Missions, conditions and the policy transfer of Smart Specialisation in the European Union

ABSTRACT

Smart Specialisation is fundamentally a bottom-up policymaking process premised on entrepreneurial discovery and peer learning. Nonetheless, the European Union has relied on coercive policy transfer through the ex-ante funding conditionality instrument to secure its implementation across regions. This paper examines how the use of conditionality has shaped the implementation of Smart Specialisation in the European Union and its regions. Furthermore, it discusses whether the use of conditionality to ensure the implementation of a particular type of innovation policy reflects a changed perception of the role of the state in the broader European integration project.

KEYWORDS:

• innovation policy
• Smart Specialisation
• conditionality
• regional policy
• European Union institutions
• industrial policy

JEL:

• O19
• O32
• O38
• R58

1. INTRODUCTION

It has become somewhat of a stylised fact that regional economic development policies, especially when conducted at the level of supranational organisations such as the European Union (EU), should reflect the heterogeneity in the myriad cultural, economic and administrative ways (Suorsa, 2014) by which one can define a region. There is thus growing acceptance of the need for place-based innovation policies that build on a regional economy’s unique resources and competences (e.g., Barca et al., 2012; McCann & Ortega-Argilés, 2013). In this vein, Smart Specialisation has emerged as a popular framework: proposing that governments at the national and/or subnational levels identify strategic areas in which to invest in innovation, building on analysis of their territories’ strengths and potential through an entrepreneurial discovery process (EDP) with broad stakeholder involvement (e.g., Foray et al., 2009). Inherent to this is the idea that a region recognises its unique opportunities to develop competitive advantage; the emphasis on uniqueness helping to avoid the tendency for regions to imitate perceived best practices in other regions, as in the emergence of ‘Silicon Somewheres’ (Hospers, 2006; Tödtling & Trippl, 2005).

For levels of supranational governance, such as the EU, which have embraced the principle (Foray et al., 2011), adoption of Smart Specialisation has ensured a differentiated innovation strategy in which regions specialise in different domains. However, implementation of Smart Specialisation mostly takes place at the substate or regional level, in line with the centrality of a bottom-up policymaking process enshrined within the EDP. Higher levels of government therefore depend on policy transfer processes to ensure implementation by lower levels.

There is by now a large body of research on what constitutes unique opportunities for regions (e.g., Balland et al., 2019; Boschma & Gianelle, 2013), and a growing literature on how regions implement Smart Specialisation (e.g., Deegan et al., 2021; Di Cataldo et al., 2022; Gianelle et al., 2020; Gianelle et al., 2020; Marrocu et al., 2023). However, the policy transfer mechanisms inherent to the movement of Smart Specialisation policy receive less attention. The EU’s version of Smart Specialisation – Research and Innovation Strategies for Smart Specialisation (RIS3) – took the initial form of an ex-ante conditionality (ExAC) within the 2014–20 Regional Development and Cohesion Policy (hereafter Cohesion Policy). Having subsequently reinforced Smart Specialisation as an enabling condition in the 2021–27 Cohesion Policy (European Commission, 2021, art. 15.1), since 2014 there has been a regulatory requirement to adopt the approach to access European regional funding for research and innovation (R&I)-related activities.

RIS3 was originally introduced to ensure more efficient rollout of the main European Structural and Investment Funds (ESIF), to derive greater ‘European added value’, and to align the direction of travel of European regional development policy with the Europe 2020 Strategy (Roller, 2011). Notably, the design and implementation has been uneven across the EU since the policy’s adoption, creating tensions between expectations and institutional reality (Benner, 2020; Morgan & Marques, 2019). This paper thus analyses how the use of conditionality as a means of policy transfer has influenced this implementation in the EU.

By employing conditions, what has been intended as an approach for regions to develop unique, bottom-up strategies, based on regional governments’ facilitating cooperation across their innovation systems (Estensoro & Larrea, 2023) – the premise of the EDP – has itself been implemented through a top-down regulatory process. In the context of recent calls for further research on issues of ‘institutional isomorphism [and] the causes and consequences of  …  the gaps between concepts and implementation’ (Benner, 2022, p. 884) of Smart Specialisation, this apparent disconnect leads us to ask: How might this mode of policy transfer shape the regional implementation of Smart Specialisation?

