THE RIS UNDER SCRUTINY

the ris under scrutiny

Methodological debate on the ‘Regional Innovation Scoreboard’

the ris under scrutiny Methodological debate on the ‘Regional Innovation Scoreboard’

Juan Carlos Salazar-Elena Universidad Autónoma de Madrid Jon Mikel Zabala-Iturriagagoitia Deusto University

WITH THE TECHNICAL SUPPORT OF COTEC’S ECONOMICS DEPARTMENT

Aleix Pons Head of Department Josep Bosch Analyst

DIEGO QUIJANO

METHODOLOGICAL DEBATE ON THE ‘REGIONAL INNOVATION SCOREBOARD’.

TA- BLE OF CON- TENTS

Executive summary

01. Introduction

02. Some methodological discussions about synthetic indices

03. The Spanish context in accordance with the EIS

04. Robustness of the ranking of regions in accordance with the RIS

4.1. Stability the ranking of regions of the RIS

4.2. The significance of RIS differences between regions according to the relative weights of the variables in the RIS

4.3. Towards a robust regional innovation

ranking system in Europe

05. Inputs, outputs and efficiency in the RIS

5.1. Indicators that characterise innovation

system inputs and outputs

5.2. Efficiency characterisation through the relationship between inputs and outputs

5.3. Efficiency characterisation through

Data Envelopment Analysis

06. “Bottlenecks” in the RIS

6.1. Ranking in accordance with the bottleneck penalisation 66 6.2. Ranking in accordance with the penalisation in business R&D expenditure and the degree of openness of the SMEs 68

07. Conclusions

Bibliography

Annex I. Methodology to obtain random weights Annex II. Rankings for all European regions according to the perspectives developed

Annex III. Methodology to obtain the efficiency measurement of a system 96

Annex IV. Summary sheets by Spanish region

METHODOLOGICAL DEBATE ON THE ‘REGIONAL INNOVATION SCOREBOARD’.

Exec- utive sum- mary

The ultimate goal of these indicators and synthetic indices is to help public administra- tors tasked with drafting innovation policies by identifying specific aspects of the innovation system that require public intervention due to their structural deficiencies or requirements. However, the literature shows that most terri- tories define their innovation policies according to their relative position in the rankings, rath- er than as a result of structured, systematic analysis of the indicators behind the synthetic indices or the conclusions that can be drawn from them. Previous research has shown that the meth- odology followed in drafting these synthetic indices has a direct impact on the rankings, and therefore on the policies subsequently implemented in most European countries and regions. We hope that this report will help inform the debate on the methodology used in drafting synthetic indices, focusing on analysis of the Regional Innovation Scoreboard (RIS), which is one of the main instruments used by the European Commission to evaluate the innovation performance of European regions. To measure the innovation performance of Eu- ropean regions, the RIS introduces a synthetic index obtained from the arithmetic mean of its 21 indicators. When constructing the RIS index using this approach, it is assumed that higher RIS values mean a better innovation system (i.e. the more, the better). However, this very nature of the RIS also means that the value of this synthetic index will increase even if the result of increasing the resources allocated to support innovation is zero growth. The decision to construct the RIS index using this method- ology may have a direct impact on the relative position of the Spanish regions in the final ranking, without forgetting the implications that this ranking may have on any participa- tion in calls for proposals or on the allocation of resources in structural policies such as the Smart Specialisation Strategies developed by the European Commission.

Innovation is one of the main drivers of economic growth and social wel- fare. Recent dec- ades have seen the development of an increasing number of indicators, with a view to charac- terising innovation and its impact on territorial com- petitiveness. Giv- en the large number of indicators re- quired to measure a phenomenon as com- plex as innovation, these are grouped together in a syn- thetic index that provides a final ranking of the ter- ritories being ana- lysed.

METHODOLOGICAL DEBATE ON THE ‘REGIONAL INNOVATION SCOREBOARD’.

The report develops three new perspectives to supplement the one already provided by the European Commission itself through its RIS index. Thanks to these new perspectives, the report provides a new ranking for European regions as a whole in terms of innovation. The first of these new perspectives consists of changing the relative weights attributed each indicator in the final synthetic index. In the RIS index, the European Commission attributes the same specific weight to all indicators (i.e. the arithmetic mean), which implicitly reflects the fact that the relative importance of the func- tions that characterise an innovation system remains the same, regardless of the territory. Assigning different relative weights to each indicator in the synthetic index will hopefully obtain a measurement of the degree of robust- ness offered by the RIS index, and, therefore, the robustness of the ranking offered by the European Commission in relation to the inno- vation performance of European regions. Are significant variations observed in the ranking of Spanish regions when different relative weights are assigned to the indicators in the RIS? report. This ranking is obtained by considering the number of European regions that have a significantly higher RIS index than that of a given region. A significant improvement in the performance of the regional innovation systems in Spain can be observed in all cases when comparing the ranking offered by the European Commission for each of the Spanish regions with the robust ranking estimated here. However, even with this robust ranking, most regions continue to occupy intermediate or low positions within Europe, thus demonstrating that the systems in place in most regions have low innovation capacity and therefore require public policies for improvement. However, such A robust RIS ranking is provided from the first of the perspectives considered in this policies should not focus solely on increas- ing the resources allocated to the innovation system, but also on improving its performance and operation in terms of producing innovation results.The second perspective involves ana -

