THE RIS UNDER SCRUTINY

METHODOLOGICAL DEBATE ON THE ‘REGIONAL INNOVATION SCOREBOARD’.

Towards a robust regional innovation ranking system in Europe 4.3

TABLE 8 Comparison between the conventional ranking and the robust RIS ranking in the Spanish regions Note: The robust RIS indicates the number of regions that have a significantly higher RIS than the region in question. The number of Spanish regions are shown in brackets. Source: Drafted in-house based on European Union (2021b) Spanish Regions

RANKINGS

ROBUST RIS

RIS

One major problem in our robust RIS estima - tion is that it does not induce a unique ordering among regions (i.e. a ranking). This is because the positions are always established from the perspective of each of the regions. For exam - ple, let us imagine the following case (relatively common in our estimates) of the following three regions: • Region 3 has a significantly higher RIS than region 1. • There are no significant differences be - tween regions 1 and 2.

ish region according to the methodology fol- lowed by the RIS, while column 3 (robust RIS) represents the number of European regions that have a significantly higher RIS than the Spanish region in question (i.e. as mentioned above, for the Basque Country there would be 45 regions with a statistically higher RIS). It can be seen that the relative position of the Spanish regions in the ranking of European regions changes substantially if we consider the robust indicator. This is particularly true for the mid-ranking Spanish regions. As shown in section 4.1, transforming and normalising the RIS indicators induces a greater agglomeration of regions in middle-ranking positions. Table 8 shows the significant change in the ranking for Catalonia, Madrid Region and the Basque Country when moving from conventional RIS (column 2) to robust RIS (column 3), with improvements in the three cases of more than 60 positions, respectively. In the case of these three regions, only around 45 European regions (of the 225 included in the study) have a signif- icantly higher RIS, while the relative position of these three Spanish regions drops significantly if we consider the position given by the Europe- an Commission’s RIS, occupying mid-ranking positions in the total 225 considered. (TABLE 8) As for the other Spanish regions, all of them have improved their relative position in the ranking from this robust approach; also note- worthy are the improvements observed in the cases of Valencia Region (62-position improve- ment), Navarre (53-position improvement), La Rioja (41-position improvement), and Aragon

Andalusia Aragon Asturias

175 (13) 150 (6) 167 (11) 176 (14) 200 (17) 169 (12) 184 (15) 159 (9) 110 (3) 186 (16) 153 (8) 151 (7) 102 (2) 161 (10) 112 (4) 97 (1) 132 (5)

155 (14) 112 (7) 134 (11) 148 (13) 179 (17) 136 (12) 159 (15) 127 (10) 50 (3) 163 (16) 124 (8) 110 (6) 42 (1) 126 (9)

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

• There are no significant differences be - tween regions 2 and 3.

59 (4) 45 (2) 70 (5)

From the perspective of region 1, we will say that first place in the ranking will belong to region 3, and regions 1 and 2 will share second place. From the perspective of region 2, we will say that the three regions share first place. And, finally, from the perspective of region 3, regions 2 and 3 will share first place, while region 1 will be second. In this sense, since there is no single order, generating something similar to a ranking will require us to find alternative ways to use the information. One possible solution is to use the number of regions with a significantly higher RIS for each of the regions. We will call this indicator robust RIS. This information is set out in Annex II for all regions of Europe. Table 8 shows the Spanish regions. Column 2 (RIS) represents the relative position of each Span-

(38-position improvement), to mention just a few. These results clearly show how the sub - jectivity associated with the relative weights of the indicators in the RIS impact the European Commission’s ranking, with this directly affect- ing the Spanish regions’ positioning in terms of both securing European funds and also attract- ing foreign direct investment, and even when it comes to communicating and disseminating their strategies. We therefore believe that the approach pro - posed in this section is more accurate in providing a ranking of innovative regions than the one offered by the European Commission, which is obtained as the arithmetic mean of the indicators contemplated by the RIS. This

arithmetic mean represents only one of the ten thousand combinations of specific weights considered in our approach. A statistical estimation such as the one followed here, in which the indicators contemplated by the RIS can adopt multiple different relative weights, therefore offers a more robust ranking than the one offered by the European Commission.

Made with FlippingBook - Online Brochure Maker