Who makes European Cohesion Policy: a practitioners’ learning perspective

Abstract Through the European Social Fund (ESF), the European Union subsidizes social inclusion and occupational trainee projects. This policy instrument is formulated by European Union institutions and implemented by regions through a call for projects which requires innovative actions and a result-oriented strategy. It is a key vantage point to observe sub-national implementation of an European Union policy instrument in a sub-national practitioner’s perspective, a topic under-investigated in the literature. For a project to receive funding, sub-national practitioners must take the European Union requirements (accounting standards, evidence of innovation etc.) and their social needs into account. Against this backdrop, this research tackles an original question: how can local practitioners adapt their policy actions to fit with European Union requirements? Taking the region of Wallonia in Belgium as a case study, the ESF implementation is considered as a specific policy process in sub-national policy sectors. Adopting an in-depth qualitative perspective, this research contends that it depends on practitioners’ learning, practices and experience of past applications, their specific socio-economic context and income maximization.


INTRODUCTION
Since 1957, the European Social Fund (ESF) has supported local development projects. It has evolved over time to reflect European-wide concerns, such as subsidiarity, additionality and partnership principles (1987), the European Employment Strategy and competitive governance tools (1997), and 'investment in human capital' following EU2020 (from 2014). These requirements were designed in Brussels in closed policy networks and imposed on sub-national implementing authorities. As the ESF demands governance based on competition, performance and sanction (Halpern, Lascoumes, & Le Galès, 2014) alongside local activation (López-Santana, 2015), sub-national authorities and practitioners must adapt European Union (EU) requirements to their local needs.
Regional StudieS, Regional Science Contemporary research has primarily focused on how EU decisions affect local implementation (e.g., Dąbrowski, 2014;Milio, 2007;Perron, 2014). ESF decisions as exogenous interventions frame rather than direct local activities. Local practitioners here play a key role defining which policy actions will be funded, translating regional objectives into policy projects, and in implementation manipulating EU requirements to meet their local social action needs. I ask the question: how do local practitioners adapt their policy actions to fit with EU requirements from a policy design process, and therefore contribute to Bachtler and Mendez's (2007) discussion of who governs cohesion policy?
This opens the black box of ESF implementation to understand firstly how EU requirements (as innovation or partnership) are operationalized in local policy actions, and specific sub-national policy process influences the EU policy instrument. Much research has followed policy perspectives that envisage top-down Europeanization rather than strong local agency, and has therefore primarily focused on instrumental production of results and outcomes rather than considering these local learning processes. I specifically focus on learning processes (creating practices and knowledge) associated with sub-national practitioners' local adaptation mechanisms around ESF requirements.
I consider whether these local learning processes can influence the overall trajectory of European policy from a bottom-up Europeanization perspective. There are two potential causal mechanisms: firstly, an incremental EU-wide effect constituted by action on multiple member states, following the argument that Europeanization can be defined as a gradual and incremental process (Coman, 2014, p. 4); and, secondly, following Bachtler and Mendez's (2007) 'renationalisation of Cohesion Policy' thesis critique, national levels produce local policy actions on behalf of Cohesion Policy through which ESF governance shapes sub-national policy process.