The third of the perspectives developed in the report focuses on analysing identifying “bot- tlenecks” that hinder the overall functioning of the regional innovation system. According to this perspective, any system can only improve to the extent that its weakest link is reinforced. Hence, while prior efficiency analysis aims to offer public administrators decisions that would make their respective innovation sys- tems perform more efficiently, the results of the bottleneck analysis focus on making these systems perform more effectively. After iden - tifying the main bottleneck for each region, we calculated a new synthetic index that corrects the performance level of each territory, based on the penalisation caused by this bottleneck. What are the main bottlenecks hindering the overall performance of the Spanish regions? And to what extent would system performance improve if such bottlenecks were improved? The results of this third analysis show that the main bottlenecks for the Spanish regions are low business investment in R&D, lack of inter- action in the system, and low technological innovation production capacity (both in terms of processes and products) in the SME s. These results are in line with those obtained in the efficiency perspective, since they point to the fact that any improvements in the performance of innovation systems in Spain must go hand- in-hand with both better use of inputs and also an improvement in the quantity of results in terms of innovation outputs.

lysing regional innovation system efficiency in converting investment in supporting the devel- opment of innovating activity (i.e. inputs) into innovation results (i.e. outputs). According to the methodological structure of the RIS index, the RIS index will grow whenever a territory increases the inputs dedicated to the innova- tion system while the outputs remain constant (i.e. the arithmetic mean of inputs and outputs grows), therefore improving this territory’s ranking. However, from a public administrator's point of view, it is not rational to assume that an innovation system will perform better by simply increasing the number of inputs without an equivalent rise in outputs. We apply non-par - ametric methodologies such as Data Envelop- ment Analysis to construct a measurement of the efficiency of regional innovation systems in Europe. To what extent do the Spanish regions behave efficiently or inefficiently? And what are the reasons for any inefficient performance? This efficiency approach has shown that there is a group of Spanish regions with high per- formance (Balearic Islands, La Rioja, Valencia Region, Catalonia and Murcia Region), another group with intermediate performance (Aragon, Madrid Region, Cantabria, Castile-La Mancha and Extremadura), and another group with low performance (Asturias, Andalusia, Galicia, Basque Country, Navarre, and Castile-León). Apart from obtaining an efficiency-based rank - ing of European regions, the analysis is also useful in identifying the main sources of inef- ficiency for each regional innovation system. This inefficiency can be due to either an excess of inputs or a shortage of outputs. In terms of inputs, the analysis shows a low return on inno- vation expenditure in most Spanish regions. As for outputs, the analysis shows that efficient regions with a similar input level to the Spanish regions obtain a significantly higher output, es - pecially when it comes to producing industrial designs and product and process innovations.

Indicators in the RIS to measure the innovation perfor- mance of European regions

New perspectives to analyse region- al innovation systems

METHODOLOGICAL DEBATE ON THE ‘REGIONAL INNOVATION SCOREBOARD’.

SUMMARY TABLE Relative position of the Spanish Regions in the RIS 2021 rank- ing according to the different analysis criteria

These three perspectives have been applied to all European regions for 2021 (the latest sta- tistical year available), focusing on the perfor- mance of the Spanish regions. The report's multidimensional approach to innovation aims to provide administrators in charge of defining innovation policies in the Spanish regions with a comprehensive overview of their respective territories and of their relative performance level in each of these perspectives. We therefore want to help administrators to better understand the strengths and weaknesses of each regional innovation system. We trust that the report's conclusions and recommendations can help give innovation policy administrators direction, allowing them to make more efficient, effective resource allocation decisions and ensure more consolidated, dynamic, competitive innovation systems.

Note: The data indicate the number of European regions ranked highest according to each criterion used.

SPANISH REGIONS

RANKINGS

ROBUST RIS

EFFICIENCY RIS

PENALISED I RIS

PENALISED II RIS

RIS

Andalusia Aragon Asturias

175 150 167 176 200 169 184 159 110 186 153 151 102 161 112 97 132

155 112 134 148 179 136 159 127 50 163 124 110 42 126

180 97 171 0 197 140 155 224 47 161 182 0 107 49 206 195 0

168 141 159 177 208 161 174 148 89 188 147 152 93 156 94 74 127

176 156 168 185 205 170 181 163 123 186 154 152 124 169 115 96 142

Balearic Islands Canary Islands Cantabria Castile-La Mancha Castile-Leon Catalonia Extremadura Galicia La Rioja Madrid Region Murcia Region Navarre Basque Country Valencia Region

59 45 70

METHODOLOGICAL DEBATE ON THE ‘REGIONAL INNOVATION SCOREBOARD’.

“Would you tell me, please, which way I ought to go from here?” asked Alice. “That depends a good deal on where you want to get to,” said the Cat. “I don’t much care where...,” Alice began to say. “Then it doesn’t matter which way you go,” said the Cat.

LEWIS CARROLL, Alice’s Adventures in Wonderland

METHODOLOGICAL DEBATE ON THE ‘REGIONAL INNOVATION SCOREBOARD’.

Intro- duction

Innovation is one of the main drivers of economic growth and social welfare (Solow, 1956, 1957). Innovation policy is key in shaping innovation systems and their per- formance (Nelson and Romer, 1996; Barbosa and Faria, 2011; Taylor, 2016; Edler and Fagerberg, 2017). Innovation systems are made up of a complex network of interact- ing organisations, institutions and pol- icies whose main goal is to improve the conditions necessary for innovations to emerge and develop, and then be disseminat- ed and adopted (Metcalfe, 1995; Palmberg, 2006), all with a view to improving the competitiveness of the territories and the well-being of their citizens.

