Idea Transcript
Ref. Ares(2017)4012141 - 11/08/2017
European Cluster Observatory
Regional Ecosystem Scoreboard Updated Methodology Report Prepared by: Lorena Rivera León, Kincső Izsak, Paresa Markianidou, Kastalie Bougas, Technopolis Group Asier Murciego Alonso (Orkestra)
April 2017
Internal Market, Industry, Entrepreneurship and SMEs
Regional Ecosystem Scoreboard – Methodology Report
European Cluster Observatory in brief The European Cluster Observatory is a single access point for statistical information, analysis and mapping of clusters and cluster policy in Europe. It is primarily aimed at European, national, regional and local policy-makers and cluster managers and representatives of SME intermediaries. It is an initiative run by the ‘Clusters, Social Economy and Entrepreneurship’ unit of the European Commission’s DirectorateGeneral for Internal Market, Industry, Entrepreneurship and SMEs and aims to promote the development of more world-class clusters in Europe, notably with a view to promoting competitiveness and entrepreneurship in emerging industries and facilitating SMEs’ access to clusters and internationalisation activities through clusters. The ultimate objective is to help Member States and regions to design smart specialisation and cluster strategies that will help companies to develop new, globally competitive advantages in emerging industries through clusters, and in this way to strengthen the role of cluster policies in boosting Europe’s industry as part of the Europe 2020 Strategy. In order to support evidence-based policy-making and partnering, the European Cluster Observatory provides an EU-wide comparative cluster mapping with sectoral and cross-sectoral statistical analysis of the geographical concentration of economic activities and performance. The European Cluster Observatory provides the following services: ■
a biannual ‘European Cluster Panorama’ (cluster mapping) providing an update of and extension to the statistical mapping of clusters in Europe, including for ten related sectors (i.e. cross-sectoral) and a correlation analysis with key competitiveness indicators;
■
a ‘European Cluster Trends’ report analysing cross-sectoral clustering trends, cluster internationalisation and global mega trends in industrial transformation; identifying common interaction spaces; and providing a forecast for industrial and cluster opportunities;
■
a ‘Regional Ecosystem Scoreboard’ setting out strengths and weaknesses of regional and national ecosystems for clusters, and identifying cluster-specific framework conditions for three cross-sectoral collaboration areas;
■
a ‘European Stress Test for Cluster Policy’, including a self-assessment tool accompanied by policy guidance for developing cluster policies in support of emerging industries;
■
a showcase of modern cluster policy practice, provided in the form of advisory support services to six selected model demonstrator regions. The services offered include expert analysis, regional survey and benchmarking reports, peer review meetings and policy briefings in support of emerging industries. The policy advice also builds on the policy lessons from related initiatives in the area of emerging industries;
■
the European Cluster Conferences 2014 and 2016, which bring together Europe’s cluster policy-makers and stakeholders for a high-level cluster policy dialogue and policy learning, and facilitate exchange of information through, e.g. webpages, newsletters and videos.
More information about the European Cluster Observatory is available at the EU cluster portal at: http://ec.europa.eu/growth/smes/cluster/observatory/.
Regional Ecosystem Scoreboard – Methodology Report
Table of Contents European Cluster Observatory in brief ............................................................................................... 2 Introduction, objectives and scope ..................................................................................................... 1 1. Capturing the Quality of Conditions of Regional Ecosystems for Innovation and Entrepreneurship .................................................................................................................................. 2 1.1
What is a regional ecosystem? ........................................................................................... 2
1.2
What is the role of clusters in regional ecosystems? .......................................................... 3
1.3
What are favourable regional framework conditions? ......................................................... 4
1.4
Existing initiatives aiming at capturing regional entrepreneurship, innovation and competitiveness: state-of-the-art ........................................................................................ 7
2.
The Conceptual Framework ................................................................................................ 10 2.1
Key dimensions of the regional ecosystem ...................................................................... 12
2.2
Regional peer groups ....................................................................................................... 15
3.
The Measurement Framework ............................................................................................ 16 3.1
The Regional Ecosystem Indicators ................................................................................. 16
3.2
Collection of primary data: survey with cluster managers and regional development agencies ........................................................................................................................... 20
3.3
Imputation of missing data ................................................................................................ 21
3.4
Composite indicators ........................................................................................................ 23
3.5
Sensitivity Analysis ........................................................................................................... 24
References ........................................................................................................................................... 40 Annex 1: Indicator framework ........................................................................................................... 42 Annex 2: Cluster managers survey questionnaire .......................................................................... 48 Annex 3: Regional Development agencies survey questionnaire .................................................. 50
Regional Ecosystem Scoreboard – Methodology report
Introduction, objectives and scope Innovation and entrepreneurship thrives in particular contexts and under particular framework conditions, and are nurtured by the interaction between actors with different resources and capabilities such as firms, users, research organisations, investors, business support providers, and public institutions as it is widely understood. Key framework conditions that determine the nature of regional innovation and business ecosystems are for instance the nature of market demand, the excellence of academic systems, strengths of available skills, access to finance or aligning different intellectual property right protection approaches. The success of regional economic development and innovation policies depends on their ability to foster industrial competitiveness, to create conditions that can give the location a lasting advantage, and to strengthen the resilience of the system so that it stays flexible and adaptive through structural change also in times of external shocks. An important feature of modern innovation policies is that they do not only focus on rigid industrial activities or narrowly defined industrial clusters, but take into account the opportunities that span across industries and business activities. The Scoreboard is designed primarily for policy-makers responsible for regional, industrial and cluster policies as a tool to identify the bottlenecks of the wider regional eco-system that could be targeted through policies as they may hold back overall performance. It gives regions an idea how their regional ecosystem is positioned compared to other regions and businesses an idea in which regional ecosystem to locate best their operations. Target group: The Regional Ecosystem Scoreboard is designed primarily for policy-makers responsible for regional, industrial and cluster policies. Objectives: The objective of the Regional Ecosystem Scoreboard is to identify, describe and capture the quality of conditions in the regional ecosystem that can foster or eventually hinder the creation of dynamic cross-sectoral collaboration spaces for innovation and entrepreneurship revealing both enabling and constraining mechanisms. It has a double feature of providing policy-makers with insights first about general framework conditions for innovation and entrepreneurship in emerging industries and secondly theme-specific framework conditions that are specifically relevant for certain types of industries and clusters. In this way the emphasis of the Scoreboard is on the Conditions and on the Dynamics that characterise the quality and nature of the regional ecosystem and it is not about measuring performance as in the case of other related regional scoreboards. The better understanding of these conditions and cross-sectoral dynamics will aim at giving policy-makers a comparative tool to design and launch more targeted and evidence-based policy measures in the next years of 2016-2020. Geographical coverage: The Scoreboard covers NUTS2 regions, respectively NUTS 1 regions in the case of Belgium, Germany and the United Kingdom within the 28 European Union Member States.
The authors would like to acknowledge significant contributions in preparation of this report from Christian Ketels (Stockholm School of Economics), Susana Franco (Orkestra) and Michal Miedzinski (Technopolis Group). We also thank the participants of the Methodological Workshop who provided relevant input for this paper: Alberto Bramanti, Alberto Pezzi, Markus Grillitsch, Hugo Hollanders, Irma Priedl, James Derbyshire, Karen Maguire, Richard Lewney, Robert Huggins, Dorota Weziak-Bialowlska, Grzegorz Gorzelak and Laszlo Szerb. We are also grateful to Carsten Schierenbeck and Domenico Gullo from the European Commission for their valuable comments.
