Understanding citizen perception of European Union Cohesion Policy: the role of the local context

ABSTRACT

The way in which Cohesion Policy is perceived by citizens is a crucial issue for the process of European identity-building. Based on the idea that citizens’ perceptions depend on the local socioeconomic context in which Cohesion Policy is implemented, the paper seeks to define alternative combinations of the economic, social and institutional features of different local policy implementation settings, and to identify them empirically in European NUTS-2 regions. The results highlight a broad variety of policy settings, whose characteristics are relevant to the outcome of Cohesion Policy implementation.

Video abstract

Read the transcript

Watch the video on Vimeo

KEYWORDS:

• perception of policies
• Cohesion Policy
• regional policy settings

JEL:

• E61
• O10
• R10

INTRODUCTION

The way in which policies are perceived by local citizens is an interesting issue, and it has gained increasing attention since the British referendum on Brexit. Many works, in fact, highlight the role of European Union (EU) actions, such as the Erasmus programme (Mitchell, 2015) and the introduction of a common currency (Risse, 2003), in promoting citizens’ identification with EU values. The same mechanism is expected to apply in the case of Cohesion Policy, due to its objectives, mainly focused on support for regional development, and its financial resources, about one-third of the EU budget.

Compared with other programmes, however, Cohesion Policy is implemented in partnership with local governments. Therefore, the perception of the outcomes of these policies is mediated by the conditions of the settings in which they are undertaken. In other words – and this is the thesis of this paper – the way in which policies are perceived by citizens depends closely on local conditions. Previous studies have already discussed the role of certain characteristics of implementation settings on the objective outcome of Cohesion Policy actions, i.e., on the socioeconomic development of regions. Examples are provided by studies on the place-based approach to Cohesion Policy (Barca, 2009) or on the effect of institutional quality on policy results (Ketterer & Rodríguez-Pose, 2016; Milio, 2007).

With respect to these studies, this paper has two main innovative goals. First, it intends to define the local characteristics that are assumed to affect citizens’ perception of Cohesion Policy actions. Subjective factors drawn from political science, social psychology and sociology are taken into account, together with objective conditions of regions that are in general conceived as filters between Cohesion Policy implementations and outcomes like gross domestic product (GDP) and employment growth (Ederveen, Groot, & Nahuis, 2006; Fratesi & Perucca, 2014).

Moreover, rather than considering individual characteristics (e.g., regional needs, institutional quality etc.), the paper seeks to identify combinations of socioeconomic, political and institutional factors that condition how Cohesion Policy is implemented and influence how it is perceived. The presence of high-quality institutions, for instance, does not necessarily generate the same citizens’ perceptions under different degrees of Euroscepticism of local authorities. Similarly, Cohesion Policy actions aimed at satisfying objective needs are likely to be positively perceived by citizens only if they prioritize the same policy fields as those in which policies have been undertaken. Hence, the simultaneous occurrence of alternative characteristics leads to completely different archetypes of policy implementation settings, and therefore to different conditions under which Cohesion Policy is implemented. The paper therefore considers possible combinations of subjective and objective elements capturing the efficiency of local institutions, their openness to EU values as well as the real/perceived needs of local areas. These combinations define different local scenarios, whose characteristics are assumed to affect the perception of EU policy by local inhabitants and, in turn, their support to the EU integration project. Hence, the second goal of this paper is to define different archetypes of policy implementation settings, and to provide an empirical measurement for EU NUTS-2 regions.¹

To this end, the paper presents a conceptual way to interpret alternative policy implementation settings through a clear definition of the various factors that are supposed to influence such settings. Empirically speaking, an exercise like this requires more appropriate measurements of elements constituting implementation settings. The identification and measurement of the ‘need of a region’ is an example in this regard; an original methodology is proposed. The work presents the application of the methodology to the EU NUTS-2 regions, and interesting results emerge.

DIMENSIONS OF COHESION POLICY IMPLEMENTATION SETTINGS

Policy implementation settings and Cohesion Policy impact

The aim of this study is to highlight conceptually alternative implementation settings, and to identify them from an empirical point of view. The reason for this undertaking is that when the role of EU Cohesion Policy in the EU identity-building process has to be assessed, the policy-implementation setting in which policies are developed has to be taken into consideration. The impact of the Cohesion Policy on the process of EU identity-building is in fact not neutral to structural characteristics of the local implementation settings. This idea is based on two streams of literature.