REGIONAL IMPLEMENTATION AS A POLICY LEARNING PROCESS
The Cohesion Policy literature has tended to distinguish the European formulation level alongside sub-national policy implementation (Bachtler, Mendez, & Wishlade, 2013) addressing how EU regulations affect sub-national authorities, administrations and local organizations. Many evaluations of ESF interventions' macroeconomic impact find significant regional differences (Ederveen, Groot, & Nahuis, 2006). 'Place' (sub-national context) clearly matters in ESF implementation and I conceptualize these local contexts, unable to influence ex ante regulation formulation, influencing ex post implementation. Following Halpern (2011), EU policy instruments can be analyzed without referring to their supranational origin because they gain autonomy in the domestic realm. In the ESF, central governments acted as gatekeepers (Bache, 1999) mediating between the formulating EU institutions and implementing sub-national authorities (respecting the subsidiarity principle). But there has been little consideration of the processes by which EU policy instruments gain in autonomy and which mechanisms underlying this process. Considering ESF implementation as a specific sub-national (if post-hoc) policy formulation process offers an alternative perspective on theories of Europeanization by separating the instrument's EU origin from its domestic implementation.
EU regulations see ESF being operationalized through local welfare projects addressing vocational training or social inclusion. The 'project' as a governance tool contributes to translating European policy objectives (as articulated in regional operational programmes) into local policy actions. I here consider ESF regulations as framing these regional implementations, which can be defined as specific non-linear and unpredictable local policy processes (following Lane, 1987, p. 263) (Figure 1). The managing authority handles the formulation of ESF sub-national objectives, then local practitioners answer the call for projects and define policy actions (step 2). Projects are monitored by a permanent interaction between the managing authorities and the local  practitioners (step 3). This research focuses on step 2, practitioner project formulation translating ESF policy objectives into policy actions.
Sub-national practitioners formulate projects such as those developing basic and employability skills for disadvantage young learners, 1 promoting Roma inclusion, 2 or offering jobs in the non-profit sector for people close to their pension. 3 This project-based implementation at the same time has to meet with the statutory EU requirements, such as partnership, innovative actions and funding additionality. Local practitioners have to draw projects that fit with those requirements and with their own needs on the ground. They have to adapt their daily actions to the EU application form and accounting standards in a specific regional institutional framework and with 'large workloads, tightly constrained budgets and multiple and competing demands' (Arnold, 2014, 392).
Regional ESF implementation therefore clearly depends on how practitioners learn how to manage and cope with EU specific requirements. I use May's (1992) definition of this learning process, making a twofold distinction between: • social learning: 'lessons about social construction of policy problems, the scope of policy, or policy goals'; and • instrumental learning defined as 'lessons about the viability of policy instruments or implementation designs'.
Such a learning process occurs through practitioners' experience of the ESF, their beliefs about the policy process, their own needs and the specific socio-economic context they face. This dual-learning process may involve trial-and-error behaviour (Lindblom, 1990), imitation of existing applications through mimicking behaviour (May, 1992), compliance with EU requirements, or competition of ideas (Sabatier, 1988) and possible resistance (Saurugger & Terpan, 2013).

CASE STUDY CONTEXT
Studying practitioners' social and instrumental policy learning is not reproducible in other policy sectors or times (Howlett, 2014), requiring in-depth case studies and in a specific socio-economic and institutional context. I use a case study of Wallonia, Belgium (Figure 2), 4 itself a founder member of the EU, where past policy learning associated with ESF implementation has been important. With a gross domestic product (GDP) of 75-90% of the EU average, Wallonia receives Structural Funds for transitional regions (2014-20 programme). With an unemployment rate around 12.3%, Wallonia uses ESF primarily for social and occupational training and initial social integration. The case study focuses on sub-national practitioners in these two sectors and who applied for ESF during of the last two programming periods (2007-13 and 2014-20).
In line with the ESF regulation, the Wallonia Region is the managing authority responsible for the allocation of its ESF budget to welfare practitioners within its territory. To do so, Wallonia organizes a call for projects; therefore, it is the key ESF implementation tool of the ESF, following principles of competition, performance and sanction. The competition call in turn defines the main investments priorities (policy goals) and eligibility criteria; investment priorities are defined in close coordination with the European Commission and validated as the 'operational program' (OP). Eligibility criteria are based on the regional accreditation system designating practitioners, such as 'trainee on the job' or 'social inclusion' centres. This system influences partnership-building: practitioners must find eligible partners willing to cover an additional element of the total project budget (following EU additionality requirements), creating opportunities but also institutional constraints. The implementation process is tailored to EU policy goals (via the OP), and the call for projects frames practitioner's opportunities and constraints for organizations.
For this particular case study, the research draws on 30 semi-structured interviews with local practitioners managing an ESF project (conducted in January-March 2014 and January-June Regional StudieS, Regional Science 2015), participant and direct observation sessions, and documentary research of ESF regional programmes and projects. Practitioners were questioned about their own individual experiences and practices (Considine, 2012) of partnerships, project management and formulation. Using NVivo10 as the coding software, practitioner's quotes were coded as social or instrumental learning following May (1992): practitioners talking about project formulation or internal organization to develop projects were coded as social learning; whilst talk of project calls as allocation, partnership-building, financial or administrative requirements were coded as instrumental learning.