The literature on policy transfer has, since the 1990s, emerged as a salient approach to analysing and understanding the movement of policies in international systems. Although recent studies (e.g., Benner, 2019, 2022; Trippl et al., 2020) have made considerable conceptual and empirical inroads in unpacking the role of underlying institutional factors in RIS3’s implementation, policy transfer as a process has more rarely been studied, save for some notable exceptions (e.g., Veldhuizen & Coenen, 2022), in the context of regional innovation and industrial policy. Addressing this, we extend discussions of conditionality common to the literature on the European integration process, as well as recent contributions on directional regional innovation policy (Molica, 2024b), into regional studies by drawing on institutional policy transfer. We first posit that the use of conditionality as a mechanism is affected by the familiar sets of institutional preconditions which have hindered the implementation of Smart Specialisation. Second, we discuss whether the compelling of a common European approach to regional innovation policy, through conditionality, reflects a more fundamental change in the role of the state in steering the direction of innovation.

The rest of the paper is structured as follows. In the following sections, we first discuss the literature on policy transfer in general and conditionality in particular. Subsequently, we present the framework of Smart Specialisation and its implementation in the EU through the RIS3 policy instrument. We then analyse Smart Specialisation and RIS3 from the perspective of policy transfer. Finally, we discuss what the use of conditionality in (regional) innovation policy means for the evolving role of the state in the broader European integration project.

2. TOWARDS DIRECTIONAL INNOVATION AND INDUSTRIAL POLICY

The discussion on where the EU stands in relation to becoming a quasi-state or not is a much wider discussion than industrial policy alone. But industrial policy has traditionally been an area where member states have been reluctant to pool sovereignty; favouring not just a general lack of Europeanised intervention, but rather the retention of abilities to forward differing approaches when it comes to national intervention (Foreman-Peck, 2006). Over the years, the EU – bolstered especially by a host of contemporary challenges of late, from COVID to the need for a transition to cleaner energy generation and usage – has however striven to reanimate ideas of Europeanised approaches in this domain.

Discussions on the need for a new(er) approach to industrial policy, which began in earnest in the early 2000s (Aiginger & Rodrik, 2020; Aiginger & Sieber, 2006; Rodrik, 2004), have migrated in step with ideas for policies focusing on the state as a strategic enabler. Such facilitatory or ‘mission-driven’, entrepreneurial (Mazzucato, 2021) models consider innovation as a core pillar both for economic development (Park, 2018) and the achievement of wider societal goals – especially in the field of ‘wicked’ problems encompassing issues such as environmental and climatic change and widening social inequality (Diercks et al., 2019).

Such approaches prescribe a larger role for the state in setting the direction of the investments required for radical and/or high impact innovation, entailing relatively centralised coordination and organisational capacity to implement and coordinate activities across multiple levels of domestic (and international) governance (Larrue, 2021; Mendonça et al., 2018). This asks for transformative approaches to governance too, whereby traditional actors adopt new ways of working, and whereby the lines between public service provision and firm ambition become intertwined through more participatory governance processes (Estensoro & Larrea, 2023; Rabadjieva & Terstriep, 2021). Before we turn to Smart Specialisation as a specific case, we return to the theoretical bedrock of policy transfer. This explains the mechanisms of how transformative innovation policy can be implemented across the EU.

3. POLICY TRANSFER AND ITS INSTITUTIONAL ASPECTS

Whilst it would be hard to find a policy which has not taken any inspiration from elsewhere (James & Lodge, 2003), exactly why and how policies move and circulate between jurisdictions has been a subject of research for decades (cf. Benson & Jordan, 2011; Dolowitz & Marsh, 1996, 2012; Dussauge-Laguna, 2012; Porto de Oliveira, 2021). And whereas various frameworks have sought to explain how and why policies move and adapt across systems (Haupt, 2023), the established use of policy transfer to analyse international cases, especially at the EU level (Bulmer et al., 2007), makes it a helpful framework to analyse the implementation of RIS3 in the EU.

A vast literature identifies three primary modes of policy transfer: voluntary, mixed and coercive (Bulmer et al., 2007; Dolowitz & Marsh, 1996; Evans, 2006, 2009). The policy transfer continuum of Dolowitz and Marsh (2000) illustrates this clearly, in the authors’ identification of differing ‘degrees’ of transfer, ranging from voluntary lesson-drawing between policymakers to the coercive mechanisms of conditionality and obligation. These distinctions are rarely clean-cut, and even policymakers who look voluntarily to other policy systems likely recognise multiple forms of pressure from multiple levels to alter things at home (Chien, 2008; Holzinger & Knill, 2005).

Policy transfer can therefore be understood as a continuous, albeit multidirectional, process, centred on a host of interactions – not all of which are led by the ‘state’. For whilst systems of multilevel governance may rely on formal institutional processes to compel policy adoption at various levels, the professionalisation of policymakers around particular topics (DiMaggio & Powell, 1983) facilitates other arenas for policy transfer. The role of non-governmental actors in such processes has been established (Stone, 2000), and in particular, the role of epistemic policy communities (Haas, 1992): transnational networks of ‘socially skilled’ and issue-linked professionals (including in governmental organisations) who might facilitate the transfer of policy substance through the opening of new ‘fields’ of action across institutional boundaries (Fligstein, 1997; Haas, 1980; Stone, 2008; Veldhuizen & Coenen, 2022).

