1. Keynote, IV Clusters, Cluster Policies & European Networking Conference
Santiago de Compostela, 7th May 2014
WHERE NEXT FOR CLUSTERS
AND CLUSTER POLICY?
James Wilson
Orkestra and Deusto Business School
jamierwilson
2. Motivation: We can’t stay still
?
The world never stops moving
and our competitiveness
landscape is always changing …
… so becoming and remaining
competitive requires continual
transformation
In competitiveness
policy we need to
constantly question
what we are doing
and how we are
doing it
Esbjerg, Den
mark
Belfast, Norther
n Ireland
3. Policy must evolve with challenges
•Industrial
Restructuring
•Response to
economic
crisis
•Investment in
key STI
infrastructure
1980s
•Clusters
•Focus on
efficiency-driven
competitiveness
•Proactive policy
to improve
business
environment
1990s
•Innovation
•Evolution of
cluster policy
•Foundations
of smart
specialisation
strategy
2000s
An example: Evolution of
competitiveness policy in the
Basque Country
Today in the Basque Country
we are at another important
juncture … where next?
4. Clusters and cluster policy
The cluster concept in its current form has been
around for a quarter of a century now, and remains
a fundamental guide for competitiveness policy …
… But like all policy constructs, it requires constant
reflection on how it should adapt and evolve to fit with
today’s and tomorrow’s competitiveness challenges
Two types of
reflections are
particularly
important at this
moment in time How are the emergence of ‘smart specialisation
strategies’ affecting how we work with clusters? Do we
need to re-think clusters in for the RIS3 era?
How can we better evaluate what we are doing with
clusters so as to generate learning around what works,
in what contexts, and why?
1.
2.
5. An important distinction
Clusters
• A real socioeconomic
phenomenon
• Agglomeration in a
certain geographical area
of firms & other agents
engaged in related
activities, characterized
by the development of
cooperative relationships
Cluster Policies
• A set of coordinated
public and/or private
actions designed to
support the development
of relationships between
agents within existing
and/or emergent clusters
• These may be formal
and/or informal, and can
use a variety of
‘instruments’
Cluster Policy Instruments
• Specific tools used to
operationalise cluster
policies
• Support for individual
cluster associations is
one such tool
• Others include support
for projects in
cooperation and direct
management of various
clusters from a central
agency
Clusters can exist with or
without supporting policies
Cluster policies & instruments are the main focus
for evaluation & learning
Clusters and cluster ‘initiatives’ or ‘associations’ are not the same thing …
6. Need for government policy?
Scenario Policy Rationale
No existing
agglomeration
Questionable rationale for government
policy, except in very specific
circumstances (e.g. strategic positioning
where there are strong related clusters)
Existing
agglomeration but
weak institutional
elements
Government policy can help to maximise
cluster potential, but in a long term
process that fits the existing socio-
institutional context
Existing
agglomeration and
functioning
institutional
elements
Government policy may generate
‘additionality’ depending on the specific
socio-institutional context, but there are
greater dangers of crowding out
Three broad scenarios:
Context is critical
and there is an
important role for
evaluation processes
that enable policy-
makers to
understand the
evolution of context
as it interacts with
the policy
Source: Aragón el al (2014)
7. Cluster policy complexity
• The explosion of cluster policies
around the world over the last
20 years has generated critique
focussed on the impreciseness
or ‘messiness’ of the concept
and its indiscriminate adoption
as a policy panacea
• There is huge heterogeneity in
cluster policy approaches and
overlaps in the implementation
of cluster policies
• The Spanish cluster policy
landscape is a microcosm of the
global landscape
Significant
differences in
approach in
the 17
autonomous
communities
Elements of
cluster policy
also at national
& European
levels
Specific cluster
initiatives at
sub-regional or
city levels in
many cases
8. Clusters are part of policy systems
• Cluster policies overlap with many other competitiveness and innovation
policies in complex ‘policy systems’
• These policy systems are also clearly ‘multi-level’
Source: Magro & Wilson (2013)
9. Cluster evaluation is challenging
• Cluster policies interact in complex, multi-level policy systems
• Cluster policies typically contain a mix of tangible and intangible objectives
• The existing socio-economic structure (the context) both conditions networking
behaviour in clusters and is changed by it
• These characteristics make it very difficult to isolating cause-effect relationships
and determine the impact of cluster policy instruments
• It is therefore not easy to demonstrate the value of cluster policies
– We rely on a mix of empirical studies with limitations, case-based analysis and
anecdotal evidence
– There is a “pick and mix of research evidence” (Perry, 2005) that arguably re-
enforces clusters as a policy panacea
10. Choosing an evaluation approach
• Some form of logic model is often used to order indicators:
• But there is an important set of questions to ask first: Why do we want to
evaluate? Who is the audience? What are their needs?