The first deals with Cohesion Policy and its role in the process of EU identity-building. Many studies, mainly in political science, have focused on the impact of EU actions on the creation of a shared identity (Risse, 2004; Risse & Grabowsky, 2008). According to these works, the implementation of policies is the means by which the EU becomes part of the everyday lives of citizens, being perceived as a tangible and useful construct. Extensive evidence is provided, for instance, by the Erasmus Programme. Empirical results generally show that students enrolled in the project are more likely than the others to identify with EU values (Mitchell, 2015). The mechanism of EU identity-building is interpreted, in this case, as a mainly utilitarian one: the programme provided students with educational, social and professional opportunities that they would otherwise not have had.

A similar process is at work in the case of regional EU policies. The primary goal of the Cohesion Policy is to improve the quality of life of EU citizens through actions in the economic and social sphere. In turn, the increase in well-being generated by these policies is expected to induce a more favourable opinion of the EU among the beneficiaries of these policies (Faludi, 2008). In other words, the positive effect generated by the EU action on individuals’ well-being is the element linking the implementation of EU policy to the creation of a common identity.

A second stream of literature behind the main idea of the paper is the one that suggests that local context conditions strongly influence the impact of EU regional policy on local development in areas where the policy actions are undertaken. Many studies on regional development have recognized the importance of policies tailored to the specific needs of different regions, to exploit the growth potential of each place (e.g., Barca, McCann, & Rodríguez-Pose, 2012; Camagni & Capello, 2015; Mairate, 2006). In other words, in order to be effective, EU actions have to match the objective needs and unexploited potentials of regions. During the 1990s, among all possible structural characteristics influencing the growth potential of each place, a particular role for institutions was highlighted (Gertler, 1997; North, 1990), and since then particular attention has been paid to the quality of institutions with regard to regional development growth (Rodríguez-Pose, 2013). Formal institutions include constitutions, laws, charters, bylaws and regulations, as well as elements such as the rule of law and property rights, and contract and competition-monitoring systems (North, 1990). Informal institutions also play an important role: they are defined in the literature as the individual habits, group routines, and social norms and values that influence the capacity of local actors to cooperate for mutual benefits (Amin, 1999; Capello, 2016).

Institutions have been highlighted as able to increase the efficiency of the returns to certain forms of intervention in regional development (Englebert, 2002). It is not by chance that the ‘Agenda for a Reformed Cohesion Policy’ has underlined the importance of conditionalities on the institutional framework as a means to reinforce the contract when dealing with core priorities, since experience shows that the weakness of the institutional framework is often responsible for the failure of interventions (Barca, 2009).

Empirical analyses exist on the role of institutions on regional growth. Crescenzi, Di Cataldo, and Rodríguez-Pose (2016) have analysed the effect of EU infrastructure investments on regional GDP growth, finding a positive impact only in the presence of high-quality institutions. The same result has been obtained by several other studies (Ederveen et al., 2006; Rodríguez-Pose & Garcilazo, 2015).

The linkage between the quality of institutions and EU identity-building is based on the assumption that better institutional quality guarantees a better effectiveness of policy actions, and therefore a greater collective benefit, leading to a more favourable perception of such policy actions.

Cohesion Policy and EU identity-building: dimensions of the implementation settings

A policy implementation setting is defined as a combination of economic, social and institutional elements constituting the local context in which a policy is implemented. In this study, these elements relate to Cohesion Policy, and can be summarized in two dimensions: the sensitivity and the receptivity of a region to a certain policy action. Sensitivity is defined as the degree to which each region is sensitive to a certain policy, since the latter is both necessary and desirable for the solution of local problems. Receptivity refers to the capacity and closeness that local policy-makers are expected to have in dealing with such policies. The former is captured through the perceived institutional quality that local institutions possess in local citizens’ perceptions, the latter in terms of actual closeness to EU values.

Sensitivity is defined by two elements, one objective and the other subjective: the real and perceived needs of a region (Table 1). The real needs of a region capture the objective need of that area for a particular kind of programme, and identify the degree to which different kinds of policies are needed by the region. In the case of Cohesion Policy, the European Commission itself, in the communication of the Strategic Guidelines for the programming period 2007–13, clearly stated that ‘…  Member States and regions should pay particular attention to (regional) specific needs in order to prevent uneven regional development from hampering growth potential’ (Commission of the European Communities (CEC), 2005). Regional needs are therefore close to the concepts discussed in the literature on the place-based approach to Cohesion Policy.

Table 1. Dimensions of regional policy implementation settings.