Leading theories in economic geography broadly posit that both the formal and informal ‘rules of the game’ (cf. North, 1990) impact the variegated nature of regional economic development (e.g., Bathelt & Glückler, 2014; Boschma & Capone, 2015; Boschma & Frenken, 2006; Gertler, 2010; Rodríguez-Pose, 2013). At an organisational level, institutional arrangements of a recipient government are important for the successful integration of external process and substance (Bulmer & Padgett, 2005), and more successful or meaningful transfer is perceived to occur when a policy is adopted because of its suitability to domestic conditions (Evans, 2009).

Conversely, policy transfer is often unsuccessful when transfer mechanisms demonstrate insufficient consideration to domestic institutional contexts (Benson, 2009), especially where a recipient might implement policy solely to comply. As Benner (2022) discusses in relation to S3, this might run the risk of a somewhat ‘ceremonial’ adoption of a policy by a receiving institution, whereby an approach is adopted on paper and without a deeper understanding of the nature of its substance. Hence, the mere adoption of a policy or programme due to conditional requirements can be a poor indicator of its implementation (Holzinger & Knill, 2005).

Drawing on Dolowitz and Marsh’s continuum, others (Bulmer et al., 2007; Bulmer & Padgett, 2005) introduce the notion of mixed transfer processes in certain institutional contexts, whereby elements of voluntary transfer interact with the coercive at various stages in policymaking processes. Voluntary processes have been suggested to reinforce greater legitimacy, avoiding the semblance of institutional and/or organisational power dilution at the receiving level associated with coercive means (Oliveira Martins, 2021).

This institutional turn in economic geography is mirrored in the policy transfer literature. Given the complex, recursive relationship between formal and informal institutions (Lawrence et al., 2009; Lehmann et al., 2022), discerning impacts of institutional preconditions on the mechanisms of policy transfer is a complex task. Though whereas approaches in economic geography have tended to concentrate on issues of policy substance, and broader issues of institutional fit, the institutional dimension of policy transfer emphasises issues of organisational process. This enables a greater understanding of why policymakers tweak and adjust policies to suit individual organisational conditions (Howlett, 2009; Peck, 2011). As regions are not uniform entities with policy actors following predictable cycles (cf. Brown et al., 2023), regional institutional environments might be construed as complex social processes, whereby the rationality of a certain policy approach is reconstructed and reconstrued into entirely new formations, depending on domestic conditions (Peck & Theodore, 2012).

3.1. Conditionality as a means of policy transfer

Conditionality as a means of coercive transfer has long been employed by the international community to set behavioural requirements in exchange for certain forms of military or financial assistance (Loomis, 1968); requirements which evolve in line with political and policy norms governing the agents affecting the conditions. This has been the case, for example, as international financial organisations have moved from the enforcement of primary conditionalities – those conditions primarily concerned with repayment of loans – towards what some (Babb & Carruthers, 2008) deem as those based on secondary, political conditionalities (PCs). As opposed to primary conditionalities, PCs might make additional requirements of recipients. In financial organisations, this might then extend beyond the repayment of credit. PCs can be seen as enforcing the common legal environments necessary to enforce organisational behavioural change (DiMaggio & Powell, 1983), and may include internal economic reforms, the further liberalisation of trade or the raising of domestic taxes in certain areas. They therefore encourage domestic change based on incentives and rewards for course changes and may also allow for suspension of aid or finance when no longer fulfilled (Fisher, 2015; Smith, 1998). Some have suggested the distinction between primary and secondary conditionalities as being that of macro- and micro-levels; with policy conditionalities differing from funding conditionalities, which govern grant funding (Molica, 2024b). Over time, as organisations such as the World Bank and the International Monetary Fund (IMF) have pivoted from project-based interventions towards more ‘programmatic’ financial intervention (grounded on degrees of domestic political change) (Bazbauers, 2018; Dijkstra, 2002; Fisher, 2015), the use of PCs has increasingly been established as a means of coercive policy transfer.

3.2. Policy transfer in the EU

The nature of policy transfer processes in the EU lies in the bloc’s multi-level, quasi-state-like character, where policies move between the supranational (and, in some cases, intergovernmental), national and subnational levels. Voluntary, mixed and coercive transfer are all at play within the large-scale bargaining process at the heart of Single Market lawmaking, where the EU’s core (though not exclusive) responsibilities lie. Given the variety in levels of government affected by EU policies, several types of policy transfer can be seen (Bulmer & Padgett, 2005), rooted in the EU’s tasks of regulating the Single Market through the removal of national barriers.

