digit SIC industries may be members of multiple clusters Each cluster includes direct and indirect trading partners Includes measure of strength of linkage between component sector and broader cluster Reasonable initial estimate of interdependent industries even where trading links dont exist Can be easily combined with qualitative cluster analysis methods
Cluster
Metalworking Vehicle Manufacturing Chemicals & Rubber Electronics & Computers Packaged Foods Printing & Publishing Wood Products Knitted Goods Fabricated Textile Products Nonferrous Metals Canned & Bottled Goods Leather Goods Aerospace Feed Products Platemaking & Typesetting Aluminum Brake Products Concrete, Cement, & Brick Earthenware Products Tobacco Products Dairy Products
No. of 2-Digit SIC Sectors Sectors

116 58 48 38 44 32 23 23 22 14 12 9 10 10 14 9 9 8 8 4 6 10 16 14 8 5 8 6 5 9 4 2 1 6 2 7 4 4 2 1 1 1
24
Aerospace: illustration of a value-chain

Cluster ID Input output code 600200 600100 130100 600400 380900 381400 (s) 430100 (s) 381300 (s) 620101 (s) 560500 (s) SIC sector 3724, 3764 3721 3761 3728, 3769 3356 3463 3511 3364, 3369 381 3663, 3669 Description Aircraft and missile engines and engine parts Aircraft Guided missiles and space vehicles Aircraft and missile equipment, n.e.c. Nonferrous rolling and drawing, n.e.c. Nonferrous forgings Turbines and turbine generator sets Nonferrous castings, n.e.c. Search and navigation equipment Communication equipment Load ing .92 .90 .82 .70 .69 .59 .58 .56 .41 .36 L1 13 13 13 13 13 13 1 10 4 4 L2 L3 4 4 10 13 13 16,1 13 13
(s) indicates secondary sector
Loading indicates strength of linkage between sector and cluster Column labeled cluster ID indicates codes for other clusters in which sector is a member Secondary sectors are those only moderately tied to the cluster
25
Methods:
Analysis with Labor-Based Industry Clusters
U.S. Staffing Patterns Matrix

1996, Projected 2006 Employment
501 Occupations 241 Industries
of 501 x 241 labor requirements matrix on shares of workforce in each occupation
Statistical Cluster Analysis
of 501 x 241 labor requirements matrix staffing patterns defined in 0/1 form
Statistical Cluster Analysis
Judgement
Analysis of regional trends with labor cluster templates similar to top-down VC approach
32 Laborbased U.S. industry clusters
26
Illustration I: North Carolina cluster project

Policy need: Economic rationale: Cluster definition: Methodology: Implementation: Target technology adoption programs more effectively Technologies diffuse through value chain, focused analysis Potential and emerging clusters, value-chain based Input-output based, top down approach Primarily analysis, planning for manufacturing sector
27
Illustration II: Scranton/Wilkes-Barre

Policy need: Economic rationale: Cluster definition: Methodology: Implementation: Develop a strategic plan for technology-related development No a priori rationale based on policy need, general analysis Existing, emerging and potential, value-chain, labor Bottom-up input-output, labor cluster analysis Primarily analysis to inform planning efforts
28
North Carolina project scope

Identify U.S. value chains (template clusters) Examine presence of chains in North Carolina
Spatial distribution
Statewide study ! 7 regional studies ! GIS-based analysis
!
Specializations and gaps

Structure (breadth) ! Size (depth) ! Technology
!
Growth and turbulence

Overall growth ! Growth sectors ! Expansions, contractions
!
29
What can be done with national-level value

chains at the state or regional level?
Look for relative presence of each chain in the state or region Look for components of the chain present Determine why some elements of the chain are missing Determine why other elements are present Combine with other methods (surveys, regional input-output) to examine regional trading patterns Do not assume regional economy should parallel national economy, but do use templates to better understand your current industry mix Chains may serve as the basis for identifying the challenges and opportunities facing the regions industry
30
Vehicle manufacturing an NC
strength revealed by top-down cluster analysis
SIC 37, transportation equipment industry accounts for 2.5 percent of state value-added and 3 percent of employment The vehicle manufacturing cluster is the second-largest in the state Growth of the cluster is most pronounced in central and western parts of the state Pattern and nature of cluster growth is consistent with national southward shift of auto production
60%
Triangle Piedmont
40% 20% 0% -20%
Northeast
Western Carolinas
Area:
Transpark Southeast
Regional shares of statewide vehicle manufacturing cluster employment, 1989
Outline: Regional shares of statewide VF cluster employment growth, 1989-1994
31
North Carolina value-chains specialized in

lower-productivity, lower wage sectors
Metalworking Vehicle Manufacturing Chemicals & Rubber Electronics & Computers Packaged Food Products Printing and Publishing Wood Products Knitted Goods Fabricated Textile Products 0.00 0.05 0.10 0.15 0.20
Share of total cluster value-added in high productivity sectors, 1994

United States North Carolina
0.25 0.30 0.35
32
Metalworking
Southeast region
linkages likely tighter in Triangle than

Most tightly linked sector in cluster
Triangle Region
Southeast Region
Length of line from origin indicates number of employees in given sector; Axis dimensions range from 0 (at center) to 3,000 at edge.
Axes are sectors that make up the cluster, ordered clockwise by the strength of the linkage between the sector and overall cluster.
33
Scranton/Wilkes-Barre project scope

Identify key sources of industrial growth during 1990s
Specializations High growth sectors relative to the U.S., controlling for national trends Start-up activity
Identify groups of industries (industry clusters) whose members a strong degree of potential interdependence and shared competitiveness Identify any emerging industries that could take advantage of existing up- and downstream industrial activity Identify supplier and purchasing sectors that might constitute viable development targets Focus on higher value, higher technology sectors
34
Study Developed overlapping evidence