• This will determine the evaluation criteria and guide the evaluation approach
– Concrete measures of Return on Investment (ROI)?
– Indicators that support learning among the cluster management and agents?
Summative evaluation:
measures outcomes for
accountability purposes
Formative evaluation:
Geared towards learning and
ongoing change
11. Impact, learning, or both?
• Impact evaluations typically seek to quantitatively measure outputs
(observable networking behaviour) and/or outcomes (firm-level productivity) to
demonstrate the impact of the cluster policy inputs:
– Statistical techniques (such as matched samples) enable us to disentangle the
impacts of specific policy instruments and arrive at ‘headline figures’ for impact
– Techniques such as social network analysis are also useful for analysing the
relationship between different elements of cluster policy in ‘policy systems’
• Yet these approaches suffer from two main limitations:
1. They rely on good data, which often is not available
2. They don’t tell us much about why or how the policy is (or isn’t) working
The key to overcoming the challenges of cluster evaluation is in using mixed methods
12. Participatory evaluation
Conventional Participative
Who? External experts. Beneficiaries, business people, policy-makers,
evaluation team.
What? Success criteria and information
necessities are pre-determined.
Evaluation by objectives.
Participants identify their own information
necessities and determine their own success
criteria.
How? Distance from the evaluation team
and other participants
Shared methods and results from the
involvement of participants
When? In general, when the policy or
programme is finished
Frequently, throughout the duration of the
policy. Continuous evaluation.
Why? Summative evaluation. Should the
policy or programme be
continued?
Formative evaluation to generate actions of
improvement. Continual learning.
Participatory evaluation is a form of ‘action research’ that can be
particularly powerful for evaluating cluster policies because its process has
synergies with the systemic nature of clusters themselves …
Source: Diez (2012)
13. Participatory evaluation in practice
Planning
• Literature review
• Development of
conceptual framework
• Interviews with policy-
makers & cluster
association
management
Application
• Participatory workshops
with cluster
stakeholders
• Definition of evaluation
framework & indicators
• Data collection
Reflection
• Analysis of results by
research team
• Results workshop with
stakeholders
• Sharing of project
experience with others
14. A social capital evaluation framework
Source: Aragón el al (2012)
The experience demonstrated how a deeper appreciation of the realities and behaviour of
targeted firms can support policy learning
15. A change in paradigm
• Evaluation is too often seen as a threat or challenge to policy because it is
associated narrowly with accountability
• We need to understand evaluation as learning
• It should be an integral part of the policy process, not something that is done
‘afterwards’ to justify activities
• The emergence of the smart specialisation paradigm is one such change, where
learning from and with clusters can make the difference between a realistic,
embedded and ‘live’ strategy and one that remains as a ‘plan’
Cluster evaluation can generate valuable learning that will help us to improve what
we are doing and to proactively embrace changes that might affect our activities
16. In a nutshell, smart specialisation is about placing greater emphasis on innovation
and having an innovation-driven development strategy in place that focuses on each
region’s strength and competitive advantage. It is about specialising in a smart way, i.e. based on
evidence and strategic intelligence about a region’s assets and the capability to learn what
specialisations can be developed in relation to those of other regions.
DG Regional Policy, European Commission (2011)
• The concept has been embraced and taken forward rapidly, and today all
European regions are required to have a RIS3 in place to receive structural funds
related to innovation
– These developments are supported by a Guide to RIS3 and an S3 Platform
• Strategies should emerge from an ‘entrepreneurial discovery process’, with priorities
identified in the nexus of economic activities, technologies and market opportunities
Smart specialisation strategies
17. Business
manufacturing and
services, primary
sectors, financial
sector, creative
industries, social sector, large
firms, SMEs,
young entrepreneurs, students
with business ideas, cluster and
business organisations, etc.
Research
public and private research
bodies, universities,
science and technology
parks, NCPs,
Technology transfer
offices, Horizon2020
committee members, regional
ESFRI roadmaps, etc.