The voluntary, ‘horizontal’ policy coordination process (cf. Rose et al., 2013) has constituted an increasingly common means of transfer since the launch of the Lisbon Strategy in 2000 and the abolition of the EU’s three pillar structure through the treaty signed in that city a decade later. The increasing ubiquity of the voluntary – albeit persuasive – soft law process of the Open Method of Coordination (OMC) since 2000 tilled the ground for broader cooperation across a wide range of areas. The area of social policy coordination being one of the most recognised (Tholoniat, 2010), though with other policies in the fields of enterprise, innovation and research foreseen as being subject to greater overall coordination at the member state level – notably, for example, through the establishment of a European Higher Education Area through the Bologna Process, as well as the ongoing moves to create a wider European Research Area (Armstrong, 2016; Benson & Jordan, 2011).

Meanwhile, a more ‘vertical’ (hierarchical) policy implementation is also underway through the formalised transfer process through legislative acts between the EU regulatory bodies in the ordinary legislative procedure (OLP) (Larouche, 2004). Within this more hierarchical strand – which covers most EU legislation with regards to the completion of the Single Market – different modes of policy transfer from across the policy continuum are often combined within single policy instruments or approaches (Bulmer et al., 2007).

3.2.1. Conditionality as a form of policy transfer in the EU

The EU, of course, is not an international financial organisation. However, the development of the Single Market itself in the 1980s can be characterised as the result of a paradigmatic shift in the role of international organisations in conducting external development by means of conditionality. In the context of such a shift – arguably epitomised by John Williamson’s ‘Washington Consensus’ principles for development aid in the early 1990s, (Babb, 2013) – the EU has been a consistent adherent to the application of conditionalities since the 1990s (de Felice, 2015). Contemporary PC has thus been argued (Babb & Carruthers, 2008) to encompass an ever-expanding range of issues, with its application at the EU level having been influenced by the arguable consolidation of elements of the Washington Consensus into the Maastricht Treaty (cf. Fligstein, 1997; Monteverdi, 2017; Lord, 2000; Ross & Jenson, 2017).

EU membership is conditional on a state’s changing of its internal institutional practices – the so-called ‘Europeanisation’ process – in exchange for the potential benefits of being in the organisation. In its external relations, PCs are mechanisms for the EU to Europeanise the systems of neighbouring states and regions in areas such as foreign policy, trade, climate and energy, and decentralisation (cf. Hughes et al., 2004; Kristensen & Pugh, 2015; Vukov, 2020; Bradford, 2020).

Internally, the increased use of conditionalities as mechanisms for policy transfer and adoption within the EU’s policy process reflects change in the EU’s internal relations, towards a more conditionality-based culture more broadly (Bachtler & Mendez, 2020; Bini Smaghi, 2015; Viță, 2017), related to the dispersal of EU funding through the multi-annual financial frameworks (MFFs). Since the 1990s, the bloc has increasingly relied on the application of conditionalities of various forms to overcome barriers to Single Market completion (Babb & Carruthers, 2008; Mattelaer, 2018). One of these has been the increasing usage internally in the EU of macro-economic and ex ante conditionalities (Jorge Nuñez Ferrer et al., 2018), of which the prime objectives are to align the actual spend of European funds with those broader policy objectives to which EU states have voluntarily agreed.

4. SMART SPECIALISATION, S3 AND RIS3

One area with a strong subnational policy transfer dimension is the EU’s recent regional innovation policy, broadly grounded in Smart Specialisation. The genesis of Smart Specialisation as a policy concept is widely regarded to be the workings of the Knowledge for Growth Expert Group (KFG) in the 2000s, an epistemic community around Smart Specialisation. It can be seen as a policy which seeks institutional change at the regional level (Benner, 2019, 2022), which for some places, inevitably, means advocating a degree of discontinuous change (North, 1990).

Whilst aspects of Smart Specialisation Strategies (S3) are no doubt transformational, Smart Specialisation as a concept more widely has been described as dichotomously embodying elements of linear and systemic innovation approaches (Coenen & Morgan, 2020; Morgan, 2019). Indeed, one of the more theoretically challenging aspects in designing a Smart Specialisation policy at the regional level has been in going beyond horizontal policies towards something more directional (McCann & Ortega-Argilés, 2019).

The Smart Specialisation approach holds that public R&I-orientated spend in a given territory ought to be orientated towards where unique comparative economic potential rests (Foray et al., 2011). The identification of priorities should build on an objective analysis of a regional economy’s assets and potential, via a bottom-up participatory process, facilitated by a capable public sector (Estensoro & Larrea, 2023). The latter is achieved through an EDP – a process which has been posited as being more important than the outcome of the strategy itself (Lehmann et al., 2022) – whereby actors in a regional innovation system cooperate to identify new opportunities for a region to develop competitive advantage. Application of the Smart Specialisation principles should serve to overcome the tendency for regions to follow similar paths when designing policy priorities (Barca et al., 2012; Navarro et al., 2014; Tödtling & Trippl, 2005). It is thus the taking place of a diversification process in a regional economy centred on an EDP (Foray, 2014; Foray et al., 2009).