Dimensions	Elements
	Subjective	Objective
Sensitivity: need of a region for a certain policy	Need for a certain kind of policies as perceived by the population	Objective need for a certain kind of policy
Receptivity: quality of local institutions and the European Union political values	Perceived quality of the local governments (low corruption, rule of law etc.)	Acceptance of the local institutions of the values, vision and strategies of the European Union

As far as perceived needs are concerned, they measure the subjective priorities of the population with regard to different policy themes. In an ideal world, real and perceived needs should coincide, but mismatches may arise. In fact, the perceived impact of policies is expected to depend on the objective outcomes of these actions, and one can assume that these outcomes will be higher whenever they meet the real needs of each region. Nevertheless, one must recognize that individuals may have misperceptions of these needs. In other words, citizens may give high priority to policy themes that are not particularly urgent for their region due, for instance, to lobbying, imperfect information or their cultural background (Van Oorschot, 2006). In the case of a mismatch between objective and subjective needs, the outcome of Cohesion Policy will be probably valued in a different way. With this reasoning, if the match between local needs and Cohesion Policy actions does not occur, we assume this inconsistency to have a negative effect on the perception of the EU policy (Capello, 2017).

Receptivity measures (1) the capacity perceived by citizens of local institutions to manage policy efficiently and effectively; and (2) the interests (political will) of the local institutions to implement Cohesion Policy actions, highly dependent on their closeness to EU values.

A growing body of literature treats the perceived quality of local governments, of a subjective nature, as a key factor in the success of policies. According to this evidence, one of the axes on which the receptivity of regions is based is citizens’ perceptions of the quality of the local institutions.

The objective element measures the openness of local institutions to EU values, depending on their closeness to the EU's values, visions and strategies. Many studies have discussed the relationship between the composition of governments and the allocation of Cohesion Policy funds. Kemmerling and Bodenstein (2006) claimed that EU funds tend to be more generously allocated to regions with Eurosceptic local governments so as to increase, through funding, the population's political support for the EU. Nevertheless, this result is rather unstable when different statistical techniques are applied (Bouvet & Dall’Erba, 2010; Dellmuth, 2011). In the context of the present analysis, our assumption is that Eurosceptic local governments have, like all regional authorities (Chalmers, 2013), an incentive to maximize the amount of Cohesion Policy funds they receive so they can implement projects in their regions. At the same time, however, they do not have any incentive to promote positive policy outcomes as a merit of the EU but, rather, may claim that they are their own responsibility. Therefore, we assume that people living in regions marked by low levels of EU acceptance are less likely than the others, keeping other characteristics of the local policy implementation settings constant, to perceive the positive impact of Cohesion Policy on their lives and, as a consequence, to increase their identification with the EU.

FROM SINGLE DIMENSIONS TO REGIONAL POLICY IMPLEMENTATION SETTINGS

Taken together, the two dimensions defined above in the previous section concur in the definition of types of policy implementation settings based on possible combinations of subjective and objective measures of both dimensions.

Table 2 identifies nine combinations of sensitivity and receptivity elements; both sensitivity and receptivity are evaluated in relative terms compared with the EU average (the dashed line in the radar charts of Table 2).

Table 2. Archetypes of regional policy implementation settings.

	RECEPTIVITY
SENSITIVITY	Appropriate policy in an ideal context	Appropriate policy in a Eurosceptic context	Appropriate policy in an inefficient context
	Opportunistic policy in an ideal context	Opportunistic policy in a Eurosceptic context	Opportunistic policy in an inefficient context
	Unrequested policy in an ideal context	Unrequested policy in a Eurosceptic context	Unrequested policy in an inefficient context

Sensitivity varies from the top to the bottom of Table 2 according to the consistency between the (objective and perceived) needs and the policies implemented:

• Appropriate policy, when the regions show a clear objective need in a certain policy field, irrespective of the perceived needs of the resident population.

• Opportunistic policy, when a perceived need arises in policy fields where the region does not have an objective need.

• Unrequested policy, when both real and perceived needs are not associated with a certain policy field.

Receptivity, on the other hand, delineates the institutional context in which policies are implemented, namely:

• An ideal institutional context, marked by the simultaneous occurrence of high perceived institutional quality and generalized support for the EU.

• A Eurosceptic context, when the good quality of institutions is not matched by widespread support for EU institutions.

• An inefficient institutional context, marked by not particularly efficient local governments.

In Table 2, receptivity changes from left to right from an ideal, to a Eurosceptic, to an inefficient institutional context.

All the possible combinations of policy scenarios (based on sensitivity) and institutional contexts (based on receptivity) define nine archetypes of policy-implementation settings. The purpose of the next section is to translate this conceptual framework into empirical terms, providing a measurement for each of the elements characterizing the policy implementation settings.

MEASUREMENT OF THE REAL NEEDS OF REGIONS

Real needs: a conceptual identification

The real needs of a region capture the objective needs of that region for a particular kind of policy. Many studies have discussed the necessity of place-based theories in order to promote economic development by exploiting the potential for growth of every territory (Barca, 2009; McCann & Rodríguez-Pose, 2011, pp. 203–213).