1
Regional Ecosystem Scoreboard – Methodology Report
1. Capturing the Quality of Conditions of Regional Ecosystems for Innovation and Entrepreneurship The objective of this Chapter is to review the literature related to the concepts of regional ecosystems, favourable framework conditions, clusters and cluster-specific framework conditions which provides the theoretical background for the Regional Ecosystem Scoreboard, and to summarise the key features of other existing scoreboards relevant for the topic.
1.1 What is a regional ecosystem? The concept of ecosystems is now often used in business and innovation contexts. Originally a concept from biology, an ecosystem is defined as a set of relationships between living and non-living organisms – companies, individuals etc. - whose functional goal is to maintain an equilibrium sustaining state (Jackson, 2011). Ecosystems also evolve naturally and without a specific plan. Ecosystems are complex, dynamic, adaptive, emergent systems where the same inputs do not always produce the same outputs, where the behaviour of a system is not the aggregation of individual parts, where disruptions and emergence happen, and where effects occur in far-from equilibrium states (Gooble, 2014). When applied to the innovation context, Innovation Ecosystems are defined as dynamic, purposive communities with strong relationships based on collaboration, trust and co-creation of value and sharing complementary technologies or competencies (Durst and Poutanen, 2013). Innovation Ecosystems are usually created around to central node - technology platform, social or economic conditionsthat put key agents together to interact. The Innovation Ecosystem idea has also been evolved towards several levels of organisation (Gooble, 2014). National and regional Innovation ecosystems are now seen as a way to boost economies (Lawlor, 2014). Three ecosystem archetypes has been identified in the literature (Moss Kanter, 2012): thinker ecosystems, where resources are drawn around new ideas; maker ecosystems, that act as manufacturing hubs; and trader ecosystems, which are financial and logistical hubs for a region. Regional Ecosystems include all of the organisations—companies, universities, entrepreneurs, customers, regulatory agencies and municipal or regional governments— and when well articulated, they create a dynamic, regional economy. A notion that gained recently much popularity related to regional innovation ecosystems is resilience. Regional innovation ecosystems as any system are prone to external shocks, endogenous breakthroughs or political changes (Martin and Sunley, 2013) that can negatively influence the regional development path and make certain industries or the entire regional economy to decline leaving behind despaired socio-economic conditions. The word resilience refers to the characteristics of any system that bounces back against shocks instead of being hit negatively in the long-term. Holling (1973) described the notion as the pace at which a system returns to equilibrium after a disturbance. An important element of resilience is the capacity to adapt and renew, and even pro-acting to external changes that otherwise could have a negative impact. Martin et al (2013) defines regional economic resilience as “the capacity of a regional or local economy to withstand or recover from market, competitive and environmental shocks to its developmental growth path, if necessary by undergoing adaptive changes to its economic structures and its social and institutional arrangements, so as to maintain or restore its previous developmental path, or transit to a new sustainable path characterised by a fuller and more productive use of its physical, human and environmental resources”. As the literature describes, the innovation and business ecosystem of
2
Regional Ecosystem Scoreboard – Methodology Report
resilient regions is favourable for structural change, for the support of emerging industrial activities, for cross-sectoral dynamics and for the development and transformation of clusters.
1.2
What is the role of clusters in regional ecosystems?
Clusters – as economic phenomena of geographic concentration of related industrial activities - are instrumental to foster the critical connections and interactions in regional innovation ecosystems. As it is well understood the co-location of firms positively affects both economic and innovation performance of countries and regions (Porter, 1990; Sölvell et al, 2000). In this report clusters are understood as being “geographic concentrations of interconnected companies, specialised suppliers, service providers, firms in related industries, and associated institutions in particular fields that compete but also cooperate” - (Porter, 1998). A clear distinction is made between ‘clusters’ as the phenomenon and ‘cluster initiatives’ or ‘cluster organisations’ that represent deliberative, often politically driven, endeavours to support national and regional strongholds. Clusters form a natural part of any regional business and innovation eco-systems even if the specialisation of regions and the strengths of clusters can be different. The relationship between the regional ecosystem and clusters is two-way. The regional ecosystem is as important for the success of clusters as the presence and dynamics of clusters influence the regional business environment. This dynamics is linked to the creation of new firms, to the rate of rapidly growing businesses and the reinvention of mature companies (Teece, 2015). To understand cluster growth, the specificities of the knowledge infrastructure, institutional set-up or policy actions should be considered. Nevertheless ecosystems where innovative clusters are present and there are a related variety of industries provide also better conditions for the emergence of new industrial activity (Trippl et al, 2014). Clusters thrive under specific business framework conditions that can result for instance from the innovation culture or from actions of national and regional institutions (Lundvall, 1988; Lundvall, 1993; Edquist, 1997; Cooke and Morgan 1998; Ketels et al; 2008). The contrary is also true; clusters can decline due to specific failures in the economic system, in policy or because of a lack of fresh entrepreneurship based on which they can renew themselves responding or even pro-acting to external pressures. Clusters can differ in their focus, their size, maturity, level of interactions, openness or their key drivers, nevertheless they can be categorised in broader groups according to pre-defined criteria. The most straightforward grouping of clusters is according to cluster categories. For instance the cluster definitions applied in the US context propose 51 traded cluster categories while treating the rest of the economy as local. Traded clusters capture those industries that are serving markets beyond their own location and that are fully exposed to competition from other locations. Traded clusters concentrate across regions; their high wages and high levels of innovative activity make them the key engines of regional economies. Local clusters combine those industries that operate mostly only locally (such as local retail and other local services), and that are present in similar density across all regions (i.e. that are evenly spread and not clustered); their high employment numbers make them an important channel for creating shared prosperity in regions (Ketels and Protsiv, 2014). Nevertheless there is no perfect cluster categorisation as the boundaries of clusters are in constant dynamics, hence real-time industry and cluster categories are just somewhat identifiable using standard classification systems (European Cluster Trends, 2015). One of the most often used typology of clusters is the one developed by Markusen (1996), who distinguished four models based on the role of the cluster members and their interactions. ■
Marshallian cluster model – In this model, clusters are homogenous and made up by small firms who both collaborate and compete, but where none of the firms has power to control the cluster. 3
Regional Ecosystem Scoreboard – Methodology Report
■
Hub and spoke cluster model – In this model, a few firms dominate, around which other smaller companies are linked. Examples are for instance cluster in the automotive industry.
■
Satellite platform model – In this model, some multinational companies and their local branches are influencing the development of the region and provide opportunities for local suppliers, but where the regional companies are not necessary connected with each other.
■
State anchored cluster model – In this model, the government’s endeavours influence the region and fosters or sometimes impedes the economic relation between the members of the cluster.
Another typology of clusters which has been also widely used for differentiating between different types of cluster policies are the one adopted in France by DGCIS (2009): ■
Industry-driven: In such clusters, there is a presence of large, international groups of companies who are investing in R&D. SMEs are connected to these companies as suppliers or local drivers of knowledge. Examples include for instance: BioValley Basel (Switzerland); CARS in Stuttgart (Germany)
■
R&D-driven: Such clusters are driven by a university or research centre around which companies, start-ups, spin-offs, larger companies are gathered. There is a strong and dynamic entrepreneurial activity ongoing. Technological parks, science campuses can play an important role. Examples include: Cambridge Silicon Fen (United Kingdom); Tel Aviv (Israel)
■
Collaboration-driven: Such clusters are characterised by intensive collaboration between public and private sectors. The public sector plays an active role in fostering cluster development. Entrepreneurship is supported by investment funds of large industrial groups present at the local level. Examples include: Medicon Valley (Denmark and Sweden); Silicon Saxony (Germany)
Despite of the usefulness of such categorisations as presented above, some authors point to the fact that in reality, clusters do not feature just one single type of typology but are more of a mix (He and Fallah, 2011). Differences between clusters are also very much influenced by the specific industry that the cluster represents, hence relying on industry classifications can be more useful. Moreover, one has to keep in mind that clusters are not static but are in constant dynamics and thus their typology also changes over time and in different stages of the lifecycle.