Although ‘S3’ is commonly used to refer to Smart Specialisation, we suggest emphasising the difference between Smart Specialisation and S3. An S3 can be the output of an EDP, which is at the core (the ‘substance’) of the Smart Specialisation concept. Its overarching goal is to ‘set priorities in order to build competitive advantage’ (European Commission, 2013, art. 2.3). The strategy is typically a policy document or equivalent, which should be constantly monitored and renewed (e.g., Laranja, 2022) and should serve as guidance for the implementation of (R&I-focused) ERDF at either the regional or national level. Still, the document ought not be considered as the sole, conclusive component of a region’s innovation policy, as many regions may codify the outputs of entrepreneurial discovery through myriad different instruments (Charles et al., 2012).

Since 2014, in the last and current MFFs, the EU has, broadly speaking, required as conditionalities that national and/or subnational authorities design and implement an S3. In EU parlance, RIS3 can be seen as the consolidation of the Smart Specialisation policy concept into a concrete policy instrument. And as Hudson (2015) prompts, policy instruments have three principal components: processes, substance and politics; all subject to distinct institutional and organisational conditions.

As much of the literature examines its application in the EU, concepts of Smart Specialisation and the strategies themselves are often conflated with the specific EU policy instrument, RIS3, regulating the process. We argue here that RIS3 can be seen as a process; the requirement to formulate a strategy as opposed to its formulation (the policy substance) or its actual (more political) content (cf. Benner, 2020). This is a process that requires different evaluation criteria than those applied to the substance of the strategy itself (cf. Veldhuizen & Coenen, 2022). This conflation creates ambiguity, as scholars and policymakers alike seek new ways of evaluating why RIS3 may or may not be delivering.

5. THE POLICY TRANSFER OF SMART SPECIALISATION

Having introduced the literature on policy transfer and the Smart Specialisation policy concept, we now turn to the question of how the EU’s chosen mode of policy transfer – the RIS3 ex ante conditionality – has influenced the implementation of Smart Specialisation by regional governments. As rationales for the transfer of policy processes can diverge from the underpinnings of a policy’s substance or its politics, varying policy transfer rationales may underpin the nature of RIS3 adoption in different institutional contexts. Given the centrality of innovation as an accepted component of economic growth and development, this policy has therefore placed considerable conditions – and a degree of coercion – on a policy area over which (some) local and regional administrations have otherwise had substantial levels of substance autonomy.

Smart Specialisation was a policy idea and not an established policy practice in place at the member state or regional level before its introduction into Cohesion Policy. In introducing the concept, the European Commission arguably played the role of a ‘policy inseminator’ (Radaelli, 2000) in drawing lessons from the KFG epistemic community. Once adopted at a policy level within the European Commission’s proposals for a reformed Cohesion Policy in 2013, RIS3 was subject to the Single Market’s negotiated decision-making processes based on qualified majority voting in the OLP (Becker, 2019, p. 153). The modes of policy transfer involved in the EU’s adoption of Smart Specialisation thus range from lesson-drawing, through the mixed (negotiated through the EU’s OLP) transfer in the formal process of legislative adoption, to a more coercive form of policymaking by means of conditionality in the RIS3 instrument.

5.1. How has the use of conditionality shaped the implementation of RIS3?

The high degree of ambition inherent to Smart Specialisation implies the need for both formal and informal institutional capacity at the regional level in the field of participatory governance (grounded in the space of voluntary lesson-drawing at the regional level), as opposed to ‘negative’ forms of governance based on regulatory compliance. The potential power of voluntary transfer can be illustrated, for example, in that several member states and regions have adopted other approaches championed by the EU, such as missions-based frameworks (e.g., Hekkert et al., 2020), outwith any EU requirement to do so. Conditionalities, however, imply a coercive logic different to the ambitious, directional state-as-facilitator role which underpins the governance of the EDP.

Within the field of policy transfer, conditionalities – although not always direct coercion – have thus been considered a form of ‘negative integration’. This creates a disconnect: Smart Specialisation as a concept has been built on voluntary knowledge exchange guided by capable institutions at the regional level (Estensoro & Larrea, 2023), favouring horizontal transfer by governments and non-governmental actors. Meanwhile, its implementation in the form of RIS3 has relied on a degree of coercion – as, theoretically, a region’s failure to comply results in withholding of ERDF. Recalling the policy transfer continuum, the theoretical underpinnings of the Smart Specialisation policy concept situate the EDP within the sphere of voluntary transfer, with an element of mixed transfer at the regional level due to hierarchies within the regional innovation systems. A tension exists then between this voluntary, dynamic EDP and the ‘stability and predictability’ (Molica, 2024b, p. 6) required by the conditionality, which places the RIS3 instrument towards the coercive end of Dolowitz and Marsh’s range.

