*4.1. SMEs Innovations across the EU*

In Appendix A, we present the figures with data on the percentage share of SMEs, which declared the implementation of a given type of innovation between 2014 and 2018 (product, process, management, or sales), in accordance with the assumed clusters of the EU member states. The data indicate that in the case of the implementation of product, process, and sales innovations, the situation could be judged as comparable. The percentage of SMEs which declared the implementation of product, process, and sales innovations was slightly lower in inner core EU countries, in comparison to the remaining core countries. In the group of the old EU countries, Finland is clearly in the leading position. Moreover, in the cluster of the old EU countries in the majority of the countries, peak levels of product and process innovations were observable until 2017, with a visible drop in 2018. A similar pattern of changes was observable in the cluster of new EU countries (2014–2017), however with a slight improvement in 2018. In the inner peripheral countries, the lowest percentage of SMEs that declared

the implementation of the product, or process innovations was in Estonia, whereas in the cluster of outer peripheral countries—in Hungary. In the remaining new EU countries, the percentage of SMEs that declared the implementation of product or process innovations could be judged as comparable, with the most dynamic changes in Cyprus, Latvia, and Malta (only for process innovations). Similar conclusions could be drawn concerning sales innovations, with a leading position in Cyprus and Romania. Cyprus and Romania are also leading in the implementation of management innovations (Figure A3). However, in the case of old EU countries, the situation is different as compared to product, process, and sales innovations. First, a slightly higher percentage of innovative SMEs operating in the cluster of inner core EU countries is observable, as compared to the outer core ones. In the old EU countries, Greece was in a leading position in the implementation of management innovations, as declared by SMEs. The data also indicate that in the majority of countries, the declared level of management innovations remained unchanged between 2017 and 2018.

The initial observations on the overall trends and differences between the declared level of the implementation of innovation in European SMEs (Appendix A) justify the examination of the statistical relevance of these differences, within the defined clusters of countries and consistently with the first research question (RQ1). In Table 3, we present the results of non-parametric tests that have confirmed that there are statistically significant differences between the defined clusters of countries, concerning the percentage of SMEs that declared the implementation of the product, management, and sales innovation, but not the process innovations. Within the comparisons between new and old EU countries, the mean ranks of the U Mann–Whitney test (Figure 1) indicate that in the cluster of new EU countries, a higher percentage of SMEs declared the implementation of product innovations, as compared to the old EU countries. However, we observe a contrary situation in the case of management and sales innovations. The percentage of SMEs, which declared the implementation of management and sales innovations was significantly higher in the old EU countries, as compared to the new ones.

**Figure 1.** Mean ranks of U Mann–Whitney test for differences in the declared implementation of innovation between the clusters of old and new EU member states. Notes: statistically significant differences at \*\* α = 0.05.

The Kruskal–Wallis test indicates the overall differences between the four clusters of EU countries. Thus, to detects which specific groups of countries differ significantly, the post hoc tests were performed (see Table 3). In the case of product innovations, the post hoc tests have confirmed that there are statistically significant differences between the inner core EU countries, as compared to the remainder clusters. The analysis of mean ranks of the Kruskal–Wallis test, illustrated in Figure 2 indicates that the percentage of SMEs that declared the implementation of product innovation was significantly lower in inner core countries, as compared to the remainder clusters. In the case of management innovation, statistically significant differences were observed in several dimensions. First of all, the post hoc tests and the ranks of the Kruskal–Wallis test indicate that the percentage of SMEs which declared the implementation of management innovation was significantly higher in the cluster of

outer core EU countries, in comparison to the clusters of the outer and inner peripheral. Moreover, the percentage of SMEs which declared the implementation of innovation in management was significantly higher in the cluster of inner core countries, in comparison to the clusters of inner and outer peripheral (Figure 2). Finally, in the case of sales innovation, the statistically significant differences were confirmed only between the cluster of the outer core EU countries and the cluster of outer peripheral countries (significantly higher percentage of SMEs declared the implementation of innovations in sales in the cluster of the outer core, as compared to the cluster of outer peripheral EU countries).



Notes: statistically significant at \*\* α = 0.05; symbols: IC—inner core, OC—outer core, IP—inner peripheral, OP—outer peripheral.

**Figure 2.** Mean ranks of the Kruskal–Wallis test for differences in the declared implementation of innovation between the sub-clusters of the EU countries. Notes: statistically significant differences at \*\* α = 0.05.

We additionally examined the possible co-existence between the types of innovations most frequently undertaken by the SMEs. In Table 4, we present the Rho–Spearman correlation coefficients to capture the associations between the percentage of the European SMEs, which declared the implementation of various types of innovations, broken by the clusters of analyzed countries.


**Table 4.** Correlations between a percentages of the European SMEs which declared the implementation of innovations.

Notes: statistically significant at \*\* α = 0.01; \* α = 0.05.

The data presented in Table 4 indicate that in general, in the cluster of the old EU countries, there were statistically significant correlations between all types of innovations, except for product and management. It suggests that the undertakings of various types of innovations co-existed in the surveyed SMEs that perform in the old EU member states. This pattern of interdependencies is repeated in the sub-clusters of inner and outer core countries. However, in the case of the cluster of inner core countries, the correlation coefficients between product innovation and process and sales innovations are visibly lower, as compared to the cluster of outer core countries. In the cluster of new EU countries, there are statistically significant correlations between the implementation of all types of innovations, and this pattern is also observable in the sub-cluster of outer peripheral countries, with visibly higher correlation coefficients. In the case of inner peripheral countries, the statistically significant correlations are observable only between the following pairs of innovations: process and product, management and sales, as well as management and product. These observations clearly indicate that the co-existence of the implementation of various innovations was higher in the cluster of outer peripheral countries (for statistically significant correlation coefficients). That can be explained by the 'spiral of innovation' phenomenon, which describes the relationship between different types of innovations: one successful innovation (e.g., product innovation) begets the other (e.g., process

innovation). One type of innovation generates demand on the other, as well as it provides solutions that can be used to create new developments.