On moving from a theoretical to an empirical framework, however, the definition of needs is extremely complex and requires careful thought. An objective need may occur, in fact, under very different and alternative circumstances. It may derive either from a low endowment of a certain asset or from its inefficient exploitation, or from lack of a critical mass, or from the occurrence of decreasing returns in its use. None of the above definitions of need is meaningful in properly capturing the real requirements of an area. For instance, the low endowment of a resource may correspond to a low demand for that asset, so that it would be inappropriate to conceive it as a need. Moreover, the analysis of the supply is, per se, not enough to define the real needs of a region, since it does not take the corresponding demand into account.

In the context of this paper, our assumption is that regions are characterized by an objective need for a good/service/resource, and therefore call for a policy intervention in a certain field² when the supply of the good/service/resource considered is inadequate to meet the demand. Hence, what matters is not the absolute level of supply of resources but, rather, the imbalances between supply and demand.

In principle, the occurrence of an asymmetry between supply and demand is an exceptional event: in a competitive equilibrium, we would expect the supply to match the demand perfectly. Cohesion Policy intervention, however, is expressly devoted to those cases where markets fail to lead to an efficient equilibrium and, as a consequence, public intervention is needed on either efficiency or equity grounds. In more detail, Cohesion Policy actions in alternative policy fields can be justified by three main reasons, leading to different interpretations of the concept of real needs as an imbalance between supply and demand (Table 3):

• In the presence of asymmetrical information: whenever the market is characterized by asymmetries in the information between economic agents, good trades are missed (Akerlof, 1970). Situations of this kind arise, for instance, when firms cannot find financial support in the private market, since credit institutions are unable to evaluate the risk of the potential investment accurately: the supply of financial resources is lower than the demand. Another example is when information and communication technology (ICT) or transport infrastructure is not sufficiently supplied by the private sector because of the high costs of the provision (e.g., construction cost of the network) and uncertainty about future revenues. Healthcare also falls within this category: given the demand for healthcare by citizens, private companies cannot correctly discriminate their customers based on their health conditions and, as a consequence, some people will not be insured. In all these cases, public intervention is needed on efficiency grounds. For these fields of policy intervention, regional objective needs arise under these conditions when the supply of the resource is lower than its demand (S < D).

• In the presence of negative externalities: in the presence of externalities, the welfare of individuals is indirectly (i.e., without the mediation of prices) affected by the actions undertaken by other economic agents. Externalities act on the quantity of the goods supplied and, in the case of negative externalities, overproduction occurs. A competitive equilibrium is, therefore, not efficient and public intervention is required. Pollution is a typical example of this situation, and environmental actions undertaken under Cohesion Policy are aimed at mitigating the impact of negative externalities on social welfare. Hence, in this case a region is in need of a public policy if the supply of the resource producing a negative externality is higher than its demand (S > D).

• In search of equity: in all the previous situations, the public intervention in the economy was based on efficiency grounds, because market forces alone were not able to reach an efficient equilibrium. However, improving the efficiency of the EU regional economies is certainly not the only purpose of Cohesion Policy. One of its main goals is to reduce inequalities and social disparities within and across regions through, for instance, policies focused on social exclusion and unemployment. Hence, in such cases public actions are justified by equity considerations. Regional needs arise when the supply of social policies is lower than the demand (S < D).

Table 3. Supply–demand imbalances by policy field and the empirical measurement of their determinants.

Policy field	Imbalance supply (S) and demand (D)	Supply	Supply indicator	Demand	Demand characteristics	Indicators of demand characteristics
Economy	S < D	Firms’ investments	Per capita private investments	Credits' demand of firms	Productive specialization	Employment in manufacturing and the service sector
					Demographic structure	Share of the active population
					Functional specialization	Population with a tertiary education
					Degree of urbanization	Population density
R&D	S < D	Firms' investments in R&D activities	Per capita R&D expenditure	Firms' demand for R&D investments	Productive specialization	Employment in knowledge-intensive sectors
					Propensity to innovate	Per capita patents
					Functional specialization	Employment with a tertiary education in science and technology, innovation behaviour
					Degree of urbanization	Population density
Tourism	S < D	Tourism facilities	Beds in accommodation facilities	Number of potential tourists	Cultural and natural heritage	Per capita monuments, Touring Club Italiano (TCI) stars**, per cent of natural areas, heating degree-days
					Degree of urbanization	Population density
					Overall economic conditions	Per capita GDP
ICT	S < D	ICT networks and services	Share of the population with broadband access	Demand for ICT services	Presence of firms	Per capita private investments in ICT
					Human capital	Population with tertiary education
					Degree of urbanization	Population density
					Demographic structure	Dependency ratio
Health	S < D	Public health	Life expectancy at birth	Demand for health (the population’s health conditions)	Health facilities	Per capital hospital beds, NHS organization, infant mortality rate
					Degree of urbanization	Population density
					Overall economic conditions	Per capita GDP
Energy and environment	S > D	Production of energy with low environmental costs	Pollution level	Demand for environmental quality	Productive specialization	Employment in agriculture
					Cultural and natural heritage	Share of natural areas, heating degree-days
					Degree of urbanization	Population density
					Congestion	Motorization rate, traffic congestion
Society	S < D	Development under equity	Per capita value added in non-market services	Demand for equity	Presence of firms	Per capita private investments
					Degree of urbanization	Population density
					Socioeconomic conditions	Net disposable income, unemployment rate

Note: GDP, gross domestic product; ICT, information and communication technology; NHS, National Health Service; R&D, research and development.