01.

Characterising innovation systems has always been a challenge for policy-makers and re- searchers alike (Janger et al. , 2017), given the difficulty in determining the number and type of indicators to use, as well as the absence of a theory on innovation systems and policies (Grupp and Schubert, 2010; Edquist and Laat- sit, 2022). While the number of indicators to measure innovation has increased over recent decades, the problem of characterising inno- vation and innovation systems still persists (Dziallas and Blind, 2019).

The European Commission has been one of

METHODOLOGICAL DEBATE ON THE ‘REGIONAL INNOVATION SCOREBOARD’.

the most active agents in developing, using and exploiting innovation indicators with the so-called European Innovation Scoreboard (EIS). New indicators are developed in order to capture the range of activities or functions oc- curring within a system, as necessary for it to reach the required outcome. Other approaches equivalent to the EIS for measuring innovation on a territorial scale are the Global Innovation Index (prepared by the WIPO-World Intellectual Property Organization), the Nordic Innovation Monitor (prepared by the Nordic Council of Ministers), or the Bloomberg Innovation Index, to mention a few. All these approaches are based on a synthetic index which summarises the set of indicators used to measure the phe- nomenon under study in a single variable. This synthetic index therefore provides a final rank - ing which can form the basis of a classification of the territories under analysis. While their relevance in terms of innovation policy practice is low (Arundel, 2007; Arundel and Hollanders, 2008; Mairesse and Mohnen, 2009), the syn- thetic indices developed to measure innovation have major implications, as many policy deci- sions are made on the conclusions drawn from them (Schibany and Streicher, 2008; Koz ł owski, 2015; Edquist et al. , 2018). when it comes to monitoring the results of the Innovation Union, which is one of the flagship initiatives in the Europe 2020 Strategy to create an innovation-friendly environment (European Union, 2021a) 1 . The EIS is offered nationally on a yearly basis, while regional analysis (called RIS- Regional Innovation Scoreboard) is offered every other year (in odd-numbered years, the latest being 2021) 2 . In order to measure the innovation performance of European countries and regions, both the EIS and the RIS introduce a synthetic index that ranks the territories con- sidered. This synthetic index, called Summary Innovation Index (SII) for national level and Re- Published yearly since 2001, the EIS is the European Commission's main instrument

gional Innovation Scoreboard (RIS) for regional level, is obtained through the arithmetic mean of all the indicators included in the EIS/RIS. Higher SII/RIS values therefore mean a better innovation system (i.e. the more, the better). The very nature of the SII/RIS also means that this will increase even if the result of increasing the resources allocated to support innovation is zero. The simplistic use of synthetic indices such as the SII/RIS can be pernicious, as rankings based on them are accepted without question- ing their validity (Grupp and Schubert, 2010). The risk is that, once accepted, synthetic indices can become political targets, thus mak- ing them ineffective at providing insight into certain situations and acting on them. This is commonly referred to as Goodhart’s law (Free- man and Soete, 2009), i.e. when a synthetic indicator becomes a target, it ceases to be a good indicator. Increasing the SII/RIS and im - proving its relative position in the ranking has thus become the target for many innovation policies, when in fact they should be looking to identify problems and improve the perfor- mance of innovation systems (Edquist, 2011). If the EIS/RIS is to have a positive impact on the performance of innovation systems, it is imperative that it first allows these systems to be exhaustively characterised, with a view to identifying their strengths and weaknesses and subsequently defining policies to meet their needs. The role of the public sector cannot be lim- ited exclusively to allocating more and more resources to support innovation. This is the logic that underlies the EIS/RIS, since assign- ing more resources will directly impact the synthetic index, without this necessarily having any effect on performance by the system as a whole. This implies following the linear inno - vation model (Rodriguez-Pose and Crescen- zi, 2008). The simplicity of the linear model

means that, despite being completely rejected in innovation research (Edquist, 2014), it still dominates innovation policy since, accord- ing to this dominant (albeit erroneous) logic, performance of an innovation system could be improved by simply increasing R&D activities. The Lisbon strategy (European Parliament, 2000) is a clear example of predomination of the linear innovation model, a “the more, the better” logic, and the risk and inefficien - cy shown by the synthetic indices. In 2000, this strategy defined that 3% of the European Union’s GDP should be allocated to R&D by 2010. As this target was not met, it was set again in 2010 (this time with a view to 2020), only for history to repeat itself and once again fall short of the target. As Rita Mae Brown said, in a phrase often attributed to Einstein, insan- ity is doing the same thing over and over and expecting a different result. This type of error in interpreting the mechanisms governing the behaviour of innovation systems, and, there- fore, in formulating innovation policies, has clear implications in terms of the inefficiency and ineffectiveness of such policies (Samara et al. , 2012). The role of the public sector cannot be limited exclusively to allocating more and more re- sources. It needs, above all, to give the system direction (Mazzucato, 2018). However, this requires us to first identify the direction we want the system to take, as reflected in the conversation between Alice and the Cheshire Cat in the introduction. Measuring innovation at territorial level is one of the most relevant challenges in the field of innovation policy. This is due to innovation be - ing a multidimensional phenomenon, to territo- rial heterogeneity, and to the lack of measures capable of capturing many of the key aspects conducive to innovation (Carayannis et al. , 2018). However, it would be utopian to assume

2021 EIS and RIS cover pages published by the European Commission.

1 All EIS annual reports are available free of charge: https://ec.europa.eu/info/research-and-innovation/statistics/performance-indicators/european-innovation-scoreboard_en

2 All RIS annual reports are available free of charge: https://ec.europa.eu/info/research-and-innovation/statistics/performance-indicators/regional-innovation-scoreboard_en

METHODOLOGICAL DEBATE ON THE ‘REGIONAL INNOVATION SCOREBOARD’.