Different departments,
if relevant at different
government levels, agencies
e.g. for regional development,
business advice,
public procurement offices,
incubators, etc.
Public administration
NGOs and citizens’
initiatives related to
societal challenges for
which innovative solutions
would be
helpful, consumers
associations,
Talents!, etc.
Civil society / Users
Entrepreneurial in Composition and
Spirit (risk-taking, broader view
beyond boundaries …)
Source: Katja Reppel / DG Regio
presentation, January 2014
Entrepreneurial Discovery Processes
18. Smart specialisation and clusters
• For many years clusters have sought to work beyond traditional
boundaries, and there are many synergies between the concepts:
– Both imply forms of cooperation between firms and other agents working in
related/complementary areas
– Both are systemic and require new forms of leadership & governance
– Both rely on place-specific assets, context and institutions
– Both seek to be transformative & require processes of prioritization
– Both are subject to debate about the appropriate role of government
– Both are characterised by challenges in evaluating their effectiveness
But pre-existing clusters & policies
embody important elements of
entrepreneurial discovery processes that
RIS3 can build from
Foray et al. (2012) argue that “vibrant
innovative clusters” are a “classic
outcome” or an “emergent property”
of a RIS3
19. Clusters & the six steps to RIS3
Steps to RIS3 design Contribution from clustering experience
Step 1: Analysis of regional context & innovation
potential
Strategy rooted in regional specificities
Looking beyond regional boundaries
Entrepreneurial dynamics: prospects for a
process of entrepreneurial discovery
Existence of cluster policy and functioning
cluster initiatives can provide a strong basis
for analysis and knowledge about regional
context, through for example existing
diagnostic processes within clusters, cluster
mapping exercises, and in-depth cluster
case analyses
Step 2: Governance to ensure participation &
ownership
“Quadruple helix”
Collaborative leadership
Boundary Spanners
Clusters themselves exhibit a long
experience with ensuring participation and
effective governance, and there is
significant potential to learn from and
improve these governance structures and
processes in the development of RIS3
Step 3: Elaboration of an overall vision for the
future of the region
Constructing the vision: scenarios…
Communicating the vision
The strategic reflection processes of
existing clusters can provide lessons in
constructing common vision, and the
clusters themselves are important vehicles
for construction and communication of a
regional vision
20. Steps to RIS3 design Contribution from clustering experience
Step 4: Identification of priorities
Combine top-down and bottom up approach
Vertical and horizontal type priorities
Inter-cluster approaches and collaboration
among and between KET actors and
clusters can play an important role in
facilitating the coordination of bottom-up
and top-down input into prioritization
processes
Step 5: Definition of coherent policy mix,
roadmaps and action plan
Experimentation possibilities
Cluster policies have followed a similar
path, and experience shows the
importance of policy flexibility and
mechanisms to ensure sophisticated policy
intelligence
Step 6: Integration of monitoring and evaluation
mechanisms
Monitoring to follow the process of
experimentation
Evolve and adjust according to changes in
economic and framework conditions
Experience with cluster policy evaluation
suggests the importance of mixed
methodologies and a policy learning focus
Clusters & the six steps to RIS3
Source: Aranguren & Wilson (2013)
21. Re-thinking clusters in the RIS3 era?
• Boundaries between clusters are becoming more porous as new
opportunities emerge in the nexus of economic activities, technologies
and market opportunities
• Inter-cluster collaboration takes on a new importance in this context
– Within region
– Across regions (a dimension often neglected in RIS3)
• Rigid clusters can present barriers to the cross-fertilisation of ideas
• In some cases cluster policies may need to adapt and be more flexible
Above all RIS3 and future cluster development should go hand-in-hand, embracing each other
in a context of learning and improvement and building from their synergies
22. Closing Message
• Cluster policies, like clusters themselves, have lifecycles
• The competiveness environment is constantly changing, requiring renewal
• The emergence of RIS3 reflects this renewal process and presents
opportunities for clusters
• Overcoming the fears and challenges of evaluation has a key role to play in
renewal processes, generating learning and ensuring that clusters are able to
adapt, improve and continue to add value to regional competitiveness
23. WHERE NEXT FOR CLUSTERS AND CLUSTER POLICY?
James Wilson
Orkestra and Deusto Business School
jwilson@orkestra.deusto.es jamierwilson
THANK YOU
Keynote, IV Clusters, Cluster Policies & European Networking Conference
Santiago de Compostela, 7th May 2014