Real needs: an empirical measurement

The empirical measurement of real regional needs requires the conceptualization and measurement of both the supply of and the demand for the resources characterizing each area of Cohesion Policy intervention.

As far as the conceptualization is concerned, the supply is quite easily identifiable. In the economic sphere, for instance, the stock of private capital invested in the production system captures the supply of investments of firms. Similarly, the available tourism facilities represent the supply of the relevant resource in the tourism area. In some cases, the resource provided is not tangible but, rather, intangible, as in the case of healthcare, where the supply is not understood as the provision of a tangible good (hospitals), but as the achievement of a certain standard in the lives of the population. Given the conceptual definition of supply in each policy field, identification of an appropriate empirical indicator is straightforward. Table 3 reports this information.

The demand side, on the other hand, raises some issues. From a conceptual point of view, in fact, we are not interested in measuring the actual demand for the resources characterizing each policy setting but, rather, their potential demand, which contains both the demand fulfilled by the current supply and the portion of demand that is (or could be) unmet by the present supply conditions. This means that we should compare the current level of supply identified above with the demand that would occur in a competitive market, i.e., in a market with no imperfections. In the economic field, for instance, the supply of private capital in the productive environment has to be compared with the potential demand for credit by firms. In this case, an imbalance between demand and supply would imply that the supply of financial support to firms is lower than its efficient level, and therefore that public intervention in this policy field is necessary.

Unfortunately, from an empirical point of view, it is extremely difficult to measure the potential demand for any resource: for instance, no data are available on the willingness of firms to receive financial support. Potential demand, however, can be indirectly detected through the presence of some economic characteristics of the local productive environment. As an example, the demand for private investments is related to the productive specialization of the region (higher specialization in industries is expected to increase the demand for credits), to the structure of the job market (lower dependency ratio in the labour market is expected to increase the demand for credits) and to the presence of urban areas (reflecting the presence of high-value, capital intensive functions typical of cities).

More in general, we assume the following relationship:(1) where the quantity of the resource a demanded on the market of region r is a function of a set of relevant characteristics of that area. As for the example of the productive environment above, a set of demand characteristics was identified for each policy axis. Besides the supply indicators, Table 3 provides information on these variables and on their empirical measurement.³

Model (1) makes the supply of a certain resource directly dependent on some characteristics of the demand. Any imbalance between the two dimensions is captured by the error term ε. In purely econometric terms, the error term represents factors other than those included in the regression that affect the dependent variable. When all the relevant factors are included in the regression model, the error term can be interpreted as a measure of a market imbalance between supply and demand. In this study, we interpret ε as a pure market imbalance, therefore assuming that there are no omitted variables in the model specification. We are in fact quite confident that the main explanatory variables are inserted in the model, since we were able to consider all variables mentioned by the broad literature devoted to the conceptual and empirical analysis of the relationship between supply and demand of the resources considered.⁴

A model taking the form of (1) was therefore run for each policy field.⁵ The results showed that the characteristics of the demand are significant predictors of the supply, even if they do not entirely explain it.⁶ Hence, as discussed above, the error term captures the imbalance between supply and demand, and it is therefore used as a proxy for the real regional needs. A negative value of the error term implies that the supply is less than the expected demand (S < D), pointing to the occurrence of a real regional need in all policy fields apart from the environment and energy area (Table 3). A positive value of ε, on the other hand, indicates the oversupply of the resource under consideration (S > D), corresponding to a regional need in the environment and energy field.

Based on this reasoning, the predicted error terms of the model [1] run across the different policy fields summarized in Table 3 provide information on the real needs characterizing EU regions in the different areas of intervention of Cohesion Policy.

MEASUREMENT OF THE PERCEIVED NEEDS OF REGIONS

In principle, if individuals were characterized by perfect information and unbounded rationality, the perceived need for a certain policy would reflect the real one. Nevertheless, these conditions rarely hold. A long stream of research has pointed out how, even when facing apparently easy choices and problems, individuals fail to behave consistently with their preferences (Kahneman, 2003). Moreover, they usually have poor information about the costs and benefits generated by alternative public policies. As a consequence, lobbies and groups of interest may exploit this lack of information by orientating the public opinion in a way that they find beneficial (Stiglitz, 1998). Based on this reasoning, we labelled the local policy implementation setting characterized by unbalances between real and perceived needs an ‘opportunistic context’.