1.3 What are favourable regional framework conditions? Favourable framework conditions can positively influence the evolution of regional business ecosystems and facilitate the creation of new enterprises, the survival and growth of business activities. The existence of a favourable business environment provides the basis for developing strength and resilience in regions. Several studies have identified the importance of framework conditions for economic success. At regional level, the most commonly studied areas include: ■
External economies for the localisation of production systems;
■
The intensity of competition;
■
Regional cultures and local institutional fabrics, or “institutional thickness” (MacLeod, 1997);
■
Locational proximity and trust.
4
Regional Ecosystem Scoreboard – Methodology Report
All these studies describe the key dimensions of favourable framework conditions and often establish a model to capture the quality of conditions. Michal Porter’s diamond model (1990) comprised of four key elements of competitiveness provides one starting point to analyse framework conditions of regional ecosystems. The model’s four elements refer to demand conditions, factor conditions, related and supporting industries, and firm strategy, structure and rivalry. Demand conditions describe the level of domestic demand depending both on the quantity of demand as well as the sophistication level of consumers. Factor conditions include any factors of production that a firm uses in its businesses such as land, labour, capital, raw materials or basic infrastructure. Related and supporting industries encompass suppliers or distributors that serve the industry. Firm strategy, structure and rivalry describe the types of actions firms rely on in order to achieve their goals such as implementing low-cost, differentiation or focus strategies. The concept of institutional thickness refers to social and cultural factors that determine the economic success of regions, including a strong institutional presence; the quality of infrastructure; the existence of effective product and labour market regulations; property rights; industrial disputes; high levels of interaction among regional stakeholders; and local industrial embeddedness. The regulatory environment, specifically in relation to product market competition, is of particular importance, as it can drive or impede entrepreneurship and/or trade and investment. The macro-economic framework at national level also shapes the regional framework conditions. In fact, evidence (Mataloni, 2007) shows that firms follow a sequential decision process when choosing where to locate their businesses, starting first at country level (or group of neighbouring countries) based on a set of location factors; to then select a smaller geographic area (region or city) based on another set of location factors. Key macro-economic conditions affecting the regional level include unemployment rates, inflation, and public debt. At regional level, population size, per-capita income, land-labour ratio, wage costs, market size, geographic distance, transportation infrastructure, average education, worker skill levels, and the presence of multi-national enterprises (MNEs) in the region are important location factors often discussed in the literature. Florida (2002) has suggested the importance of a more human-centred approach to location, or having a favourable and effective people climate. For him, this is provided through “plug-and-play communities”, or cities/regions that have low entry barriers for people, where newcomers are accepted quickly into all sorts of social and economic arrangements. These regions have usually more diversity, higher levels of “quality-of-place”, and provide active (outdoor) participatory recreation. The rationale is that these are places where people can find opportunities and build support structures quickly. These creative centres usually have thick labour markets, providing many employment opportunities for people. Other studies highlight the importance of an agile entrepreneurial community where failure is also accepted and learning from mistakes are encouraged, they stress the need for a highly educated workforce, access to university research, favourable labour laws, existence of start-up programmes, good quality of life and the free flow of information (Teece, 2015). In summary the following key dimensions can be highlighted that are most often mentioned by the related literature: ■
The presence of a strong entrepreneurial culture: entrepreneurship is seen as a key ingredient of regional ecosystems that keeps industrial dynamics in motion. Entrepreneurs form a critical element both in the formation and transformation of regional industries;
5
Regional Ecosystem Scoreboard – Methodology Report
■
Knowledge linkages within the region are necessary by nurturing both fruitful internal linkages and external connections so that regions remain open for new ideas and do not fall in the trap of a regional lock-in. Combining regional innovation opportunities with global linkages has been a positive characteristic of many successful regional economies (Ketels and Sölvell, 2006; Ketels and Memedovic, 2008);
■
The availability of specialised high-quality inputs, the most important of which is the availability of highly skilled human resources and related infrastructures;
■
The availability of supportive financing and funding conditions such as favourable bank loans, public grants, monetary funding schemes, venture capital, crowd-funding opportunities or a network of business angels;
■
The creation of new ventures is very much supported by the presence of strong social capital and business networks that bind the innovation actors together and can trigger positive regional dynamics (Verspagen, 1998);
■
The existence of demand conditions, that serve as motors to the system, and that includes consumers (enterprises and individuals) of the products and services being produced;
■
The importance of having a dynamic local context and support structures that favours investment, upgrading (of processes and products), and a business policy that promotes competition-cooperation-and knowledge sharing;
■
Change is often triggered by new cross-sectoral business combinations or by reconfigurations on-going in industrial leadership patterns. A favourable regional ecosystem is not focused on rigid, clearly identified clusters, but supports dynamic processes so that industries and actors can constantly evolve and renew themselves.
Framework conditions can be general that are valid across all industries and clusters but are often cluster-specific or industry-specific, where only a set of them are the ones which are instrumental in fostering the development and emergence of a certain type of cluster. For instance in the evolution of science-based clusters different factors play a decisive role as in the development of clusters composed of small manufacturing firms and again others in new emerging, services-based clusters. Partially reflecting the differences in relation to strategies, location factors also vary by industry. For instance, according to the World Investments Prospectus Survey (UNCTAD, 2007), the size and growth of the market are the most important location factors for manufacturing and services; whereas access to natural resources and government regulation are very important for the primary sector. Even within the manufacturing sector, there are also marked differences. For instance, the automotive industry gives high importance to labour costs and labour productivity; while the quality of labour is very important for industries such as pharmaceuticals, equipment, machinery and chemicals. In the case of the services industries, access to markets and the openness for foreign investments are of top-priority for location. The importance of location factors also differs between industries, the functional activities, and the strategic motivations of the private sector. Dunning (2000) and Faeth (2009) distinguish that marketseeking companies naturally emphasise market-related location factors such as market size, market growth potential, and consumers’ buying power; resource-seeking companies give emphasis on locations that have relatively cheap and abundant scarce resources (natural, physical or human); efficiency-seeking companies are more interested in regions that can offer cost advantages to achieve economics of scale; while strategic asset-seeking companies aim to gain access to technology and productive capabilities in regions where they locate.
6
Regional Ecosystem Scoreboard – Methodology Report
Key framework conditions that influence the chances of cluster development can be the nature of market demand, the excellence of academic systems, strengths of available skills, access to finance or aligning different intellectual property right protection approaches. Other literature stresses the importance of the access and use of human resources, dynamism from new companies, the availability of venture capital or communication infrastructure. The role of rivalry and competition must be also highlighted. As already Porter noted (1998), it is often the competition between rival firms in the cluster that drives growth because it forces firms to be innovative.
