**3. Materials and Methods**

Driven by the relevance of innovation activity of SMEs and the literature evidence on the existence of SMEs financing gap, this study is designed to shed some light on types of innovations of SMEs that perform in the European Union and its relationship with the relevance of a given type of fund. The differences in the implementation of given types of innovation are captured on the cross-country level (**RQ1**). Further, we investigate the relationships between the types of innovation and the relevance of internal funds, external equity, and debt finance (**RQ2**).

In the empirical investigations, we rely on the data provided on a regular basis as SAFE (Survey of Access to Finance of Enterprises) reports by the European Commission and European Central Bank. The SAFE dataset consists of aggregated survey results obtained for each of the European Union (EU) member states. In this study, we have focused on SAFE results obtained for five consecutive years in a 5-year time span (2014–2018). Since 2013, the results of the SAFE survey are published annually, but since 2014 in a unified format that allows time comparisons. At the moment of the research investigation, the latest results were available for 2018. The SAFE survey sample includes randomly selected SMEs from each EU member state, from various sectors and of various size (micro, small, and medium-sized enterprises). In 2018, there were over 17,000 survey respondents (SAFE 2018).

For this study, we cluster the EU members states to construct the research sample. First, we grouped the countries following the prevailing classification scheme by distinguishing between the "old" and "new" EU members states, as explained in Table 1. This distinction is guided by the fact that "old" EU member states have a long history of performance as a union and are regarded as better developed, in comparison to the new member states. On the other hand, the "new" member states (in this, the SMEs sector in these countries) have benefited from numerous programs that were aimed at enhancing the removal of disparities (before and shortly after the EU accession).


**Table 1.** The clusters of the examined European Union (EU) member states.

Secondly, guided by the studies of Bartlett and Prica (2017) and Bruha and Kocenda (2018), we clustered the EU countries into four groups that consider the existence of core and peripheral EU member states. In particular, consistently with Bartlett and Prica (2017), in the cluster of "old" and "new" EU countries, we further distinguished between the inner core, outer core, inner peripheral, and outer peripheral countries. In the cluster of the "old" EU member states, we identified 7 "inner core" countries: the founders of the EU and the UK, as one of the largest net contributors to the EU budget (Kovacevic 2019). In the cluster of the "new" EU countries, we identified 7 "inner peripheral" countries: the members of Eurozone (see the reasoning explained in Table 1). The adopted scheme of clustering the EU countries is justified by the results of prior works related to innovation activity and capital-structure related issues. The Anwar (2018) study confirmed that the majority of the old EU countries are typically innovation leaders, whereas the post-communist countries (that are a majority in outer and inner peripheral EU countries) are low-moderate innovators. Anwar (2018) also addressed the types of innovations (product and process), including SMEs. The K ˛edzior (2012) study confirmed that there are significant differences in capital structure-related issues between the old and new EU member states.

The list of the examined variables is provided in Table 2. However, it is substantial to explain the methodology behind the SAFE survey and the presentation of its results, as it remains relevant for the design of the empirical investigations performed in this study. The SAFE database presents the percentage structure of SMEs (the respondents) answers for a given question. Moreover, the percentage structure of answers remains aggregated on country level (in other words—it is provided separately for each of the EU member states). In our empirical investigations, we merge two aspects (problems) that were subject to the SAFE study: innovation and sources of funds in SMEs. Within the first aspect (innovation), the survey incorporates the set of questions that refer to four areas of improvements implemented by firms in the past year. In this respect, we are able to identify the percentage of companies which declared the implementation of product, process, management, and sales innovations, as explained in Table 2. The second aspect (financing mix) refers to the relevance of a given source of funds. The SAFE survey provides information on the percentage of respondents (SMEs) who declared that a given source of funds was relevant in their activity in the past six months. In this respect, in statistical examinations, we refer to the percentage of firms (SMEs) which declared the relevance of internal funding, external equity, and debt financing, as explained in Table 2.

In the empirical investigations, we use the non-parametric methods, due to the nature of the available data. In particular, to examine the differences on country-level (in accordance with the defined clusters), the non-parametric ANOVA is used (Kruskal–Wallis test). The relationships of SMEs innovation and the use of particular sources of funds are captured by the Rho–Spearman correlations.


**Table 2.** The definitions of the examined variables.