From an empirical point of view, perceived regional needs can be measured using Eurobarometer (EB) data. Since 1973, the EB has conducted survey analyses on behalf of the European Commission in order to monitor the evolution of public opinion across member states. Millions of EU citizens have been asked about a broad variety of issues, and many of these questions have been repeated over time. One of these recurrent topics concerns the policy fields that should be the object of EU intervention. The question is as follows: ‘European integration has been focusing on various issues in the last years. In your opinion, which aspects should be emphasized by the European institutions in the coming years, to strengthen the European Union in the future?’ Each respondent had to indicate his/her preference by mentioning no more than three items among a list of options.⁷

Since, among other characteristics, EB respondents are asked to declare their region (NUTS-2) of residence, we were able to calculate, for each of the policy fields considered, the share of people that, in a certain region, mentioned it as a primary axis of intervention for EU actions.

To be noted is that the sample of respondents in EB studies is representative of the national population, but not of the regional communities. In order to alleviate this issue, we pooled several EB surveys, conducted between 2007 and 2009, including the same question about the fields of intervention of EU institutions. As a result, we ended up with a data set of 181,380 individual observations, with an average of 788 respondents for each NUTS-2 region.⁸

MEASUREMENT OF EUROPEAN UNION ACCEPTANCE AND INSTITUTIONAL QUALITY

The receptivity of an area to a certain policy is represented by the subjective and objective elements explaining the efficiency of an institutional context in handling EU policies. Receptivity is not assumed to change across different kinds of policies: whatever the quality level of local institutions and/or their openness to EU values, these characteristics are invariant across different types of actions.

Receptivity consists of two elements. The first is the degree of Euroscepticism characterizing a local context. Some scholars argue that support for Eurosceptic parties is mainly an instrument for voters to protest against the national governments (Prosser, 2016): the broad political coalitions governing several countries in Europe have led to the rise of populist parties able to attract the preferences of dissatisfied citizens, as in the case of the UK Independence Party (UKIP) in the UK (Clarke, Whiteley, Borges, Sanders, & Stewart, 2016). On the other hand, some researchers (Ferrara & Weishaupt, 2004) suggest that the EU parliamentary elections have lost their ‘second-order’ status: because EU policies and rules have gained momentum and influence on EU citizens’ lives, voting for EU institutions is now an opportunity to confront alternative views on the EU itself. Therefore, the support for Eurosceptic parties is not regarded as a form of disapproval of national governments but, rather, as dissent with the programmes and strategies undertaken by the EU. The conditions characterizing a ‘Eurosceptic context’ are expected to weaken the residents’ perception of the impacts of Cohesion Policy (e.g., Lubbers & Scheepers, 2010; Prosser, 2016; Ferrara & Weishaupt, 2004; Van de Wardt, 2015; Treib, 2014). Empirically, we measured the level of EU acceptance with the share of votes for non-Eurosceptic parties in the European Parliament elections held in 2009.⁹

The second element of receptivity is the quality of local institutions, as perceived by citizens. The role of institutional quality in economic growth is well recognized in the literature (Easterly, Ritzen, & Woolcock, 2006). More recently, its impact on the EU regional policy has been addressed by several studies (Ederveen et al., 2006), generally pointing out an increasing return on public investments in the presence of efficient institutions. The limitation of these studies is related to the unavailability of data at the regional level.

The research team at the Quality of Government Institute of the University of Gothenburg has prepared, on behalf of the European Commission, a detailed study on the subnational variations of the quality of government in EU countries (Charron, Lapuente, & Rothstein, 2010). The aim of this study is to measure perceived institutional quality at the regional (mainly NUTS-2) level. In more detail, the quality of government is defined along four dimensions: corruption, rule of law, bureaucratic efficiency, and the strength of democratic and electoral institutions. In a survey study conducted in 2009, 34,000 individuals in the EU were asked to answer some questions about their perceptions of these four dimensions. The results of this survey made it possible to calculate an overall index of perceived institutional quality.¹⁰

Several studies have adopted this indicator in order to study the relationship between the quality of regional institutions and the provision of transport infrastructure (Crescenzi et al., 2016), the return on Cohesion expenditure (Rodríguez-Pose & Garcilazo, 2015) and the occurrence of regional disparities (Ezcurra & Rodríguez-Pose, 2014). Following this literature, the perceived institutional quality of regional governments is also captured in the present paper by the overall index developed by Charron, Dijkstra, and Lapuente (2014).