1.4 Existing initiatives aiming at capturing regional entrepreneurship, innovation and competitiveness: state-of-the-art Although contributions to measure the regional eco-systems as well as innovation systems are numerous, each of them pursues its own overall approach with its corresponding dimensions. Their main objectives can be outlined into four categories: the measurement of innovation performance; the focus on sectors, industries and clusters; the focus on entrepreneurship and the focus on territorial competitiveness. The first category of studies relies on the measurement of innovation performance, in qualitative and quantitative ways. The Regional Innovation Monitor Plus produces indicators to measure innovation performance along with related baseline regional profiles, events, news, organisations, policy documents and support measures and aspires to share knowledge on innovation systems and policies, while the Regional Innovation Scoreboard mainly aims at providing a comparative assessment of the innovation performance across different EU regions and provides statistical facts on the matter. A second category of contributions focuses on sectors, industries or clusters’ analysis and provides mostly sectorial insights. The US Cluster Mapping presents data on regional clusters and economies to support businesses, innovation and policy in the United States, through interactive tools and robust data. The European Cluster Observatory explores and captures cross-sectorial dynamics in a first phase and then provides a mapping of these clusters in a second phase. The Observatory tackles the topics of clusters, clusters organisations and regional microeconomic framework conditions across Europe. The European Cluster Excellence Scoreboard measures the regional strength of three selected emerging industries, namely creative industries, eco industries and mobile services and the European Service Innovation Scoreboard measures the extent to which service innovation can impact sectors and thereby contribute to structural changes within regions, i.e. the transformative power of service innovation. The third category grasps the stakes behind entrepreneurship’s effects and developments. The Global Entrepreneurship Monitor (GEM) seeks to address the link between entrepreneurship and economic development through its conceptual model connecting entrepreneurship and economic growth. In this perspective, the GEM highlights country differences, indicates drivers to adequate levels of entrepreneurship and provides policy advice on national entrepreneurial activity. The Flash Eurobarometer Survey provides insights on the development of entrepreneurship and allows for a comparative assessment of the entrepreneurship level between EU Member States as well as with non-EU countries. As Acs et al (2009) points out one shortcoming of entrepreneurship indicators that they often capture the quantity but not the quality of entrepreneurial activity, such as opportunity recognition, skills, creativity, or innovation and high growth. There have been nevertheless several other indicators constructed such as high growth potential, business discontinuation and the environmental factors of entrepreneurship perception that can make up for this shortcoming. The OECD Entrepreneurship Index consists in a methodological tool and in an analytical model providing a collection of indicators on entrepreneurship. Eventually, the World Bank’ Entrepreneurship Database essentially focuses on a specific dimension of entrepreneurship, by providing data on new business registration and new business entry density, in order to bring out the dynamics and patterns of private enterprises. The last category of scoreboard encompasses the Regional Competitiveness Index, and focuses on 7
Regional Ecosystem Scoreboard – Methodology Report
territorial competitiveness. This last publication aims at discovering the overall strengths and weaknesses of EU regions. The overarching differences in approaches of these contributions rely on the various dimensions they rely on to support their analysis, whether at the macroeconomic and/or microeconomic levels. The US Cluster Mapping has the particularity to be essentially based on microeconomic dimensions: it dissociates traded clusters – which serves markets in other regions/countries – from local clusters – which sell in a local market and are present in all regions – and bases its analysis according these two features for dimensions like employment/unemployment, job creation, income/wages, prosperity, etc. Similarly, the Regional Competitiveness Index mostly relies on macroeconomic dimensions, including macroeconomic stability, wider framework conditions, health, infrastructure, labour market efficiency and market size, etc., but is however supplemented with micro dimensions such as business sophistication, technological readiness, etc. At the other end of the scale, the Flash Eurobarometer survey only explores firm-level dimensions on employment status (e.g. feasibility and desire to be selfemployed), starting up a business (e.g. experiences and key considerations), etc. Nevertheless, most studies provide a mix of macroeconomic and microeconomic dimensions, as it is the case for the European Cluster Excellence Scoreboard, the Global Entrepreneurship Monitor, the Regional Innovation Survey, the Regional Innovation Monitor Plus, the European Service Innovation Scoreboard, the Cluster Observatory and the OECD-entrepreneurship Index. A common set of dimensions can be observed throughout the different studies. Most publications indeed take into account the financing of innovation/growth and access to finance, which deals with the presence of actors to provide funds and the availability of financial resources (equity and debt); the knowledge creation and transfer, which considers how national or regional R&D could possibly lead to commercial opportunities within firms; the entrepreneurship culture, activities and capabilities, on the extent to which norms can practically encourage the launch of new business methods and/or activities, and on the entrepreneurial efforts; the business support services, focusing on the presence of measures to stimulate training, advice, internationalisation and to foster collaboration and networking; and eventually an innovation dimension, coping with technology readiness, business sophistication and specialisation. On a more macroeconomic level, the most recurrent dimensions are the regulatory framework/government policies and programs, providing indicators on the presence of favourable regulatory environment and on the size-neutral or SMEs-oriented aspect of regulations; the market dimension, which includes insights on market size, critical mass of consumers, etc and the education/training dimension, related to the quality of education and lifelong learning. Table 1: Summary of existing initiatives aiming at capturing regional entrepreneurship, innovation and competitiveness Category
Scoreboard
Description
Coverage
Comparative assessment of innovation performance among EU member states based on regional statistical facts.
Regional (190 regions out of 22 EU countries + Switzerland and Norway)
Platform providing a series of innovation performance indicators in order to share knowledge on policy trends.
Regional (30 EU regions)
Interactive platform providing regional data on clusters and business environments.
Regional (only US regions)
Providing data and analysis of clusters, cluster organisations and regional microeconomic framework conditions
Regional EU28 + Norway)
Focus on sectors, industries and clusters
Measuring innovation performance
Organisation Regional Innovation Scoreboard European Commission – DG GROW Regional Innovation Monitor Plus European Commission – DG GROW US Cluster Mapping Harvard Business School’s Institute for Strategy and Competitiveness and the US Department of Commerce European Cluster Observatory European Commission – DG GROW
8
(Regions from Switzerland and
Regional Ecosystem Scoreboard – Methodology Report
Category
Scoreboard
Description
Coverage
Measurement of the transformative power of service innovation on regions and clusters.
Regional (280 regions out of EU28 + Iceland, FYROM, Norway, Serbia, Switzerland and Turkey)
Measurement of four broad markets (B2B, B2C, B2G, G2C)
National (pilot 8 countries)
Focus on the link between entrepreneurship and economic development; measurement of differences in the level of entrepreneurial activity among countries.
National (85 countries)
Comparative assessment of entrepreneurship development among EU countries and with non-EU countries.
National (all EU28 countries + BRICS, Japan, Norway, US, Israel, Iceland and South Korea)
Collection of indicators on entrepreneurship, harmonized on international level.
National (OECD countries)
Source providing comparable crosscountry data on new business registration in order to cope with dynamics of private companies.
National (worldwide coverage)
Overview of the territorial competitiveness at a regional level; emphasis on regions’ strengths and weaknesses.
Regional (270 regions out of EU28 countries)
Organisation European Service Scoreboard
Innovation
European Commission – DG GROW Demand-side Innovation Scoreboard European Commission – DG GROW Global Entrepreneurship Monitor Global Entrepreneurship Research Association Flash Eurobarometer Survey
Focus on territorial competitiveness
Focus on entrepreneurship
European Commission – DG GROW OECD Entrepreneurship Index The Organisation for Economic Cooperation and Development World Bank’s Entrepreneurship Database The World Bank Regional Competitiveness Index European Commission – Joint Research Centre
9
Regional Ecosystem Scoreboard – Methodology Report
2. The Conceptual Framework The objective of this Chapter is to present the Conceptual Framework of the Regional Ecosystem Scoreboard that is based on the literature review presented above and further develops the dimensions of favourable general framework conditions for the purposes of the Scoreboard. The Regional Ecosystem Scoreboard consists of 50 indicators and composite indicators comprises across six key dimensions, namely Knowledge basis and skills, Collaboration and internationalisation, Access to finance, Demand conditions, Entrepreneurial conditions and the Quality of governance. These are further divided into seventeen sub-dimensions that determine the quality of conditions of regional ecosystems (See Figure 1 and Figure 2). Figure 1: Conceptual model: dimensions of region-specific framework conditions
Key dimensions of the regional ecosystem
Knowledge basis and skills
Collaboration and internationalisation
Entrepreneurial+ condi/ons+
!