The effectiveness of PC as an instrument of policy transfer for Smart Specialisation has also been mixed. Not all regions developed a tailored S3 during the period 2014–20. Some submitted other innovation policy documents, or wider economic development strategies, to comply with the conditionality. Yet, this has not had major consequences. In theory, funds can be withheld, and infringement procedures started (Koch, 2022), although regions (or member states) have been unlikely to be sanctioned due to non-compliance with the RIS3 process. The period 2014–20 also saw a lack of enforcement of PC more widely at the EU level, and the ‘softness’ of the implementation of conditionality can be seen within a broader context of the degrees of discretion traditionally afforded to member state authorities in Cohesion Policy’s implementation (Bachtler et al., 2014).

5.1.1. Institutional conditions as a limit to conditionality

Benner (2022) notes that much of the process of adoption and implementation of RIS3 has tended to focus on ‘ceremonial’ adoptions of the policy to comply with the ExAC, something compounded by the ambiguity in the Smart Specialisation concept and myriad differences in understanding its implementation processes (Marques & Morgan, 2021). RIS3 is complex: underpinned by a theory which ideally sees a regional innovation system manage two concurrent functions of governance simultaneously (Esparza-Masana, 2022). First, the EDP advocates a substantial degree of participatory governance. For some governments, this may involve considerable updating and reforming of entrenched and established practices. Second, the Smart Specialisation process further requires a government to maximise opportunities within a complex regional innovation policy mix and regulatory landscape, administered across multiple levels of government (Magro & Wilson, 2019). Such complex legislative and policy environments tend to function as a counterforce to innovation more generally (Hollanders & Wintjes, 2003).

Issues of institutional capacity are oft addressed within the prevailing literature on RIS3. Critique has considered this confusing, complex nature of Smart Specialisation; and namely, its being unsuited to regions with lower levels of administrative capacities, such as economies with exhibiting lower levels of public sector coordination (Lehmann et al., 2022). Other studies note its advocating the adoption of a ‘one-size-fits-all’ process (Benner, 2020; Hassink & Gong, 2019), if not content. ‘Lagging’ and/or ‘peripheral’ regions are associated with lower absorption potentials for complex innovation policies (Hollanders & Wintjes, 2003), and can struggle with implementation of the EDP (Papamichail et al., 2023). Less-developed regions are often said to face challenges in regional innovation system governance, interregional collaboration, and administrative issues critical to the implementation of Smart Specialisation (Wibisono, 2022). This furthers the question around some regions’ abilities to fulfil the condition’s substantive nature – the implementation of transformative innovation policy.

In some cases, as Lehmann et al. (2022) note in their study of Smart Specialisation in Bosnia and Hercegovina (a country outwith the ExAC requirements), deficiencies in both formal and informal institutions make the introduction of the EDP, a process which could help build institutional capacity, particularly difficult. This is in line with the ‘regional innovation paradox’, with those places’ most in need of innovation capacity being typically those with lower overall institutional and/or organisational capacity to design and implement more far-reaching policies (Esparza-Masana, 2022; Oughton et al., 2002; Trippl et al., 2020).

Regional development policies have been suggested to work best when they attempt not to change the idiosyncrasies of domestic institutional culture in a prescriptive manner (Rodríguez-Pose, 2013). But might ‘an institutional culture’ be more orientated towards a compliance process or the nature of the substance? From the EU accession literature dealing with the uptake of the aquis (cf. Bachtler et al., 2014; Hughes et al., 2004; Vukov, 2020), the level of ‘fitness’ at an institutional level may be more in step with organisational abilities to comply with conditionality (Grabbe, 2002); a rationalist, cost–benefit motivation for access to the wider benefits of being in the EU (Sedelmeier, 2016) rather than broader political or institutional commitment to transform ways of doing things. By contrast, some (Prange, 2008) have asserted that institutional fit with the European model itself will have an impact on regions’ ability to implement an EU regional innovation policy, as a certain degree of legitimacy will be associated with it.