LOCAL POLICY IMPLEMENTATION SETTINGS OF EUROPEAN UNION REGIONS

By combining the four different indicators built in previous sections, the regional implementation settings of Table 2 were identified for all NUTS-2 regions of Europe. Regions were allocated to the alternative policy settings according to their relative value (with respect to the EU average) on the four indicators of real and perceived needs, institutional quality and EU acceptance.¹¹ Since objective and subjective needs assumed different values for different policy axes in the same region, suggesting a regional policy setting for each policy axis, an additional aggregation was necessary, one based on similarities in policy interventions, namely interventions in the productive environment (economy, research and development (R&D) and tourism), in infrastructures (health, transport, energy and environment and ICT) and in the social field (society). The aggregation of different policy fields was obtained by choosing the dominant policy setting emerging within the sub-themes.¹² In the case of non-existence of a dominant situation, the opportunistic case, conveying the most relevant implications for policy-making given the risk of rent-seeking positions, was chosen as the one prevailing over the others.

Figure 1(a) presents the policy-implementation settings as regards the actions on the productive environment. The first finding that emerges from Figure 1(a) is the strong divide between Southern and Eastern Europe and the rest of the EU. Southern and Eastern regions belong, generally, to appropriate policy settings in an inefficient institutional context. This situation is rather homogeneous, even if with some exceptions in Italy, where in a poor area like the south of Italy, policy implementation settings proved not to be appropriate. This apparently contradictory result can be explained by the fact that in those regions pure endowment policies in the form of higher funds spent in this field do not seem the most appropriate means to overcome the inefficiency gap that these regions face in the productive environment (Aiello & Pupo, 2012).

Figure 1. (a) Policy implementation settings for the interventions in the productive environment; (b) policy implementation settings for the interventions in infrastructures; and (c) policy implementation settings for the interventions in the social field.

Interventions in the productive environment are unrequested in most of Scandinavian, Austrian, and some French and British regions, even if the institutional context is ideal, i.e., both efficient and supportive of EU policies. To be noted is that the majority of areas in which Cohesion Policy is unrequested are also those characterized by efficient institutions. Rather rare, in fact, is the case in which unrequested policies are matched with the poor quality of local governments.

This evidence conveys important policy implications. As discussed above, previous literature (Rodríguez-Pose, 2013) has pointed out that the effectiveness of Cohesion Policy is positively associated with the quality of local institutions. This implies, however, that allocating funds to the regions with the highest real needs may not reinforce EU identity. On the one hand, in fact, citizens in recipient regions may not perceive the positive impact of EU actions on their well-being, due to the difficulties of local institutions in effectively implementing Cohesion Policy. On the other hand, citizens from the net contributors to the EU budget may raise concerns about the allocation of the resources and the low effectiveness of EU actions.

A similar reasoning can be applied to several regions in the EU-12, classified among those areas where Cohesion Policy would be appropriate but the degree of Euroscepticism is above the average (the UK, the Netherlands, Denmark and East Germany). Since most of these regions are included among the transition or more developed regions in the programming period 2014–20, the relatively low (compared with less developed areas) amount of funds received might lead to a further weakening of EU acceptance.

Figures 1(b, c) show the regional policy implementation settings when policy actions are devoted to infrastructure and social policies. Figure 1(b) presents two main messages. First, appropriate policy settings in an inefficient context characterize most Eastern country regions, and some Spanish, Italian and Greek regions. Secondly, the occurrence of opportunistic policy settings is higher than in the previous case, especially in Central Europe. This is potentially alarming, since these kinds of actions often concern the implementation and provision of large-scale infrastructures, and the onset of rent-seeking behaviours is therefore likely.

Finally, Figure 1(c) shows the regional policy implementation settings in the case of social policies. Again, Eastern and Southern country regions (with the exception of the south of Italy) are those with the highest levels of appropriateness, and the lowest efficiency in implementing policies.

One result is particularly interesting: opportunistic settings occur with high frequency in the contexts characterized by strong Eurosceptic parties. The rent-seeking mechanism suggested above seems to be operated, in the case of social policies, by political movements rather than private economic agents. Immigration and the issues related to this phenomenon, for instance, are among the most popular themes emphasized by Eurosceptic parties in order to exploit niches in the local political arena. Therefore, it is not surprising that, in areas with low levels of EU acceptance, the perception of social issues is higher than what would be justified by the objective socioeconomic conditions.

CONCLUSIONS

The goal of this paper has been to build a conceptual and empirical taxonomy of EU regions based on the local characteristics that are expected to influence citizens’ perception of Cohesion Policy.