!
Access to finance
Demand conditions
Quality of governance Source: Technopolis Group
10
Regional Ecosystem Scoreboard – Methodology Report
Figure 2: Sub-dimensions of the Regional Ecosystem Scoreboard
First of all in terms of the users’ experience with the online Regional Ecosystem Scoreboard the following steps are to be highlighted: 1. At the Scoreboard’s interface the user will be asked to provide information about the name of its region, and the peer group that it would like to compare itself, besides the automatic peer group categorisation of the regions according to similar industrial structure and level of development. 2. After providing the requested information, the user will see first of all its performance on the six regional ecosystem dimensions, sub-dimensions and indicators (presented in Section 2.2. in detail). The performance of the user’s region will be compared to the average and the best score of all the regions in its peer group (automatically uploaded according to similar industrial structure and level of development) and the user can also select a comparison to self-identified peer group of regions. We also provide below a specific example: Let’s assume that the region that would wish to use the Regional Ecosystem Scoreboard is Catalonia from Spain. The objective of the regional policy-maker is to see what are the dimensions and subdimensions in which it performs well or less well. The values of Catalonia on the specific indicators and sub-dimensions are compared to the ■
Average of regions in the same peer;
■
Best score of the region in the same peer group;
■
Average of any custom-made peer group that the policy-maker selected at the front page of the Scoreboard.
The policy-maker will be able to see if for instance the regional conditions concerning entrepreneurship are good, but there is a missing dynamism in cross-sectoral linkages and more would need to be
11
Regional Ecosystem Scoreboard – Methodology Report
done on opening up the region to external linkages. The indicators and results can then be translated into arguments for specific policy measures.
2.1 Key dimensions of the regional ecosystem The core of the Conceptual Framework builds upon six key dimensions, including the horizontal dimension of the quality of governance, and seventeen sub-dimensions that determine the quality of conditions of regional ecosystems aiming at fostering regional innovation-driven entrepreneurship and structural change. These are built over a total of 58 individual indicators. The selection of the dimensions is based on the review of the literature and an analysis of available indicators that can meaningfully reflect the conditions of the ecosystem that have an impact on business and industrial dynamics. All favourable regional ecosystems certainly need people and money, hence the quality of the knowledge basis and skills and the availability of access to finance are two primary dimensions that we cannot disregard. It is also well understood that even if these two dimensions are of high-quality innovation-driven entrepreneurship can be hampered by the lack of collaboration between the system actors and can be held back by insufficiently harnessing the potential in internationalisation and global knowledge flows. A further key condition is the nature of the demand including both private and public demand that can stimulate and reward innovative activities. Entrepreneurship is very much influenced by the entrepreneurial conditions of the institutional framework for doing business and by the culture that can foster or act as a barrier for innovative-minded people. Last but not least the quality of governance will be also taken into account that has an impact on the business environment and possibilities for entrepreneurial activities. Below we present each dimension in detail.
Dimension of Knowledge basis and skills For a long time it has been claimed and proved that high levels of human capital lead to increased productivity as well as wages. A recent OECD report on skills and jobs (OECD, 2012) highlighted the increasing importance of skills in the knowledge-based global society, in particular skills that are dynamic and that are able to evolve with labour markets and that are continuously developed throughout life. Moreover, in order for the private sector to become more productive and competitive, the educational and vocational training system should prioritise the needs of enterprises. Vocational education in particular helps improving labour market skills and addresses important needs such as the upgrading of skills of students and employees and the re-skilling of the unemployed population. Lifelong learning is also of importance in particular for regions with high rates of adult population. This dimension of the Regional Ecosystem Scoreboard is linked and contributes to the entrepreneurship dimension. However, the main focus of the knowledge basis and skills dimension is on (higher) education, human capital development and training, and the availability of skilled human resources in the region.
Dimension of Access to Finance Access to finance is probably the most significant challenge for entrepreneurs as well as for the creation, survival and growth of enterprises. Access to finance has been more exacerbated by the global and financial crisis and the credit crunch, particularly hitting SMEs, and new and high-growth firms. The OECD calls this phenomena the “(SME) financing gap”, or the situation where a sizeable share of enterprises, notably SMEs, cannot obtain financing from banks, capital markets or other suppliers of finance; and thus entrepreneurs that have the capability to use funds productively if they were available do not have access to these funds.
12
Regional Ecosystem Scoreboard – Methodology Report
Moreover, not only the amount of funding has been more limited in the last years but also the conditions for credit have worsened (i.e. shortened maturities, requests for collaterals and guarantees). This has affected the level of cash flow available in the private sector, and has forced the weakest of enterprises into bankruptcy, ultimately raising unemployment levels in the less resilient regions and countries. This dimension of the Regional Ecosystem Scoreboard will look at issues related to the availability of private and public financing; as well as the existence of a conducting framework that enables the access to finance by enterprises.
Dimension of Collaboration and internationalisation Knowledge transfer reflects the connections and collaboration between universities, businesses and the public sector. Knowledge transfer can happen through many ways. The most straightforward form is through the mobility of researchers, innovators and entrepreneurs in the regional or broader institutional fabric. Universities and research centres can be involved in projects financed by the private sector and as such deliver technology and knowledge. Besides to this they can also participate in join collaborative research projects. Academics can become entrepreneurs; establish spin-off companies and commercialising their research results. They can also give the right to use their research outputs in the form of licensing to other firms or organisations. Collaboration between academia and industry has been described both by the Triple Helix (academia-industry-public-sector) and the Knowledge Triangle models (research-education-innovation). An important aspect of knowledge linkages is the dynamism happening between different sectors. These cross-sectoral linkages can enable the region to build novel industrial profiles. Thus, the interactive learning processes between heterogeneous firms and other regional actors create a momentum for new cluster formation or transformation (Trippl et al, 2014). Knowledge transfer mechanisms do not only exist within the specific regional boundaries, but are sometimes with an interregional or international dimension. This is why it is also important to investigate the connectedness of the regional actors to geographically broader knowledge and innovation networks. In this context, the objective of the Regional Ecosystem Scoreboard is to measure the level of knowledge transfer existing between academia and industry in the region and to capture the availability of public knowledge transfer mechanisms. It will also include the measurement of the international connectedness of the region (for instance through the participation in international research, technological or innovation collaboration, platforms, networks, etc.). We differentiate between sub-dimensions of knowledge linkages such as 1) general system linkages 2) specialisation 3) cross-sectoral linkages and 4) openness to extra-regional knowledge-flows.
Dimension of Demand Conditions For the purposes of our conceptual framework, market transformations are understood as mechanisms that can alter the behaviour of market actors and can create specific demand for innovation. To this end, we distinguish between dynamics influenced by private and public demand. Lead users play an important role in the development and diffusion of innovations. Lead users are referred to as individuals or organisations that experience and express needs for a given innovation earlier and are willing to take up and use innovations first (von Hippel, 1986). In the role of lead users
13
Regional Ecosystem Scoreboard – Methodology Report
both private and public actors can act that are well aware of their needs and proactively collaborate with firms in the stimulation and co-creation of ideas and innovations (see also Izsak and Edler, 2011). This dimension of the Regional Ecosystem Scoreboard will look at specific patterns in buyer sophistication, lead users, market dynamics and demand-related factors that the public sector can stimulate such as government procurement as a driver of business innovation or government procurement of advanced technological products. We differentiate between public and private demand factors.
