Regional ICT Innovation in the European Union: Prioritization and Performance (2008–2012)

The European Union incentivizes the prioritization of specific policy fields by regional governments, especially with its 2014–2020 framework for the European Regional Development Fund (ERDF). To this end, the European Commission revised the relevant regulation and introduced a new legal framework for the allocation of ERDF by requiring regional and national authorities to develop dedicated regional strategies and focus their ERDF investment on existing strengths and promising economic activities like ICT development (EU Regulation 1303/2013/EU). This poses questions about the effectiveness of policy prioritization: Do regions with strategies for specific policy fields perform better in these policy fields? Does focused investment of ERDF in specific policy fields yield better performance?

To answer these questions, we examine regional policy prioritization and performance during the ERDF framework 2007–2013. Although the concern for policy prioritization was less pronounced in the previous ERDF framework, the Council of the EU already stressed greater ownership at local and regional level for the Lisbon competitiveness agenda and in its 2006 strategic guidelines for Cohesion Policy (Council Decision 2006/702/EC). We focus our inquiry on ICT policy and performance, since a number of regions had developed dedicated ICT strategies despite the absence of an explicit conditionality or legal obligation in the ERDF regulations. As ICT is one of the nine priority fields for the ERDF, the European Social Fund and the Cohesion Fund, the EU implicitly recommended this policy field for regional policy prioritization (Council Regulation 1083/2006/EC). ICT as a prioritized policy area at EU level goes to the 1990s when the paradigm of ‘information society’ was increasingly streamlined into EU regional development policies (Dabinett 2001).

Using data from the ERDF framework 2007–2013, we take the opportunity to study the effects of regional political prioritization of ICT even before the new incentives for policy prioritization of the 2014–2020 framework took hold. Therefore, our findings can possibly anticipate some effects of the new framework ex ante.

The overarching goal of this article is to analyse the effect of policy prioritization of ICT by regional governments in the EU on their ICT performance as an example of policy prioritization by regional governments. For this, we first conduct a quantitative analysis of the effects of ICT strategies and ERDF expenditure on regional ICT performance for the period 2008 to 2012. Based on this analysis, we map EU regions according to their ICT prioritization and performance to identify interesting cases for future investigation. Finally, we discuss the implications of our findings for the current work on the new generation of smart specialization strategies and policy frameworks for digital growth.

The increasing attention given to dedicated regional strategies and targeted funding for regions in the EU is the result of the emergence of the concept of smart specialization. This concept, as developed by the European Commission and adopted by the European Parliament and the Council of the EU, builds on the accumulated knowledge from a broad range of regional innovation initiatives since the 1990s,¹ general developments within regional and innovation policies, as well as more recent work carried out by Dominique Foray and the Knowledge for Growth Expert Group in the framework of the European Research Area (Foray et al. 2009). This group argued that research investments in Europe have been overly fragmented, lacked critical mass and were plagued with a pronounced ‘me too’ syndrome, i.e. regions investing in highly similar and fashionable areas like ICT, nano- and bio-tech.

In policy guidance documents for the next period of European Structural and Investment Funds that include ERDF, the European Commission urges regions to prioritize innovation investments in areas of strategic potential to systematically support structural change by investing in areas of strategic potential (European Commission 2012). The guidelines for developing regional innovation strategies also strongly emphasize the importance of priority setting and point out how regions should identify their unique assets, challenges, competitive advantages and potentials for excellence.

The aforementioned guidelines carry the implicit assumption that the prioritization of policy goals and the development of detailed strategies are crucial to effective policy-making. However, such a straightforward link between goal formation and policy outcomes presumes a high degree of rationality in strategic management and organizational behaviour. The classic theories of rational administrative behaviour that support these assumptions are heavily contested by concepts like bounded rationality, the influence of politics and power, and the effects of ‘randomness’, e.g. in the garbage can model (Olson 1971; Nelson and Winter 1982; Simon 1991; Eisenhardt and Zbaracki 1992; Pagels-Fick 2010). The orientation provided by general goals and prioritization has been shown to be helpful for policymakers to mediate between institutional inertia and dynamic changes in the environment (Mintzberg 1978). Both provide also useful steering for strategic management even if detailed plans fail (Kay 2010). In contrast, concepts like Lindblom’s ‘muddling through’ have shown that policy implementation is more difficult given the need for coordination and the incremental and emergent nature of strategic management (Lindblom 1959). This calls into question if strategic policy prioritization can be meaningfully applied ex ante. Indeed, it highlights one of the main concerns of organization theorists, namely the inherent tension between policy strategies (what is ought to be done) and their actual implementation (what is done on the ground).