The rest of the document is structured as fol- lows. Chapter 2 provides an overview of some of the most representative methodological debates around synthetic indices, and presents the three perspectives addressed in this paper. Chapter 3 frames the Spanish situation within the context of European countries, based on the data provided by the EIS at national level. Chapter 4 analyses the robustness of the RIS ranking for European regions in the event of variations in the specific weights of the indica - tors in the RIS index, firstly noting its stability, and secondly offering a robust indicator for it. This robust RIS indicator provides a measure- ment of the “significance” of the differences between the indices of the European regions by assigning different weights to the RIS in- dicators. Chapter 5 analyses the efficiency of innovation activity for the European regions as a whole, identifying those factors which explain the inefficiency of regional innovation systems from an input or output perspective. Chapter 6 focuses on analysing the bottlenecks that constrain the performance of the whole in- novation system. This analysis has relevant implications for public action, since, by iden- tifying the factors that limit the performance of the entire system, it provides clear direction for innovation policy administrators in each territory. Far from being partial, these policies would focus on those system elements that require most support, since their effects would generate positive network externalities for the entire system. In other words, while chapter 5 offers policy-makers decisions to make their respective innovation systems perform more efficiently, the conclusions set out in chapter 6 focus on making these systems more effec- tive. Finally, the main conclusions are set out in chapter 7, identifying the main areas of public intervention required in order to improve per- formance for each regional innovation system in Spain.

that we will reach a point in time when we have a “complete” number of indicators available to provide the final characterisation of an inno - vation system. As a scientific community, we therefore find ourselves in a context in which we must make the most of the indicators available today. This implies that we must put on as many sets of “glasses” as possible to interpret a single reality (measured through the indicators we have at this moment) from different perspectives (Filippetti and Peyrache, 2011). Each set of “glasses” represents a meth - odological approach with a particular view of a multidimensional reality. Existing research in this regard shows that, depending on the perspectives (i.e. “glasses”) used, the overview of performance of an innovation system can vary significantly (Zabala-Iturriagagoitia et al. , 2007a, 2007b, 2008; Edquist et al. , 2018; Barbero et al. , 2021). Having a comprehen - sive understanding of an innovation system, its challenges, problems and needs therefore requires us to use as many “glasses” as possi- ble and combine the conclusions drawn from them, ready to feed the definition of policies. The aim of this report is to analyse perfor- mance of Spain’s regional innovation systems through a range of perspectives (“glasses”), giving an overview of its relative position in Europe in order to define the innovation pol - icies to be implemented. To this end, we will address three perspectives of analysis, all of them based on the data provided by the 2021 edition of RIS for all European regions: (I) change in relative weights of each indicator to be included in the synthetic index; (II ) efficiency of the innovation activity, considering the ratio between inputs and outputs; and (III) analysis of bottlenecks hindering overall system perfor- mance. These three perspectives have been applied to the 225 European regions reporting all the indicators in the RIS for 2021, and their results have been compared with the official results published by the European Commission in its RIS ranking 3 .

Measuring innova- tion at territorial level is one of the most relevant chal- lenges in the field of innovation poli- cy.

The aim of this re- port is to analyse the performance of Spain’s innovation system from three different perspec- tives

3 The fifteen European regions excluded from the analysis due to missing data include two Spanish regions (the autonomous cities of Ceuta and Melilla).

METHODOLOGICAL DEBATE ON THE ‘REGIONAL INNOVATION SCOREBOARD’.

Some methodo- logical discussions about synthetic indices

A first point of discussion with regards to calculating synthetic indices is the relevance of the choice of indicator weights vector (Grupp and Schubert, 2010; Becker et al. , 2017; Hauser et al. , 2018). A key characteristic of synthetic indices is that the sum of the relative weights to be given to all indicators in the analysis must be equal to 1. In other words, a relative weight must be assigned to each of the indicators, such that the sum of all of them equals 100%. Choosing the weights vector assigned to each indicator is critical in constructing synthetic indices (Cooper et al. , 2011). This is because each set of weights defines a particular func - tion or aggregation model under which per- formance is measured differently, therefore producing alternative results and unequal interpretations In the case of innovation analysis and meas- urement, this is a debate that is difficult to resolve. Is private R&D more important than public R&D? Do technological innovations represent innovation performance better than non-technological innovations? And in any of these options, how much more important is one variable than the other? Furthermore, the relevance of each factor depends on its terri- torial context, as what may be important in a specific territory may not be so in another. For example, public R&D expenditure tends to have

Scientific research on synthetic indi- ces identifies a set of debates (Salt- elli, 2007); Nar- do et al. , 2008; Lane, 2010; Greco et al. , 2019), some of which are dealt with in this docu- ment.

02.