Dimension of Entrepreneurial Conditions As the European Commission’s Entrepreneurship 2020 Action Plan states Europe needs more entrepreneurs if it wants to bring Europe back to growth and higher levels of employment. Entrepreneurship is indeed a key ingredient both of regional innovation systems and clusters as it can spark regional industrial change, adaptation and self-organisation (Feldman, 2005). Entrepreneurship can be defined as “the capacity and willingness to develop, organise and manage a business venture along with any 1 of its risks in order to make a profit” . It is often understood as the mix of entrepreneurial attitudes, entrepreneurial activity, and entrepreneurial aspirations (Acs et al, 2009). By measuring the level of entrepreneurship in a region, more information becomes available with regard to business dynamics of starting a new business or further developing existing ones. The relationship between entrepreneurship and economic development is not straightforward. First of all a linear link clearly does not exist between the two, but can be rather depicted in a mildly S-shape (Acs et al, 2009). Secondly entrepreneurship should not be equalled with start-ups (Isenberg, 2014), although emerging niches are often triggered through new businesses. Many studies also point out the difficulty in finding comparable, available and recent data that can describe the current situation and not something in the very past. There is also a difference between necessity entrepreneurship that can be found in developing regions and opportunity entrepreneurship that characterises innovating regions. Cross-sectoral dynamics should be also highlighted that play a special role in entrepreneurship. Dynamics in one sector can have a positive or negative effect on the dynamics of other sectors. Intended or unintended cross-sectoral encounters can create new opportunities for so far non-existent ventures, new technological combinations or new business models. Nevertheless they can also shake the industrial value chain and weaken the conditions for other related industries. In this context, the objective of the Regional Ecosystem Scoreboard is to capture the ease of doing business, infrastructure (such as transport, broadband or support services) and entrepreneurial culture. Under entrepreneurship culture we understand the attitude of people towards creating their own business and having the necessary skills to recognise opportunities and realise them in practice. Under entrepreneurship infrastructure we understand the availability of broadband, transport infrastructure and the access to supporting services. Under ease of doing business we understand factors that can ease or amplify the administrative burden of would be entrepreneurs who would like to create and run a new venture in the specific region.
Dimension of Quality of governance By measuring the quality of governance one can capture the extent to which national or regional governments and administrations perform their activities in an impartial and incorrupt manner (Charron et al., 2014). This is an important dimension since many empirical studies found a positive correlation between high corruption and weak rule of law with low-level of economic development (Mauro 2004). 1
http://www.businessdictionary.com/definition/entrepreneurship.html
14
Regional Ecosystem Scoreboard – Methodology Report
The lack of an impartial legal structure can unfortunately hinder many entrepreneurial minded people to further develop or expand their ideas since they bump into barriers that cannot be overcome unless the institutional frameworks change (Rothstein and Teorell, 2008). In the framework of the European funded ANTICORP research project, a European Quality of Government Index has been designed which measures the perceptions and experiences with public sector corruption. The Index is based on survey data on corruption and governance at the regional level within the EU, conducted in first in 2010 and then again in 2013. The Regional Ecosystem Scoreboard will make use of the above-mentioned project and will take into account the quality of governance in the regional benchmarking.
2.2 Regional peer groups With the objective to allowing the user to compare the quality of conditions of its regional innovation and entrepreneurial ecosystem to relevant peer regions, we will rely on the benchmarking tool and reference regions developed by Deusto Business School, the European Commission JRC-IPTS and Orkestra in the framework of the regional smart specialisation process. This regional benchmarking classifies regions according to their structural similarity evaluated on the basis of 42 variables representing different factors such as socio-demographic characteristics, sectoral structure, technological specialisation or business size. The model makes a distinction among seven dimensions, which are geo-demography, human resources, technology specialisation, economy and industry specialisation, firm structure, openness, and institutions and values. The definition of a region relies on NUTS statistical units: in general NUTS2 level has been used, except in Belgium, Germany and the United Kingdom, where NUTS1 is applied instead. In order to determine the distance between NUTS regions, a distance matrix was developed that is 2 calculated according to specific weighting methods, based on a single Principle Component Analysis 3 analysis on the 22 components. The benchmarking tool allows for the creation of peer groups of 10 to 35 peer regions according to their distance. For the purposes of the Regional Ecosystem Scoreboard we will use the 20 nearest regions to identify the peer group of each region. Besides relying on this already developed grouping of peer regions as presented above, the online system of the Regional Ecosystem Scoreboard will also allow for the user to determine its own peer group and compare itself to specific self-selected regions. Figure 3 Example: peer regions of the Basque Country
2
Principal component analysis (PCA) is a statistical procedure that uses an orthogonal transformation to convert a set of observations of possibly correlated variables into a set of values of linearly uncorrelated variables called principal components. The number of principal components is less than or equal to the number of original variables.
3
For more information about the methodology http://s3platform.jrc.ec.europa.eu/regional-benchmarking
15
of
the
regional
benchmarking
please
see:
Regional Ecosystem Scoreboard – Methodology Report
Source: Methodological Report, Navarro et al. (2014). Regional benchmarking in the smart specialisation process: Identification of reference regions based on structural similarity, S3 Working Paper Series No. 03/2014
3. The Measurement Framework 3.1 The Regional Ecosystem Indicators Following the review of each of the six dimensions as described in Section 2.2, specific indicators have been selected that can capture the dynamics in the regional ecosystem. These indicators are presented below (please note that the detailed presentation of the indicators can be found in Annex 2). In line with the objective of the Regional Ecosystem Scoreboard, the indicators have been selected so that they capture conditions and dynamics but not the performance of regional innovation systems such as SME’s innovation activities or innovative products and services on the market, which are seen rather as the result of favourable conditions. Compared to the 2014 edition of the Regional Ecosystem Scoreboard, certain indicators have been deleted that captured national level conditions and other regional indicators could replace them. The revision resulted in 50 indicators that are listed in the following sections.