Value Chain Analysis Basic Industry Trends Labor Cluster Analysis
35
Step 1: Basic Industry Analysis

Specializations (LQ>1.2) Positive Competitive shift
Strong Start-up Activity
36
Step 2: Labor cluster analysis

Specializations (LQ>1.2) Positive Competitive shift
Strong Start-up Activity
37
Step 3: Bottom up cluster analysis

1. Identification of 1st and 2nd tier supplier and purchasing chains for each of 413 basic industries Supplier sector i Supplier sector j Supplier sector k Study sector i (i, . . . , j=413) Purchasing sector i Purchasing sector j Purchasing sector k
2. Examination of characteristics of each chain
Size Breadth (sectors present in region) Specialization, current and trend Expansion (aggregate and share growth sectors) Depth (diversity, trend in diversity)
3. Identification of core regional industries 4. Reduced Set of regional industries with strong or emerging potential up- and/or downstream linkages
Core Industry Analysis Value Chain Analysis

Size Specialization, current and trend Expansion (absolute and relative)
Potential regional cluster industries
38
Step 4: Interpret overlapping evidence

Look for consistent results across analyses: VC clusters should overlap with labor clusters Weakness of approach
Interpretation involves considerable art
Strength of approach
The notion of a single set of clusters for a give region unreasonable except at the most aggregate level Provides a level of detail that can facilitate strategic planning Can serve as a framework for continuous monitoring
39
Summary
Summary thoughts
Think policy needs first Identify appropriate clustering dimension to focus on Are localized clusters what youre after? Existing, emerging, potential? Utilize existing studies (e.g., templates), at least as a start If you commission a study, demand the following:
Full and detailed disclosure of all methods and data An actual assessment of interdependence, not just specialization Follow-up interpretation and support
40
Resources
Available guides
Old and New Theories of Industry Clusters, by E. Feser. In Clusters and Regional Specialisation: On Geography, Technology, and Networks, edited by M. Steiner, Pion Ltd., 1998. Networks, (www.pion.co.uk/ep/errs/errs1st.html) Industry Clusters: a Methodology and Framework for Regional Development Policy in the U.S., by E. Development Bergman and E. Feser, Forthcoming in Cluster Analysis and Cluster Based Policy: New Perspectives and Rationale in Innovation Policymaking, edited by T. Roelandt and P. den Hertog. Paris: OECD, Policymaking, Hertog. 1999. National industry cluster templates: A framework for applied regional cluster analysis, by E. Feser and regional E. Bergman, Forthcoming Regional Studies, 1999. (Contains all data in this presentation.) Studies, Industry Clusters: An Internet Resource for Economic Developers, by J. LeVeen, 1998. Developers, (www.unc.edu/depts/dcrpweb/courses/261/leveen/index.html) Targeting North Carolina Manufacturing: Understanding the States Economy through Industrial Cluster States Analysis, by E. Bergman, E. Feser, and S. Sweeney, North Carolina Alliance for Competitive Analysis, Alliance Technologies, 1996. (Contains set of U.S. manufacturing value chains based on 1987 input-output chains inputaccounts.) Industrial Strength Strategies: Regional Business Clusters and Public Policy, by S. Rosenfeld, Aspen Public Policy, Institute, 1995. The Competitive Advantage of Nations, by M. Porter, Basic Books, 1990. Nations,
41

Intro To Clusters

Uploaded by

Copyright:

Available Formats

Intro To Clusters

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Intro To Clusters

Uploaded by

Copyright:

Available Formats

Regional Industry Cluster Analysis

What are industry clusters?

Introduction to industry cluster

What are industry clusters?

Porters model of competitive advantage

Local context encourages investment and sustained upgrading

Demand Demand conditions conditions

Factor Factor conditions conditions

Sophisticated and demanding local customers

What are industry clusters?

A cluster consists of interdependent firms and

Interdependence: Each member firms competitive

position depends on one, some, or all other members of the group.

What is industry cluster policy?

Why care about industry clusters?

What is industry cluster policy?

Policy uses vary according to degree to

What is industry cluster policy?

approach to cluster policymaking

The most common

Cluster definition Analysis Findings

Policy design Implementation

Example: Environmental technologies cluster

What is industry cluster policy?

Users: Build demand

Environmental Equipment Sustainable goods Intermediate suppliers Environmental services

Compliance assistance Credits and incentives

Arrow indicates trading linkage

What is industry cluster policy?

approach to cluster policymaking

Policy problem Cluster definition Analysis Findings Policy design Implementation

Framing Rationale Existing/potential Spatial concentration Comprehensiveness Generalizability

Input to analysis Output from analysis

Industry involvement Analysis

How to do a cluster analysis

types of applied cluster analyses General cluster analysis

Typical policy needs imply two major

Focused cluster analysis

How to do a cluster analysis

Conducting a cluster study: Challenges

How to do a cluster analysis

Begin by relating concepts to measures

How to do a cluster analysis

Labor-based clusters: Firms that share similar labor

Innovation clusters: Firms that exchange key

information, knowledge that leads to technological improvements or changes

How to do a cluster analysis

Non-localized: Linked firms that are not localized in any

How to do a cluster analysis

Potential: Clusters that might emerge given trends, luck,

How to do a cluster analysis

General Features of different methods

How to do a cluster analysis

Framing helps determine approach

Policy problem: What will you use the study for?

How to do a cluster analysis

Focus on value-chain and labor clusters

How to do a cluster analysis

Value-chain clusters: Identify interdependencies