Following these conceptual discussions, we adopt the assumption of rational strategic management as our working hypothesis, which also reflects EU guidelines and even more pronouncedly the requirements of the 2014–2020 ERDF framework for smart specialization. That is, our analysis assumes that the political prioritization and strategic goals of a region are expressed through dedicated ICT strategies and targeted ERDF allocation, both of which are likely to result in better policy performance.

Whether the existence of a dedicated regional ICT strategy has a positive impact on ICT performance is tested with an analysis of variance. The F value (0.006) for the relationship between the binary variable Dedicated ICT Strategy and Household Internet Access 2008/2012 is not statistically significant (p < 0.938). Therefore, we cannot confirm an effect of the adoption of a dedicated ICT strategy by a region on its ICT performance. In fact, a simple descriptive analysis shows that the regions without dedicated ICT strategy (39.71 % improvement in Household Internet Access) performed on average slightly better than regions with a dedicated ICT strategy (39.35 % improvement in Household Internet Access); see also Fig. 2.

Open image in new window

• RQ2: Does the investment of ERDF funds in ICT have an effect on regional ICT performance?

Whether regions allocate ERDF to ICT (specifically telephone and broadband infrastructure) or does not have a significant impact on their ICT performance (Fig. 2). In our model, we find that regions that invest ERDF funds to telephone and broadband infrastructure on average improved their Household Internet Access by 46 %, while regions that did not allocate any ERDF funds to this category only increased their Internet access by an average of 34 %. The analysis of variance shows this correlation to be significant (F = 7.46; p < 0.008).

So far, our analysis has shown that a dedicated regional ICT strategy alone did not have an impact on regional ICT performance. On the other hand, we observed that allocating ERDF expenditure to Internet infrastructure has had a positive effect on a region’s improvement in Household Internet Access. As stated before, one of the objectives of this article is to map the prevalence of regional ICT strategies and identify interesting cases for further research. In this regard, our finding that simply having dedicated regional ICT strategies is statistically not significant for ICT performance certainly is a puzzle that deserves further investigation.

In Table 4, we map the ICT performance of all 97 regions in our sample within the matrix of their expected performance based on our hypotheses.⁵ Only 35 % of the regions behave in the expected way (7 % with high ICT performance, 12 % with moderate performance and 16 % with low performance), while 65 % of the regions displays performance patterns in contradiction to our hypotheses, which are based on the assumption that policy prioritization has an impact policy performance (see Fig. 3).

Table 4

Distribution of regions by Household Access 2008/2012, ERDF and dedicated ICT strategy

		Dedicated ICT strategy
		ICT strategy exists	No ICT strategy
ERDF investment	ERDF investment in category 10	High performance (7 %): Extremadura (ES43), Región de Murcia (ES62), Valle d’Aosta (ITC2), Liguria (ITC3), Toscana (ITE1), Umbria (ITE2), Centro (PT16) Moderate performance (3 %): Voralberg (AT34), Aragón (ES24), Algarve (PT15) Low performance (2 %): Mellersta Norrland (SE32), Scotland (UKM)	High performance (15 %): Castilla y León (ES41), Castilla-La Mancha (ES42), Andalucía (ES61), Border, Midland and Western Ireland (IE01), Bolzano (ITD1), Veneto (ITD3), Friuli-Venezia Giulia (ITD4), Marche (ITE3), Abruzzo (ITF1), Molise (ITF2), Campania (ITF3), Puglia (ITF4), Calabria (ITF6), Sicilia (ITG1), Sardegna (ITG2) Moderate performance (7 %): Brabant Wallon (BE31), Hainaut (BE32), Lombardia (ITC4), Lazio (ITE4), Basilicata (ITF5), Norte (PT11), Yorkshire and the Humber (UKE) Low performance (7 %): Östra Mellansverige (SE12), Småland (SE21), Norra Mellansverige (SE31), Övre Norrland (SE33), West Mindlands (UKG), East of England (UKH), South East of England (UKJ)
	No ERDF investment in category 10	High performance (3 %): Brandenburg (DE4), Sachsen-Anhalt (DEE), Galicia (ES11) Moderate performance (5 %): Kärnten (AT21), Mecklenburg-Vorpommern (DE8), Cantabria (ES13), Navarra (ES22), Alentejo (PT18) Low performance (9 %): Niederösterreich (AT12), Wien (AT13), Tirol (AT33), Baden-Württemberg (DE1), Bayern (DE2), Hamburg (DE6), Nordrhein-Westfalen (DEA), Schleswig-Holstein (DEF), Comunidad de Madrid (ES30)	High performance (6 %): Ceuta (ES63), Piemonte (ITC1), Trento (ITD2), Emilia-Romagna (ITD5), Azores (PT20), North West England (UKD). Moderated performance (18 %): Steiermark (AT22), Bruxelles-Capitale (BE10), Niedersachsen (DE9), Sachsen (DED), Thüringen (DEG), Asturias (ES12), País Vasco (ES21), La Rioja (ES23), Comunidad Valenciana (ES52), Illes Balears (ES53), Melilla (ES64), Canarias (ES70), Southern and Eastern Ireland (IE02), Lisboa (PT17), Madeira (PT30), North East England (UKC), South West England (UKK) Low performance (16 %): Burgenland (AT11), Oberösterreich (AT31), Salzburg (AT32), Liège (BE33), Luxembourg (BE34), Berlin (DE3), Hessen (DE7), Rheinland-Pfalz (DEB), Saarland (DEC), Cataluña (ES51), Stockholm (SE11), Sydsverige (SE22), Västsverige (SE23), East Midlands (UKF), London (UKI), Wales (UKL)