WHEN PRACTITIONERS MANAGE AN ESF PROJECT
Practitioners apply for projects in order to increase their human and financial resources, driving a combination of instrumental and social-learning processes. Practitioners must find eligible partners all willing to contribute matching project resources to provide a part of the global budget for the projects. Partners must also agree on the policy action (rather, the details of particular individual outcomes that are delivered by individual actors). It was reported that the organization with most ESF experience took the lead to improve success chances: Each project has a slightly different origin but the project leader will be the organization that have the specific skills to manage an ESF project. I mean financial aspects, writing applications, administrative management, etc. A European project is not the same management logic than as usual for a director or a trainer. (director, 25 May 2015) Finding partners and agreeing those different policy crucial dimensions is not easy. Therefore, in many organizations, practitioners formulate projects with former or existing local partners with whom they had complementarities (one partner worked on inclusion programmes, another as an on-the-job-trainee). Partnerships are built reflecting past trust and confidence drawing on accumulated shared mutual experiences, based on instrumental learning where practitioners learn lessons regarding ESF viability and eligibility: Being a small practitioner, we are already in a partnership logic. The ESF brings more convergence and complementarity between long terms partners, on a close territory.
[…] In order to increase our social actions, we manage one project dedicated to specific categories, and one project that covers a broader public. (social assistant, 2 February 2014) Partners must then meet regularly to prepare a proposal, often even before the official start of the call for projects, because of extremely short application periods. They adapt their desired projects to specific OP requirements and application forms. Interviewees argued that project ideas were driven from specific needs such as particular trainee courses or inclusion programmes for unqualified workers) or indeed emergency situations (e.g., following a major industrial relocation). Firstly, they define their needs, then shape their ideas in order to fit with EU requirements and forms, reflecting a micro-process of policy translation from local observations and needs to EU innovative and sustainable project management and tools: At first, it was only coordination and social action. Gradually, we learned to make diagnosis. We had projects with universities that have taught us to gather our data and develop better pedagogical tools. […] The innovation is in our pedagogical tools. (director, 23 March 2015) We have working methods, objectives, we know the socio-economic reality of our beneficiaries, and therefore, we check how the OP objectives match with our actions. Fortunately, in the new OP we can ask subsidies for a project in four axes […] the ESF strengthens our activities. If we had not the ESF, we could not develop and maintain such a broad spectrum of policy actions. (trainer, 15 April 2015) Regional StudieS, Regional Science Finally, in formulation, feasibility must be tested with trainees, even more so if an innovative project. Project design at this stage is more of a social-learning process and crucial because feasibility is a key factor determining whether or not to apply for an ESF grant. Practitioners ensure their outcomes in order to cope with policy goals and to test the viability of their policy actions: Writing applications is the crucial step. We can only do it on the basis of the reality you deal with on the ground […] then, the last step is to test the feasibility with the trainees. If you set up a project and nobody wants to participate, your project failed. (trainer, 5 March 2014) These social and instrumental policy-learning processes combine to effect ground-level shaping of implementation. Following that, the practitioners may have to support and mentor workers with significant labour market disadvantages that demand more time than ever. This requires other kinds of subsidy from Wallonia, seeing the ESF integrated into a single delivery but reported on against the accounting standards, administrative tools and target groups of each subsidy stream. The ESF becomes juxtaposed alongside other funding instruments with collective single-delivery activities.

CONCLUSIONS
Welfare practitioners highlight two key mechanisms for observing sub-national implementation. First, applying to calls for projects demands that more experienced partners lead, their instrumental and social learning signalling to other partners a likelihood of winning this complicated funding. These partnerships are based on sharing knowledge and practices among experienced practitioners and other partners, with the dual learning process functioning as a collective risk-management mechanism. Second, dual-learning processes influence project formulation at the practitioner's level. They analyze their needs and define a specific policy action to be funded by the ESF, then translate this action to ESF vocabulary, forms and goals, thereby putting the ESF into action and operationalizing the Cohesion Policy's macro-socio-economic goals. Project formulation is determined by the partnership-building capacity, local needs and former experience of applying and managing an ESF grant, revealing how EU requirements affect a local practitioner's participation in the Cohesion Policy process, but also how EU administrative and financial requirements frame practitioner action. This helps address the question of who determines Cohesion Policy. EU institutions formulate ESF institutional frameworks and objectives, but policy content is largely defined at the sub-national level by local practitioners who create policy content and translate targets and requirements into actions. Social and instrumental learning are one mechanism of incremental changes at the sub-national level influenced by EU instruments, and it would be helpful to understand how these learning models function outside European policy (for welfare policies decided nationally or regionally). Further ESF governance modes could be compared with other sectoral tools and regulations in an instrumentation analysis (Halpern et al., 2014). Finally, a similar approach applied to other cases could explore the sub-national variation of constraints and opportunities for practitioners.
Finally, because EU institutions have institutionalized project-based implementation, it is significant to discover that it functions as a sub-national policy process, with the EU apparently here outsourcing its actions' policy legitimacy to local actors who become policy-producers formulating ESF policy action content. It appears to reduce the risk of failure whilst minimizing criticism of EU policy investments goals and controls. The continual adaptation of regulations at the European level does stimulate these local learning processes even though their requirements may be diluted by multilevel requirements. This paradox is significant for understanding Cohesion Policy: when EU requirements frame the ESF policy process from the European regulation to