In line with policy transfer theory, RIS3’s status as a budgetary conditionality without clear sanctions (Bulmer & Padgett, 2005; Dolowitz & Marsh, 2000; Evans, 2009) creates confusion as to whether the EDP is an institutionally legitimate or relevant concept for industrial policy across many of the EU’s member states and regions. As an example, central and eastern member states are home to many of the least-performing regions when it comes to innovation (Hollanders & Es-Sadki, 2023). However, they may be also amongst those places with the greatest institutional ‘muscle memory’ of implementing conditionalities, which were heavily imposed on the new member states during the accession process. But levels of broader post-accession aquis compliance of newer member states (including many of those often associated with low qualities of governance in the Smart Specialisation literature) have been shown to be higher in certain cases than in the EU15 (Sedelmeier, 2016). Conversely, when conditionalities within previous iterations of Cohesion Policy have been imposed, some member states with more administrative capacity have tended to view conditionalities as being cumbersome and adding minimal value to domestic approaches (Hollanders & Wintjes, 2003; Jorge Nuñez Ferrer et al., 2018; Koch, 2022).

Turning back to RIS3, certain lagging and/or peripheral regions have demonstrated compliance with the requirements to put strategies in place (e.g., Healy, 2016; Potter & Lawton Smith, 2019). Molica’s (2024b) consideration of a Calabrian ESIF operational programme supports this; demonstrating the ability of a low-income region in applying a micro-level funding conditionality drive directionality. As has also been shown, more-developed regions with highly entrenched innovation policy approaches may be more prone to eschewing EU RIS3 conditionality. This might be due to existing path dependencies, or a ‘lack of space’ in the policymaking process to countenance transformative approaches, as demonstrated by Kristensen and Pugh (2023) in their study of Central Jutland. Thus, we might distinguish between those member states with institutional cultures which are adept at applying transformative innovation approaches, and those whose governmental bodies have substantial capacities to transpose regulation. Hence, a nuance emerges as to whether institutional capacity relates to the abilities to inform and execute an EDP and the resulting Smart Specialisation process, or whether it relates to the ability of a government department to comply with conditionalities.

6. POLICY IMPLICATIONS: ENTER THE (EUROPEAN) ENTREPRENEURIAL STATE?

So, does the use of conditionality reflect a broader emergence in the EU of a Mazzucatian entrepreneurial state? The increased focus on internal conditionalities within the Cohesion Policy in recent years has served to highlight the challenges surrounding the implementation of RIS3, particularly around regional institutional capabilities, and issues surrounding the application of the RIS3 ExAC in those regions most affected by the ‘regional innovation paradox’.

Broader mission-oriented innovation policy approaches have increasingly been seen as a core way to tilt state intervention towards more sustainable development with a view to addressing the ‘wicked problems’ often posed in policy dialogues. As these policies evolve, consideration should be given to the evidence concerning existing policy transfer in the context of RIS3. The adoption of RIS3 is the first time in the history of the European project – if one excludes the European Coal and Steel Community in the 1950s as an industrial policy endeavour in itself (Aiginger & Sieber, 2006) – that a particular type of policy approach in this area has been obligated on member states as a conditionality of drawing down European funding. The application of Smart Specialisation as a legal condition has thus begot a change in EU intervention in the economic development policy approaches of its member states.

At the EU level, outcomes of this can be seen in the bloc’s taking forwards transformative approaches on grand societal challenges (cf. Schot & Steinmueller, 2018; Lundvall, 2023), where a clear result of this has been the uptake of missions-orientated actions within programmes such as Horizon Europe (Mazzucato, 2018). More recently, policy directions such as Partnerships for Regional Innovation (PRI) have advocated for greater directionality in the realm of regional innovation policy, building upon and extending the scope of policies such as Smart Specialisation (Kelchtermans et al., 2022). This is arguably also being reflected in longer term policy discussions around the provision for coordinated and ‘synergetic’ cooperation in future EU programmes between competitive instruments such as the Framework Programmes, and the (mainly allocated) ESIF (e.g., Jokelainen & Novo Guerrero, 2023).

In terms of additional insight in preparing the next generation of RIS3 policy instruments, this raises a critical issue for further research regarding the transferability and the applicability of RIS3 across different regional typologies. Some argue that Smart Specialisation is an appropriate framework for taking forwards transformative approaches (Foray, 2018), giving rise to a need for further debate on the role of the EU’s Cohesion Policy (Molica, 2024a). But the question of whether and how such directional approaches translate to entities below the level of the member state (and larger ones at that) highlights well-worn issues (Benner, 2020, 2022; Hassink & Gong, 2019) in adopting, adapting, and implementing transformational approaches. Recent discussions around challenge-orientated innovation systems (e.g., Tödtling et al., 2022) have highlighted potential frameworks to move forwards here, as have contributions on more localised solutions to societal challenges (e.g., Bours et al., 2022; Wanzenböck & Frenken, 2020). Others have noted that the governance of Smart Specialisation processes has some way to go before transformative innovation policy processes can be integrated (Reid et al., 2023), especially as the rise in ambition has tended to map a decline in public resources more generally (Coenen & Morgan, 2020).