Source: own elaboration

Open image in new window

• Among regions with an ICT strategy and ERDF investment on Internet infrastructure, 58 % (7 regions) displays high performance in accordance with our hypotheses. 25 % (3) of the regions displays moderate performance and 17 % (2) shows low performance.

• Among the regions without ICT strategy but with ERDF investment on Internet infrastructure, 24 % (7 regions) displays moderate performance as expected, 52 % (15) high performance and 24 % (7) low performance.

• Among the regions with ICT strategy, but without ERDF spent in category 10, only 29 % (3) perform moderately as expected. Fifteen percent (5) presents high performance, and 53 % (9) presents low performance.

• Among the regions with neither ICT strategy nor ERDF spent in category 10, 41 % (16) presents low ICT performance as expected, 15 % (6) presents high performance and 44 % (17) moderate performance.

Comparing the expected performance to the actual distribution of the regions, we identify several particularly interesting cases, which are shortly highlighted below.

The Europe 2020 strategy has introduced smart specialization and evidence-based innovation strategies as key processes for fostering structural change in order to drive sustainable growth. This framework strongly incentivizes EU regions to set clear policy priorities in order to access ERDF funds. To anticipate some of the consequences of the new framework, we have examined the effects of policy prioritization in EU regions on their performance with regard to Internet infrastructure during the previous funding period.

Our findings question this general push for priority setting, since we could not find a straightforward link between policy prioritization and improved performance. In our mapping of EU regions, their ICT prioritization and performance, especially the lack of a discernible link between the development of a dedicated ICT strategy and high ICT performance is baffling. In line with the broader strategic management literature, the link between rational strategic management and policy outcomes seems to be less robust than partly put forward by management science and policy initiatives.

Yet, we must be cautious not to over-interpret these initial findings. Our measure of ICT strategies is still very simple and does not take into account their quality and their embeddedness within the broader innovation ecosystem. Only having a formalized document called ‘strategy’ in place does not necessarily mean it is evidence-based, well designed and its implementation is effective. All these are elements that are promoted by the idea of smart specialization for the EU funding period 2014–2020. As other policy studies remind us, any strategy document is embedded in a broader set of other existing documents and initiatives that are inter-related (Rayner and Howlett 2009). Indeed, it is important to expand the analysis of mere broadband infrastructure and access to Internet towards issues of ICT uptake and the ecosystem of advanced broadband applications. All these issues necessarily go beyond a narrow focus on broadband or ICT strategies and relate to pertinent questions of ICT-enabled innovation more broadly (OECD 2008).

Moreover, our sample covers only Western European regions. But most ERDF investments in the past and current funding period have been and will be made in Central and Eastern Europe. There, most countries have undergone a difficult transition from centrally planned communist systems to free market economies. Given this highly volatile environment and insufficient state capacities, public policies are more likely to resemble a patchwork of incongruous elements rather than rational design (Stark 1997). This is related to the degree of complexity of regional innovation systems. Less complex regional societies are more likely to have effective ‘incremental strategies of problem solving’, whereas more complex ones must rely on ‘trial and error’, which may in the best case also lead to larger performance leaps (Kitschelt 1991: 462–63). Adding East European countries and regions to future analyses will most likely require adaptations of our model and the addition of other possible explanatory factors.

Still, our mapping not only reveals puzzling results with regard to policy prioritization but also offers a promising starting point for further investigation and worthwhile case studies. Investment in Internet infrastructure from sources other than the ERDF should be taken into account. Not only the existence but also the quality of regional ICT strategies should be assessed. Aspects from the context of smart specialization, such as stakeholder involvement in policy development and implementation, mechanisms for monitoring and evaluation could also be explored. Further, national and regional characteristics and path dependencies may be of great relevance (Thelen 1999; Streeck and Thelen 2005; García et al. 2014).

With the data provided in this article, EU regions can also identify reference regions similar to them and possibly learn from good and bad practices in the strategy and implementation process. Thus, the baseline we have established will hopefully serve further studies on ICT strategy design and performance in the context of multi-level policies supporting digital growth and innovation at EU, national and regional level.