3.1. Overall process of convergence

Figure 3 shows current innovation performance as measured by the SII on the vertical axis against short-run trend performance of the SII on the horizontal axis. [1] There is a process of convergence in innovation performance in Europe with the Member States with below average performance having positive trends. Most of moderate innovators and catching-up countries are closing the gap with the EU and the innovation leaders and followers. The innovation leaders and followers are experiencing a relative decline in their innovation lead. Notable exceptions include Luxembourg which combines a moderate level of performance which a high SII growth rate; Italy, Spain, Greece and Croatia which all have relatively low SII growth rates; and Norway and Turkey which are experiencing very low SII growth rates. The following section will analyse in more detail if this overall process of convergence is taking place between and/or within the four identified country groupings.

3.2. Stable membership of country groups

As set out in Section 2.1, countries have been classified into different innovation groups based on their SII scores over a 5-year period. Changes in group membership within the 5-year period of time are shown in Figure 4. Group memberships are largely stable but we do see some changes:

• Luxembourg is in the process of moving from the innovation followers to the innovation leaders;
• Cyprus and Malta have moved from the catching-up countries to the moderate innovators;
• Latvia and Romania are first part of a cluster with Turkey and then moved to the catching-up countries.

Cluster membership is more stable than the ranks of countries; ranks within a cluster are far from stable, as shown by for example Belgium in the cluster of innovation followers and the US in the cluster of innovation leaders. These results show that one should not focus too much on changes in rank from one year to the next within the same cluster. It is better to focus on cluster membership and the countries within the same cluster and to identify for each country peer countries. This is consistent with the Strengths and Weaknesses report of 2005 where peer countries were identified based on comparable relative performance levels.

3.3. Convergence between country groupings

The previous section showed that despite the general process of convergence, cluster membership is stable over time. This suggests that the observed convergence is a general trend rather than the result of exceptional single countries' developments. This can be shown by plotting the evolution of the innovation performance of the different clusters (Figure 5). We observe increasing relative performance for the catching-up countries and the moderate innovators, stable relative performance for the innovation followers and declining relative performance for the innovation leaders. Convergence between the country groups is shown in the lower half of Figure 5 where the differences in the cluster SII scores have been plotted over time. The results show a strong process of convergence taking place between the innovation leaders, innovation followers and moderate innovators. There is also some convergence between catching-up countries and moderate innovators. We can estimate the theoretical time of convergence for each of these processes using a simple linear approach which will be discussed in Section 3.4. On this simplified basis, it would take almost 30 years for the catching-up countries to close the gap with the moderate performers, and almost 40 years for the latter to close the gap with the innovation followers and about 25 years for the latter to close the gap with the innovation leaders. In conclusion one can see that convergence between clusters is taking place, but it is likely to take many years before this convergence process is completed.

3.4. Expected time to convergence

Using a simple linear extrapolation of current performance levels and growth rates [2], an estimate can be made for how many years it would take countries to either catch up or decline to the EU average level of performance based on current trends. Figure 6 shows the estimated years to catch up to or decline to the EU average for European countries only. For four of the moderate innovators and catching-up countries a short-term convergence to the EU average performance level could be expected in about 10 year’s time. These countries are Estonia, Czech Republic, Lithuania and Cyprus. For Slovenia short-term convergence could be expected in about 15 year’s time, for Poland, Latvia, Bulgaria, Slovakia, Malta and Romania convergence would take at least 20 years. For Portugal, Hungary and Italy the catching up process would take more than 30 years. On the other hand, countries like Belgium, France, the Netherlands and Denmark: these countries still show an average value of the SII above the EU average, but might regress to the EU average, possibly within the next 5 to 10 years, as the average EU performance increases faster than their individual innovation performance. Finally, based on this analysis, some countries seem to stay outside the convergence process (and are therefore not represented in the chart) as they are either moving away from the average in a negative direction (Spain, Greece, Croatia, Norway and Turkey) or in a positive direction (UK, Iceland, Austria and Luxembourg).

However, linear extrapolations of trends are likely to become less reliable over longer time periods, as maintaining the above EU growth rates will become more and more difficult when countries start to approach the EU average level of performance. A non-linear catching-up process was therefore modelled by assuming that the growth rate of each country would diminish over time [3]. The catching-up process now looks different, with only Estonia and the Czech Republic as likely candidates to complete their catching-up process in the short-run. Belgium, France and the Netherlands are still in danger of falling back to the average EU level of performance within a relatively short time period. While Sweden was predicted to fall back to the EU level in 17 years time using the linear approach, in the non-linear approach it would take more than 100 years.

Understanding how countries' innovation performance can change over time is one of the key future challenges identified in Section 6 (Future challenges). The analysis conducted in this Section shows that over a five year time period there has been a relatively stable grouping of countries, which each group at a different level of innovation performance. This finding points to the difficulty of bringing about major changes in overall innovation performance. This may be because innovation has many dimensions along which countries need to improve in order to increase their overall performance; but also because changing innovation performance simply takes time.

However, over a longer time period we do observe a more dynamic situation. First, there are some countries that appear to have made a transition between different levels of innovation performance and it would appear that some other countries are on track to making such a transition in future. Second, there appears to be a long-term trend towards convergence between the different groupings. If this continues, it may mean that the different groupings merge over time or alternatively it may lead to new patterns and trends emerging.