METHODOLOGICAL DEBATE ON THE ‘REGIONAL INNOVATION SCOREBOARD’.

a lower impact on private R&D and business innovation in environments where companies' capacity to absorb knowledge is lower (Castel- lacci and Natera, 2013), while a positive inter - action between public expenditure and private R&D expenditure is observed in environments where companies have high capacity to absorb knowledge. Despite recognising their impor - tance, the RIS does not go further in assigning different relative weights to each indicator, but rather gives all variables the same relative weight, which implies, in operational terms, that the same relative importance is assigned to all factors of the system, regardless of its context (i.e. uniform weights for disparate territories). This trade-off between ensuring consistency in the relative weights in order to compare the territories being analysed and capture the differences between them (i.e. not all variables have the same relative importance in each territory) is particularly complex in the European sphere, with its 27 countries and 240 regions. The decision to assign identical weights to all variables and all territories (as the European Commission does in the SII/RIS calculation) is undoubtedly one way to remain neutral in this debate. However, this arbitrary position does not clarify when the differences in the positions of the two regions in the final inno - vation performance ranking can be attributed to this random decision or to significant differ - ences in the performance of their associated systems. In this first analysis, we will try to answer whether it can be concluded that the differences in the innovation performance of the European regions measured using the RIS synthetic index are due to the relative weight of its variables, or whether some regions always perform better than others, regardless of the weights assigned to these variables. To shed light on this first debate, we will study the sta - bility of the RIS in the event of changes in the weights of the variables, and propose a robust ranking for these weights.

Finally, a third debate envisages the possibility of interactions (i.e. complementarity effects (positive), or substitution effects between factors (negative)) between the indicators that measure an innovation system’s different activities and functions (Tarabusi and Palazzi, 2004); Lafuente et al. , 2021). The synthetic indices calculation assumes perfect substitut- ability between the components of an innova- tion system. However, as seen in the previous paragraph, considering that inputs can be sub- stituted by outputs could be problematic from an economic and public policy perspective. Furthermore, the fact that there are comple- menting factors in the innovation system (cap- tured in the RIS variables) could lead to “bottle- necks”, with development being slowed down by certain aspects of the innovation system (Ács et al. , 2014). These bottlenecks represent the weakest functions of the system (i.e. the function that acts as a constraint for the whole system), and therefore penalise system perfor- mance. Identifying these bottlenecks ensures poilcy-makers in each territory count on clear guidelines in terms of the interventions re- quired in order for the overall system to benefit from public action (i.e. additionality).

A second debate, not completely unrelated to the first one, refers to the problem of adding variables that are not designed to measure exactly the same phenomena (Edquist and Za - bala-Iturriagagoitia, 2015; Edquist et al. , 2018). Although the RIS indicators are undoubtedly related to innovation activity, there is a consen- sus that they measure different aspects of the innovation system. One of the most commonly used classifications (although not the only one) of the indicators in the RIS divides them into three groups: facilitators, business activities, and results (European Union, 2016). From the conventional perspective of the production function, the first two groups can be seen as “inputs” (internal and external to the company, respectively) of the regional innovation sys- tem's innovation activity, while the third one is for the “outputs” produced by the associ- ated system. Despite these clear differences between types of indicators, the synthetic index offered by the RIS does not differentiate between input and output variables, attributing the same relative weight to all of them. This methodological decision could lead to unrea- sonable situations from the point of view of economic analysis. For example, let us imagine a hypothetical context where two regions find themselves in opposing situations: the first has very high values for input indicators and very low values for output indicators, while the second shows the opposite, i.e. very few inputs and many outputs. Although the RIS synthetic index would attribute the same performance level to both regions (since they would have the same arithmetic mean), we will all agree that both cases represent two opposing examples of a system working efficiently in generating innovation results. In a second analysis, we will therefore develop a methodological approach to analyse the efficiency of regional innovation systems in Europe, allowing us to see, from this perspective, which territories make better use of the investment in their respective sys- tems, indicating that their respective innovation systems work better.

METHODOLOGICAL DEBATE ON THE ‘REGIONAL INNOVATION SCOREBOARD’.

The Spanish context in accordance with the EIS

As mentioned in the introduction, the Eu- ropean Commission offers an innovation lev- el diagnosis at both national (via the EIS synthetic index) and regional (via the RIS synthetic index) levels. This section aims to analyse the Spanish situation in the context of European countries, based on the data provided by the EIS at national level. This chapter can therefore help contextual- ise the results to be included in sections four, five and six of this report, which an - alyse the situation of the Spanish regions in the European context.

The coloured columns in Figure 1 show the performance of European countries in 2021 relative to that of the European Union average (EU = 100 for 2014). The figure is accompanied by the performance shown by each country in 2014 (thin black line) and 2020 (thick black line), in both cases also relative to the EU average for each year. The dashed lines show the thresholds that allow European countries to be classified in categories by performance: innovation leaders (countries with performance over 125% of the EU average): Sweden, Finland, Denmark and Belgium), strong innovators (be- tween 100 and 125% of the EU average: Neth - erlands, Germany, Luxembourg, Austria, Esto- nia, France and Ireland), moderate innovators

03.

METHODOLOGICAL DEBATE ON THE ‘REGIONAL INNOVATION SCOREBOARD’.