Compared with previous works, suggesting the importance of certain territorial conditions in fostering the effectiveness of Cohesion Policy, the present study has two innovative aspects. First, it provides a conceptual framework for understanding the mechanisms through which specific context conditions are expected to mediate the perception of EU policies. The framework is built so as to highlight the mix of different territorial conditions that favour/hamper the success and image of a policy action, which the previous literature treated separately. Instead, reading the conditions in a comprehensive framework, there emerge different interesting policy settings in which policy actions can be developed. Thanks to this framework, the second innovative aspect is achieved: that of providing policy-makers with the structural local conditions across EU regions in which their normative interventions take place. These differences are expected to lead to a highly diversified perception and success of similar policies and actions.

Further analysis is certainly needed to understand fully the mechanisms of EU perception formation and its associations with the regional types defined in this study, but the preliminary findings presented here already raise several implications and suggestions for future research.

The first involves the association between need, understood as a gap between the demand for and supply of resources, and institutional inefficiency. Regions receiving most of the funds (Eastern and Southern EU regions) objectively need them but, at the same time, present an inefficient implementation context. This may have negative implications for the perception of Cohesion Policy in both recipient areas, since residents do not feel any positive impact on their lives, and net-contributing regions, because taxpayers are aware that their money is being spent inefficiently.

The second consideration concerns the occurrence of opportunistic policy settings. While these situations are rather limited in the case of interventions in the productive environment, they become much more frequent in that of infrastructural and social policies. In the former case of infrastructure, this is particularly critical due to the amount of funds generally allocated to these actions. In the latter case, it may reflect the instrumental use of social issues by Eurosceptic parties to gain the support of the local voters.

Finally, an additional consideration concerns the association between Euroscepticism and other territorial characteristics. The analysis has pointed out that, especially as regards the actions on tangible private goods, some regions marked by low levels of EU acceptance show a high need for policies. Nevertheless, the funding for these areas is, in relative terms, below the average.

SUPPLEMENTAL DATA

Supplemental data for this article can be accessed at 10.1080/00343404.2017.1397273

DISCLOSURE STATEMENT

No potential conflict of interest was reported by the authors.

ORCID

Roberta Capello http://orcid.org/0000-0003-0438-6900

Giovanni Perucca http://orcid.org/0000-0001-5582-1912

Additional information

Funding

This work received financial support from the European Union's Horizon 2020 Framework Research and Innovation Programme, COHESIFY project (February 2016–April 2018) [grant number 693427].

Notes

1. NUTS = Nomenclature des Unités Territoriales Statistiques.

2. With the expression ‘policy field’, we refer to general areas of intervention of public policies. Based on the classification suggested by CEC (2015), the areas of intervention of Cohesion Policy are the following: economy, R&D, tourism, transport, ICT, health infrastructure, energy and environment, and social welfare. Each field is characterized by a specific good/service/resource supplied to the market, such as private investments in the economic field, or tourism facilities in the tourism area of intervention.

3. For a more detailed discussion on the demand characteristics and the indicators chosen to measure, see Appendix A in the supplemental data online.

4. For a detailed discussion of this literature, see Appendix A in the supplemental data online.

5. With the exception of the field of transport infrastructure, for which the imbalance between supply and potential demand is captured by an indicator of road congestion, defined as vehicle-kilometres by road over the total length of lanes (length of road × number of lanes). This indicator already expresses the asymmetry between the supply (surface of roads) and the demand (number of commuters) for transport services.

6. The results for all the policy fields are available in Appendix A in the supplemental data online, jointly with a thorough discussion of the choice of the variables employed. All the data used for the empirical identification of the policy implementation settings refer to 2009. The number of regions covered in the analysis changes across policy fields, due to data availability, from a minimum of 236 to a maximum of 267 (see Table A2 in Appendix A in the supplemental data online).

7. The options are reported in Appendix B in the supplemental data online.

8. For more details, see Appendix B in the supplemental data online.

9. The classification of parties between Eurosceptic and non-Eurosceptic is based on internet-based research and the analysis of Treib (2014). Both soft and hard Eurosceptic movements are classified together as anti-EU. The source of the data is the European Election Database managed by the Norwegian Centre for Research Data.

10. Descriptive statistics on this indicator are reported in Appendix C in the supplemental data online.

11. For instance, a region is characterized by a high real need if the relevant indicator (i.e., the predicted error term reported in Table A2 in Appendix A in the supplemental data online) is higher than the EU average. The same applies to the other three dimensions of the local implementation settings. A high perceived need in a certain policy field corresponds to a higher-than-average value of the indicator previously defined. Finally, regions are characterized by high EU acceptance and perceived institutional quality if the value of the relevant indicators is higher than the EU average.

12. In particular, economy, R&D and tourism were merged under the heading ‘productive environment’, while health, transport, energy and environment, and ICT were grouped under the heading ‘infrastructures’.