E: Entrepreneurial conditions This dimension includes a total of ten indicators. The guiding principle behind the selection of indicators that can capture entrepreneurial conditions was to find available data related to the conditions to set up and grow a new business, entrepreneurial culture and infrastructure. We deliberately did not take into account indicators related to new business formation, since the goal of this Scoreboard is not about capturing entrepreneurship activity, but the conditions under which entrepreneurship can thrive (or eventually decline). Regulatory framework conditions such as starting a business are usually set at national level and applicable across all regions within the country except for some such as Germany, hence these indicators were taken into account at national level. Title of the indicator
Source
E1. Regulatory framework for starting a business E1.1. Number of days for starting a business
World Bank
E1.2. Difficulties encountered when starting a business
Flash Eurobarometer
E1.3. Barriers to entrepreneurship
OECD
E1.4. Ease of doing business
Regional Competitiveness Index
16
Regional Ecosystem Scoreboard – Methodology Report
E2. Entrepreneurial culture E2.1. Entrepreneurial motivation
Global Entrepreneurship Monitor
E2.2. Business and entrepreneurship education
Global Entrepreneurship Monitor
E2.3. Cultural and social norms
Global Entrepreneurship Monitor
E2.4. It is important to think new ideas and being creative
European Social Survey
E2.5. Most people can be trusted
European Social Survey
E3. Attractiveness of the region and quality of infrastructure E3.1. % of total households with access to broadband
Eurostat
E3.2. Transport infrastructure (Motorway/railway potential accessibility)
Regional Competitiveness Index
K: Knowledge basis and skills This dimension includes a total of eight indicators. The indicators selected cover three different key aspects of knowledge basis and skills: the availability of human resources, the quality of these resources as well as the available infrastructure to train the regional human resources; the availability of vocational training in the region as well as life-long learning in order to measure how adaptive is the region to changes and to renew the quality of its human resources; and the availability of skills in the private sector, in order to understand the capacities of the private sector to respond to market changes, maintain competitiveness and become more innovative. Title of the indicator
Source
K1. Human resources K1.1. Population having completed tertiary education
Eurostat [edat_lfse_11]
K1.2. Total R&D personnel (% active population) - Business enterprise sector
Eurostat [rd_p_persreg]
K1.3. Employment in knowledge intensive services
Eurostat [htec_emp_reg2]
K2. Vocational training and lifelong learning K2.1. Availability of training and coaching programmes for innovative skills in companies
Own survey
K2.2. Lifelong learning. Participation of adults aged 25-64 in education and training, % of population aged 25-64
Eurostat [trng_lfse_04]
K3. Skills K3.1. Availability of technical skills in enterprises
Own construct based on CIS 2010 data [[inn_cis7_csk]
K3.2. Availability of design/creative skills in the private sector
Own construct based on CIS 2010 data [[inn_cis7_csk]
K3.3. Employment in ICT
Eurostat [lfst_r_lfe2en2]
C: Collaboration and internationalisation This dimension includes a total of 13 indicators. The indicators have been selected according to different types of knowledge linkages that can exist among innovation actors and which are critical in stimulating the development of emerging industrial niches. Some of these indicators relate to spillovers 17
Regional Ecosystem Scoreboard – Methodology Report
between businesses or science and business linkages, which are used in general to capture regional knowledge flows. Cross-sectoral linkages have been singled out as a separate sub-dimension, since connections across technological disciplines and industrial sectors can reflect dynamics in related variety within a region. Extra-regional collaboration and international openness is a further sub-area that was considered as important given the fact that it can contribute to a free flow of new ideas and it can foster new entrepreneurship. Title of the indicator
Source
C1. General system linkages C1.1. Firms cooperating with HEIs and PROs
CIS 2012
C1.2. Number of spin-offs
Own survey
C1.3. Innovative SMEs collaborating with others (CIS)
CIS 2012
C2. Cross-sectoral linkages C2.1. Number of cross-sectoral innovation projects within clusters
Own survey
C2.2. Number of cross-technological patenting
Own construct PATSTAT
C2.3. Number of co-working spaces
Own Survey
based
on
C3. Specialisation C3.1. Specialisation in (strong) clusters
European Cluster Observatory
C3.2. Specialisation in knowledge-intensive emerging industries
European Cluster Observatory
C4. Openness of the region C4.1. FP7 & H2020 leverage
CORDA data, DG RTD European Commission
C4.2. SMEs participation in private sector in FP7
Own construct based on CORDA data, DG RTD European Commission
C4.3. FDI and technology transfer
Regional Competitiveness Index
C4.4. Foreign nationals in skilled occupations
European Cluster Observatory, based on Labour Force Survey, Eurostat
C4.5. Number of international co-publications
Own construct based on Leiden Ranking data
F: Access to finance This dimension includes a total of ten indicators. The indicators chosen for this dimension take into account the access to finance by regional enterprises from different perspectives: private financing, public funds from the public sector; as well as the legal framework available to support access to finance. Indicators at regional level in relation to this dimension are very limited and most of the available secondary data is only at national level and when available at regional level, only for a small subset of the EU28 regions. The indicators for this dimension propose a mix of secondary sources, as well as some collected through cluster managers and regional development agencies. As a whole, all ten indicators should provide a good picture of the ease of accessing funding by enterprises at regional level.
18
Regional Ecosystem Scoreboard – Methodology Report
Title of the indicator
Source
F1. Attitudes of regional investors and private financing F1.1. Availability of venture capital
Own survey
F1.2. Availability of bank loans for capital investments to firms
Own survey
F1.3. Share of bank loan applications that were not successful
Own construct based on CIS 2010
F1.4. Investor protection
OECD Entrepreneurship Index / World Bank (Doing Business)
F2. Legal framework supporting access to finance F2.1. Strength of legal rights
Small Business Act Initiative / World Bank
F2.2. Country Credit Rating
OECD Entrepreneurship Index, IMD World Competitiveness Yearbook
F3. Availability of funds from public sector F3.1. Direct loans by government to SMEs
OECD Scoreboard on financing SMEs and entrepreneurs
F3.2. Share of innovators receiving public financial support
Own construct based on CIS 2012
F4. Support from Structural Funds F4.1. Structural Funds in support of services for training in connection with restructuring of sectors
Own construct based on DG REGIO European Commission data
F4.2. Structural Funds dedicated to entrepreneurship and SMEs
Own construct based on DG REGIO European Commission data
D. Demand conditions This dimension includes a total of seven indicators. Two sub-dimensions are considered notably private and public demand conditions. Despite the fact that it is a very important condition that creates opportunities for an emerging niche to expand and for businesses to cluster, available meaningful data are scarce. This is the reason why the list of included indicators is short and some of the data only available at national level. Title of the indicator
Source
D1. Private demand D1.1. Market dynamics
Global Entrepreneurship Monitor
D1.2. Buyer sophistication
OECD / World Economic Forum
D1.3. Spending on innovative products
Own construct based on Eurostat
D1.4. Access of lead users in clusters
Own survey
D2. Public demand D2.1. Role of government in purchasing innovative goods and services 19
Own survey
Regional Ecosystem Scoreboard – Methodology Report
D2.2. Government procurement as a driver of business innovation
Own survey
D2.3. Government procurement of advanced technological products
Public Sector Innovation Scoreboard
QG: Quality of government This dimension includes one indicator, a composite on the quality of the regulatory environment. Other relevant indicators as originally included in the RES conceptual design have been dropped as they 4 were highly correlated indicators within the dimension. Title of the indicator
Source
QG1.1. Quality of regulatory environment
Own construct based on QoG EU Regional Data
3.2 Collection of primary data: survey with cluster managers and regional development agencies The objective of the survey with cluster managers and regional development agencies was to collect data for a set of indicators in the areas where secondary data was not available and to increase the coverage and robustness where data at regional level was limited. Data was gathered using a structured questionnaire online survey. The target population of the survey was the around 1,000 cluster organisations, with which the European Cluster Observatory team has close collaboration; and the network of regional development agencies members of the European Association of Development Agencies (EURADA). A total of 457 responses of regional stakeholders were collected covering 153 regions in the EU.
3.2.1
Questionnaire design
Annex 2 and Annex 3 present the draft survey questionnaires. Both questionnaires were divided in six different sections (plus an introductory section), corresponding each to one of the regional ecosystem dimensions. Each of the questions is linked to one indicator. Implementation The survey was implemented in 4 steps: Step 1: Validation and cognitive testing The draft survey questionnaires in Annex 2 and Annex 3 were firstly validated and/or updated by a group of experts during the validation workshop of the Regional Ecosystem Scoreboard in April 2015. Both questionnaires presented in the Annexes were adapted following the expert comments received. In addition, a total of 10 cognitive testing interviews with a selection of cluster organisations (6) and regional development agencies (4) were undertaken. The objective of these interviews was to ensure that all questions were understood as intended by all potential respondents and that all of them could give an accurate answer. The interviewees were asked to probe the survey questions and their inter4
Originally: QG1.2. Rule of law; QG1.3. Government effectiveness; QG1.4. Product market regulation.