(between 70 and 100% of the EU average: Italy, Cyprus, Malta, Slovenia, Spain, Czech Republic, Lithuania, Portugal and Greece), and emerging innovators (below 70% of the EU average: Cro - atia, Hungary, Slovakia, Poland, Latvia, Bulgaria and Romania). On average, the EU’s innovation performance has increased by 12.5% since 2014. Innovation performance has increased since 2014 in all EU member states, being highest for Cyprus, Estonia, Greece, Italy and Lithuania. More recently, between 2020 and 2021, performance has improved in twenty Member States, nota- bly Cyprus and Estonia, and declined in seven: France, Ireland, Latvia, Luxembourg, the Neth- erlands, Portugal and Slovakia. (FIGURE 1) Based on the findings of recent EIS reports, Spain has been included in the moderate innovators group (light-pink group in Figure 1), as its innovation system is on a smaller scale than that of the average EU country. Analysing growth by the SII index for Spain shows that it grew by 16.2% between 2014 and 2021, which is above the EU average (12.5%). This seems to point to the fact that Spain is immersed in a convergence process within the EU, where countries with lower innovation performance grow faster than those with higher perfor- mance. However, the trend shown by those countries found alongside Spain in the moder- ate innovators category (Figure 2) points to the need to qualify the above statement. Although Spain (black line) is among the countries with a higher SII in the moderate innovators category in 2021 (only Italy, Cyprus, Malta and Slovenia show a higher SII in this category), it is striking that countries with a lower SII show significant - ly higher growth rates than Spain. For example, Lithuania’s SII has grown by 50.4% between 2014 and 2021, Cyprus’ by 45%, Greece’s by 41.3%, Italy’s by 31.8%, and Malta’s by 17.2% 4 . This fact shows that other moderate innovat- ing countries' convergence rates with the EU average are significantly higher than for Spain, which points to the need, firstly, to reconsider the amount of resources invested in the nation-

FIGURE 1 Innovation performance of European countries and country categories

16/27

Score relative to EU27 average in 2014 (=100) Source: European Union (2021: 6)

Spain remains in the third ranking category, i.e. mod- erate innovator

LEADING INNOVATORS STRONG INNOVATORS MODERATE INNOVATORS EMERGING INNOVATORS

IN 2014

IN 2020

EU 27 100

150

16.2%

SWEDEN FINLAND DENMARK

1 2 3 4 5 6 7 8 9

BELGIUM HOLLAND GERMANY LUXEMB O URG

Spain’s EIS growth rate for 2014-2021, higher than the EU average

AUSTRIA ESTONIA FRANCE IR E LAND EU AVERAGE ITALY CYPRUS MALTA SLOVENIA SPAIN CZECH REPUBLIC

10 11

12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27

LITHUANIA PORTUGAL GREECE CROA T IA HUNG ARY SLOVAKIA POLAND LATVIA BULGARIA R O MAN I A

METHODOLOGICAL DEBATE ON THE ‘REGIONAL INNOVATION SCOREBOARD’.

al innovation system and, secondly, to improve innovation activity results at national level. (FIGURE 2) The 2021 EIS indicators are grouped into 12 blocks: (1) Human resources; (2) Attractive re- search systems; (3) Digitalisation; (4) Funding and support; (5) Business investment; (6) Use of information technologies; (7) Innovators; (8) Relationships; (9) Industrial property rights; (10) Impact on employment; (11) Impact on sales and (12) Environmental sustainability. The image in Figure 3 shows Spanish evolution in each aspect between 2014 and 2021. Spain has performed positively in several of these aspects between 2014 and 2021, most notably improving human resources (with growth of 42.6%), business investment (40%), and rela - tions (30.8%). Meanwhile, we find poorer per - formance for impact on employment (-15.3%), industrial property rights (-12.1%), and attrac - tive research systems (-1.4%). 5 . (FIGURE 3) The conclusion we can draw from the EIS analysis at national level is that the Spanish innovation system is small in size. From this perspective, the Spanish science, technology and innovation policy should focus on those functions that require either more funding or an improvement in system operation, as a way to ensure better results in terms of innovation output. However, it should be remembered that increasing resources without taking into consideration their performance and assuming that they will naturally bear fruit is not a good strategy, since this would mean assuming that the linear model of innovation is true, when we know that it is not (see Section 2). That is why it is important to include the efficiency perspective when evaluating performance of innovation systems, as we argue in this docu- ment.

The literature on measuring innovation and characterising innovation systems e.g. Ed - quist et al. , 2018; Zabala-Iturriagagoitia et al. , 2021) has shown that the relationship between innovation system size and performance in terms of efficiency is not linear, since higher innovation capacity (i.e. the size or scale of the system) does not automatically imply that the system will be able to convert this growing capacity into a larger number of innovation results, which would mean lower efficiency of the system as a whole. One of the latest nationwide studies in this area (Barbero et al. , 2021) points to four different patterns for these two dimensions (Figure 4): (a) countries with low innovation capacity and low performance (LILP); (b) countries with high innovation capacity and high performance (HIHP); (c); countries with high innovation capacity and low performance (HILP), and (d) countries with low innovation capacity and high performance (LIHP). Barbero et al. (2021) show that most national innovation systems, including Spain’s, operate under diminishing returns to scale. In other words, the results (outputs) would grow at a lower rate if the system resources (inputs) were doubled. This shows that the current resource allocation is far from adequate, since in most cases resources are still being allo- cated to “saturated” system dimensions and functions, meaning the marginal gain from increased investment in these functions in terms of overall system performance will be practically nil. This requires in-depth recon - sideration of those dimensions and functions that require greater effort in order to achieve a balanced, consolidated, efficient innovation system. (FIGURE 4) Based on Barbero et al. (2021), the results are considerable, even despite innovation capacity (size) in Spain being small. As can be seen in Figure 4, Spain would be ranked 22nd in terms of inputs, while in terms of efficiency it would be in 7th place in the ranking of European countries. These results show the low innova -

FIGURE 2

FIGURE 3

Innovation performance of moderate innovating countries

Evolution of Spain's main national innovation system dimensions Value received by Spain in the Summary Innovation Index 2021 in each aspect of the EIS Source: Drafted in-house based on European Union (2021a)