The conditional implementation of RIS3 has reflected a changing role of the EU in setting innovation and industrial policy – from the fixing of market failures in favour of guiding societal transformation (cf. Borrás & Edler, 2020). The use of conditionality has evolved over time from the enforcement of primary financial conditionalities to secondary PC and is increasingly being employed as a potential tool for industrial policy (Mazzucato & Rodrik, 2023), reflecting a broader, evolving role of supranational governance structures in steering the direction of industrial policymaking. Its use in the field of regional innovation policy may reflect an evolution from secondary conditionality towards a third form; going beyond the traditional nature of EU industrial and innovation policy interventions. With an emphasis on certain methodologies for transformation (e.g., Foray et al., 2012), such an extended concept has ushered in a new top-down methodological standard for how regional institutional reconfiguration ought to take place (e.g., Benner, 2020). The additional directionality inherent to Smart Specialisation entails a new requirement for complex policy redesigns at the regional level to inseminate the approach across multiple territories of the EU’s member states.

Compliance with conditions tends to be smoother if the required approach in question has a greater local legitimacy in terms of its process, substance, or political aims (Sedelmeier, 2016). It is by far from clear that the retention of universal conditions in Cohesion Policy, which require such directional transformation, can succeed without substantial redesign considering local institutional structures (Molica, 2024b), especially given the heterogeneity of European regional institutional structures and industrial-innovation policy traditions (Kölling, 2023). Thus, increased reliance on conditions might risk creating a degree of domestic institutional imbalance, between the requirements of the formal ExAC and the nature of informal institutions already present (Lehmann et al., 2022). This is particularly pertinent when considering activities for building on the 2014–20 RIS3 experience, given the context of recent policy discussions around sustainable S3, the subsequent moves at the EU level to create innovation-driven territorial transformation through the PRI policy approach (Kelchtermans et al., 2022) and uptake of the enabling conditions in the current 2021–27 MFF.

7. CONCLUSIONS

This paper has examined the policy transfer of Smart Specialisation. We show how the EU’s use of conditionality as a mode of policy transfer is at odds with the logic of lesson-drawing and peer learning inherent to the theory of Smart Specialisation. In line with discussions on the limitations of conditionality (e.g., Molica, 2024b, p. 4), we contend that this has contributed to variegated implementation of the policy, as enforcement has been rather soft. Furthermore, we forward the discussion on the role of institutional conditions in understanding the varying degrees of implementation across different regions. Notably, there is a disconnect between the regions which are good at complying with conditionalities, those which have the institutional capacity to implement a complex policy such as Smart Specialisation, and those which are the most in need of innovation-led transformation. We discuss whether the use of conditionality in the field of innovation policy reflects a new, tertiary, type of conditionality and even the emergence of an entrepreneurial state at the EU level.

Although small given the overall size of the EU’s budget, given RIS3’s inherent grounding in the ‘third paradigm’ of innovation, this reflects a new development in the European integration project. In the long history of EU R&I policy, from the first Framework Programme onwards, RIS3 marks the first time a regional approach to industrial policy has been legally compelled. Recalling what can realistically be expected from innovation policy and transformational innovation policy (e.g., Laatsit et al., 2022), the use of conditionality in directional innovation changes what conditionality in Cohesion Policy might reasonably be expected to accomplish.

Although the nature of this ‘tertiary’ conditionality does not denote a new method of policy transfer, it has to some extent obligated a new and experimental approach in what was primarily a voluntary method of transfer through the OMC – that of enterprise and innovation policy (Armstrong, 2016). In line with Fagerberg’s (2018) reflection, the capital, resource base and capacity required to effectively implement a transformative innovation approach may entail its limitation to those countries, or larger, advanced regions with the institutional ability and competence, to undertake such policy projects. This marks a core issue at the heart of the problem of RIS3 implementation: the ability and competence to implement and respond to a conditional obligation does not necessarily equate with the ability and competence to develop a transformative regional innovation policy. Whilst the former is a question of the policy process, the latter is a question of the policy substance.

ACKNOWLEDGEMENTS

The authors are grateful to the anonymous reviewers for their valuable comments, and to those who have extended their feedback on this work during conference and workshop presentations. Particular thanks to Professor Ron Boschma and Éva Somogyiné dr. Komlósi for their constructive comments and criticisms of earlier versions of this paper.

DISCLOSURE STATEMENT

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by the European Union’s Horizon 2020 Research and Innovation programme under the Marie Skłodowska-Curie Actions [grant number 860887]. The views and opinions expressed here are those of the authors alone and do not necessarily reflect those of the European Union. Neither the European Union nor the granting authority can be held responsible for them.