20
Regional Ecosystem Scoreboard – Methodology Report
pretation of specific words. Once validated and tested, the survey was uploaded onto an online survey software. Step 2: Distributing the survey The next step in the implementation process, after making the final revisions based on the expert validation and cognitive testing was to distribute the survey to the target population. All cluster organisations and regional development agencies received a personalised email invitation to take part of the survey. The survey was accessible via a weblink over the period July-September 2015. Step 3: Monitoring response rates Responses to the survey were monitored on a weekly basis and targeted reminders in different national languages were sent (e.g. Spanish, French, Italian, Greek). When necessary, telephone reminders were also conducted. Step 4: Survey analysis and construction of indicators Finally, all collected data was used to construct the regional ecosystem indicators defined in the Indicator framework of Section 3.1. All survey-collected indicators were weighted using an adjustment for non-response on the sample in order to reduce differences between the collected responses relative to the ‘population’ of cluster organisations and regional development agencies available in each region. The weight applied to all survey indicators is equal to the reciprocal of the regional response rate, or the ratio of selected sample to achieved sample size in each region.
3.3 Imputation of missing data For calculating the regional ecosystem composite indicators, data should be available for all indicators and regions. Imputation techniques were used in order to increase data availability at both country and 5 regional level. The techniques used are the following : First case: missing data at the country level, ■
If the country data for the initial year of the period is missing and the data for the following year is available, then the missing country data will be replaced by: 𝐷!! = 𝐷!!!!
■
If country data for a given year is not available but the data for both the previous and following year are available, the average of both years will be used to replace the missing country ta: 𝐷!! =
■
!!!!! ! !!!!! !
If the data for the final year of the period is missing, the data for the previous year will be used to replace the missing country data: 𝐷!! = 𝐷!!!! where C is the country, Y is the current year, Y-1 is the previous year
Second case: missing data at the regional level (NUTSi), ■
5
If regional data for a given year is not available and both regional and that at a higher aggre6 gate level data for the previous year are available, the ratio between the two last available da-
These techniques have been tested and implemented broadly in other EU Scoreboard initiatives such as the Innovation Union Scoreboard (http://ec.europa.eu/enterprise/policies/innovation/files/ius/ius-2014_en.pdf) and the Regional Innovation Scoreboard (http://ec.europa.eu/news/pdf/2014_regional_union_scoreboard_en.pdf).
21
Regional Ecosystem Scoreboard – Methodology Report
ta for the previous year is multiplied by the current value at the higher aggregate level. The following equation describes the technique used: 𝐷!! =
𝐷!!!! ∗ 𝐷!! 𝐷!!!!
where R is the region, C is the country (as the higher aggregate level), Y is the current year and Y-1 is the previous year. ■
If regional data for a given year is not available and both regional and that at a higher aggre7 gate level data for the following year are available, the same above-described procedure will be used but using the ratio between the corresponding NUTS level and that at a higher aggregate level for the following year: 𝐷!! =
𝐷!!!! ∗ 𝐷!! 𝐷!!!!
where R is the region, C is the country (as the higher aggregate level), Y is the current year and Y-1 is the previous year. ■
If there are no regional data for both the previous and following year and regional data on high correlated indicators are available, the missing regional data will be substituted by the predicted values obtained from regression. This method is known as regression imputation (OECD, European Commission, Joint Research Centre (2008)) and the dependent variable of the regression will be the regional indicator containing the missing values (𝐷!! ), and the regressors will be regional indicators showing a strong relationship with the dependent variable: !
𝐷!!
!
𝛽! 𝐻 ! !
= 𝛼! + !!!
corr(𝐷!! , 𝐻′! )>0 where R is the region, Y is the current year, H’ are regional correlated indicators and j is the number of available regional correlated indicators. ■
If there are no regional data for both the previous and following year and regional data on high correlated indicators are not available, the missing regional data will be replaced by the higher 8 level aggregate for the current year, and, if not available, then that will be substituted either by the previous or the following year depending of the data availability: D!! = D!! or D!! = D!!! or D!! = D!!! where R is the region, C is the country (as the higher aggregate level), Y ! ! is the current year, Y-1 is the previous year and Y+1 is the following year.
Third case: Indicators based on CIS data Some indicators were constructed using CIS data. Following the CIS regionalisation technique developed in the Regional Innovation Scoreboard 2014 (Hollanders et al., 2014), the missing regional data was estimated under the assumption that the observed shares at the industry level also applies at regional level and using available regional data on employment or number of firms at industry level. For example, the share of firms with product innovations for region R is estimated as the corrected 6
NUTS1 for NUTS2 regions, NUTS0 for NUTS1 regions.
7
NUTS1 for NUTS2 regions, NUTS0 for NUTS1 regions.
8
NUTS1 for NUTS2 regions, NUTS0 for NUTS1 regions.
22
Regional Ecosystem Scoreboard – Methodology Report
average of the share of firms with product with innovation in the country multiplied by the share of local enterprises of Industry in the region). The following equation is applied:
!"#$%!! ! 𝐷!" =
!"#
%$! !!∈! !" ! ! !! !" !
∗
! !!" ! !!
∗ 𝐶𝑜𝑟𝑟𝑒𝑐𝑡𝑖𝑜𝑛 𝑓𝑎𝑐𝑡𝑜𝑟
Where R is the region, C the country (as the higher aggregate level), F is number of firms, I is the industry (eg. NACE 2-digit industries), i is the subset of firms with product innovations in the industry I, E is persons employed, Y the current year, Y-1 the previous year and Y+1 the following year. A correction factor is applied and it is equal to
!"#$ !"#$% !" !"# !" !"#$%&' !"#"! ! ! !"#$%!! !!"∈! ∗!!" !"#$% ! !! !" ! !! !
. This factor is needed in order to
correct the estimated value of the contribution of all industries into the real value reported by CIS2010. When the regionalisation of the data is estimated using the data on number of firms, the same procedure above-described is used but instead of data on employment, the number of firms (local units) is used.
Fourth case: No regional and no country level data available ■
If there are no regional and no country level data available for the current year, previous and following year, missing regional data will not be imputed.
3.4 Composite indicators 3.4.1
Normalisation of data
After the imputation of the missing data, composite indicators were estimated for each of the Scoreboard sub-dimensions (seventeen in total). Since individual indicators were expressed in a variety of statistical units, ranges or scales they were put in a common scale before adding them in composite indicators. These indicators were calculated assuming that all individual indicators compiled follow a normal distribution. To do so, all individual indicators were transformed using a square root transformation with power N if the degree of skewness of the raw data exceeded 1 such that the skewness of the transformed data is below 1 (Hollanders et al., 2014). After that, data were normalised using the min-max procedure. The transformed score was calculated as follows: for positive indicators, the score was estimated as the ratio of the difference between the raw indicator value and the minimum score observed for all regions and all years divided by the range of the observed score for all regions across the entire time frame. For negative indicators, the score was estimated as the ratio of the difference between the maximum score observed and the raw indicator value for all regions and all years divided by the range of the observed score for all regions across the entire time frame. The obtained normalised scores range from 0 to 1. The estimation of the normalised scores is described in the following equations:
If 𝐷! >0 ,
23
Regional Ecosystem Scoreboard – Methodology Report
𝐷! =
𝐷! − 𝑀𝑖𝑛(∀, 𝐷! ) 𝑀𝑎𝑥 ∀, 𝐷! − 𝑀𝑖𝑛(∀, 𝐷! )
If 𝐷!