Score relative to EU27 average in 2014 (=100) Source: Drafted in-house based on European Union (2021a)

EU AVERAGE

SPAIN

DIGITALISATION

ITALY

CZECH REPUBLIC

ENVIRONMENTAL SUSTAINABILITY

CYPRUS MALTA

LITHUANIA PORTUGAL

HUMAN RESOURCES

USE OF INFORMATION TECHNOLOGIES

SLOVENIA

GREECE

IMPACT ON SALES

ATTRACTIVE RESEARCH SYSTEMS

120

FUNDING AND SUPPORT

RELATIONSHIPS

BUSINESS INVESTMENT

110

IMPACT ON EMPLOYMENT INDUSTRIAL PROPERTY RIGHTS

INNOVATORS

100

0,9

0,8

0,7

0,6

0,5

0,4

0,3

0,2

0,1

4 Of the 38 countries covered by the EIS 2021, Spain’s SII level growth for 2014- 2021 ranks 17th, with Lithuania (50.4%) growing the most in its synthetic index, and Ukraine (-13.7%) the least. 5 Details of the EIS 2021 results for Spain are available at https://ec.europa.eu/docsroom/documents/45936.

2014

2021

2014

2021

METHODOLOGICAL DEBATE ON THE ‘REGIONAL INNOVATION SCOREBOARD’.

adopt the knowledge and innovations devel- oped by other countries, since their innovation systems do not enjoy the conditions neces- sary to effectively generate and disseminate innovations internally. This strategy requires lower levels of development, and therefore less investment in system operation, while at the same time producing more efficient behaviour (since, firstly, risk is reduced and, secondly, it is faster to adopt the knowledge and innova- tion developed by others than to produce it oneself). It should also be noted that countries with fewer resources should pay more atten- tion to how these are used, as they cannot afford to waste the scarce resources devoted to innovation activities, with caution always leading to greater efficiency. One possible strategy for Spain, whose in - novation system is dominated by SMEs and mini-SMEs, would be to exploit the advantages brought by this industrial fabric (e.g. flexibil - ity and use of external networks); this would require a clear improvement in the system’s ability to effectuate, merging internal and exter- nal networks that allow SMEs to quickly access knowledge and technology and to integrate them in their production processes (Daven- port and Bibby, 1999). In other words, Spanish innovation policy should fundamentally con- sider creating and consolidating technical-sci- entific networks with a view to maximising the country's capacity to absorb and transform its innovation system. As part of the COTEC-fund - ed Open Innovation Programme, a study was carried out between 2018 and 2020 to analyse open innovation patterns in the three most dynamic regions: Catalonia, Madrid and the Basque Country (Salazar-Elena et al. , 2020). The study shows the importance of having agents acting as a bridge between scientific research and business innovation, with tech- nology centres emerging as a key institution in terms of developing a more productive and in- novative business fabric. In line with the results set out in this report, the study by Salazar-Ele- na et al. (2020) shows that Madrid Region and Catalonia still have much room for improve-

FIGURE 4 Relationship between the system size and performance in terms of efficiency Horizontal axis: Ranking based on the innovation inputs scale (4 inputs) Vertical axis: Ranking based on innovation performance (RCII) Source: Barbero et al . (2021: 10)

In terms of in- puts, Spain would rank 22nd, while in terms of efficiency it would climb to 7th position

THE MORE THE BETTER PARADIGM EFFICIENCY PARADIGM

LILP

HIGH HILP

LOW INNOVATION , LOW PRODUCTIVITY

INNOVATION , LOW PRODUCTIVITY

LITHUANIA

SWEDEN

ESTONIA

CROATIA

FINLAND

ISLANDIA

BULGARIA

POLAND

NORUEGA

ROMANIA

BÉLGICA

R. UNIDO

HUNGARY

ment in promoting open innovation. As for the Basque Country, although it has made a much clearer commitment to technology centres, it also has its weaknesses, since its collaboration networks are mainly local. This diagnosis at national level is reproduced by the European Commission at regional level (European Union, 2021b). In line with the struc - ture provided by the EIS, where countries are classified in four groups (see Figure 1) accord - ing to their SII, the RIS also classifies Europe’s regions in innovation leaders (38 regions), strong innovators (67 regions), moderate innovators (68 regions), and emerging innova- tors (67 regions). However, given the territorial diversity in the 240 European regions, the RIS provides a more detailed breakdown of each of these groups by dividing them into three subgroups: a “+” is assigned to the subgroup with the best performance, and a “–” to the subgroup with the worst one (see Figure 5). The most innovative regions will therefore be “Innovation Leaders +”, while the least inno- vation will be “Emerging Innovators -”. Table 1 shows the logic followed in characterising European regions into these groups and sub- groups. (TABLE 1)

LATVIA

CZECH REPUBLIC

SLOVENIA

SWITZERLAND

IR E LAND

FRANCE

PORTUGAL

HOLLAND

SPAIN

CYPRUS

DENMARK

SLOVAKIA

GERMANY

GREECE

HIGH INNOVATION , HIGH PRODUCTIVITY

LOW INNOVATION , HIGH PRODUCTIVITY

LUXEMB O URG

MALTA

ITALY

AUSTRIA

HIHP

LIHP

tion capacity of the system as a whole (i.e. low inputs), which increases efficiency in the short term (i.e. high output/input ratio). However, this transitory result should not be misinterpreted in a self-interestedly triumphalist way, since innovation performance must be evaluated from a long-term rather than a short-term per- spective. Evidence in this regard is clear, as it

shows that the lack of system investment and the absence of policies focused on improving system operation will lead to a loss of competi- tiveness in the long term. Countries classified in the Spain group (i.e. countries with low innovation capacity and high performance: LIHP) tend to absorb and

METHODOLOGICAL DEBATE ON THE ‘REGIONAL INNOVATION SCOREBOARD’.

TABLE 1

TABLE 2

Definition of performance groups and subgroups in the RIS

Innovative regions groups and sub- groups by country

Source: European Union (2021b: 18)

Source: European Union (2021b: 20)

HIGHER SUB-GROUP (+)

HIGHER SUB-GROUP (-)

HIGHER SUB-GROUP

GROUP

COUNTRY

EIS GROUP 2021

INNOVATION LEADERS

STRONG INNOVATORS

MODERATE INNOVATORS

EMERGING INNOVATORS

Innovation leaders

1 1 1 1 + 5

- 21

- 24

+ 25

+ 35

- 26

+ 19

- 10

INNOVATION LEADERS + Above 144.8% of EU average STRONG INNOVATORS + Between 116.7% and 125% of the EU average MODERATE INNOVATORS + Between 90% and 100% of the EU average EMERGING INNOVATORS + Between 52.1% and 70% of the EU average

INNOVATION LEADERS Between 134.9% and 144.8% of the EU average STRONG INNOVATORS Between 108.3% and 116.7% of the EU average MODERATE INNOVATORS Between 80% and 90% of the EU average EMERGING INNOVATORS Between 34.1% and 52.1% of the EU average

INNOVATION LEADERS - Between 125% and 134.9% of the EU average STRONG INNOVATORS – Between 100% and 108.3% of the EU average MODERATE INNOVATORS – Between 70% and 80% of the EU average EMERGING INNOVATORS – Below 34.1% of EU average

3 2 9

Innovation leaders Innovation leaders Innovation leaders Innovation leaders Innovation leaders Strong innovators Strong innovators Strong innovators Strong innovators Strong innovators Strong innovators Strong innovators Strong innovators Strong innovators Moderate innovators Moderate innovators Moderate innovators Moderate innovators Moderate innovators Moderate innovators Moderate innovators Moderate innovators Moderate innovators Emerging innovators Emerging innovators Emerging innovators Emerging innovators Emerging innovators Emerging innovators Emerging innovators Emerging innovators

Switzerland Sweden Finland Denmark Belgium Netherlands United Kingdom Germany Luxembourg

3 1 1 1 1 3 2 9

Strong innovators

1 1 1 1 3 4 5

1 2 1 3 4 5

Moderate innovators

1 2

1 8

Emerging innovators

Austria Norway Estonia

1 1

1 2 1

Ireland France Italy Cyprus Malta Slovenia Spain Czech Republic

1 3 4

1 1 6

UE27 100

1 4

1 2

3 4

150

–

INNOVADORES EMERGENTES EMERGING

MODERADOS MODERATE

FUERTES STRONG

LÍDERES LEADERS

1 5 4

1 3

2 1

2 1 1

2 2

One of the conclusions drawn from the RIS is that all regional innovation leaders come from countries identified as innovation leaders or strong innovators, while most of the regions considered moderate or emerging innova- tors are in countries identified respectively as moderate and emerging (see Table 2). In other words, the characteristics of the national innovation system have a direct impact on the performance of the regions in this country. The only exceptions are those regions that the RIS considers to be “excellence hubs”, i.e. whose performance level is higher than for the country they come from. These are Prague in the Czech Republic, Attica and Crete in Greece, the Basque Country and Madrid in Spain, Emilia-Romagna in Italy, Budapest in Hungary,

Warsaw in Poland, the Bratislava Region in Slovakia, and Belgrade in Serbia. (TABLE 2)

Lithuania Portugal Greece Croatia Hungary Serbia Slovakia Poland Latvia Bulgaria Romania

1 3 5 1 4 3 3 6

1 4

1 2

2 2 3 1 1 1

In the case of Spain, a high level of diversity is evident in the regions, it being three times high- er in the best performing region (the Basque Country) than in the worst one (Ceuta) (see Table 3). Two of the regions are considered strong innovators, 10 moderate innovators, and 7 emerging innovators. (FIGURE 5)

1 1

2 7

Page 1 Page 2-3 Page 4-5 Page 6-7 Page 8-9 Page 10-11 Page 12-13 Page 14-15 Page 16-17 Page 18-19 Page 20-21 Page 22-23 Page 24-25 Page 26-27 Page 28-29 Page 30-31 Page 32-33 Page 34-35 Page 36-37 Page 38-39 Page 40-41 Page 42-43 Page 44-45 Page 46-47 Page 48-49 Page 50-51 Page 52-53 Page 54-55 Page 56-57 Page 58-59 Page 60-61 Page 62-63 Page 64-65 Page 66-67 Page 68-69 Page 70-71 Page 72-73 Page 74-75 Page 76-77 Page 78-79 Page 80-81 Page 82-83 Page 84-85 Page 86-87 Page 88-89 Page 90-91 Page 92-93 Page 94-95 Page 96-97 Page 98-99 Page 100-101 Page 102-103 Page 104-105 Page 106-107 Page 108-109 Page 110-111 Page 112-113 Page 114-115 Page 116

Made with FlippingBook - Online Brochure Maker