Stakeholder Empowerment in Sustainable Rural Development Partnerships: Two Case Studies from Italy

Abstract

The funds allocated through the EU Rural Development Programme provided the engine for local stakeholders to interact, forming mixed collaboration partnerships. This paper investigates the structure of such partnerships with the aim of verifying whether (R1) there are significant differences between the various stakeholder categories in influencing the rural development process, and (R2) which categories of stakeholders are more empowered in directing the design of sustainable rural development. The study is focused on two Italian cases: the regions of Apulia and Veneto. Using a combination of SNA and nonparametric tests, the results demonstrate that the empowerment of the stakeholders followed unequal paths in the two cases; i.e., a central role is clearly played by economic associations in Apulia, while this power is more distributed between different kinds of stakeholders in the case of Veneto. Agricultural associations in Apulia play an important role in the densely connected rural development network, promoting information flow and collective action. On the other hand, the weakness of this configuration lies in the fact that the rural development agenda can receive strong pressure from the agricultural sector, pushing more sectoral strategies in turn. Private companies play a key role in Veneto’s rural development, bridging the network gaps between more clustered local groups and increasing pluralism and inclusion. However, the network is sparse and shrinking, posing challenges in terms of coordination and collective action. This kind of evaluation makes policymakers and managers aware of both the most influential and weakest actors. This is crucial to improving the effectiveness and sustainability of the project, as they can involve the most influential groups from the early stages of the design process to ensure support as well as address the needs of the lagging stakeholder categories to reinforce tacit rules, trust, accountability, and responsibilities.

1. Introduction

Bottom-up or endogenous approaches have been the reference model for rural development in the EU since the early 1990s with the LEADER Initiative. The bottom-up approach implies that local actors participate in decision-making about a strategy and in the selection of the priorities to be pursued in their local area. The involvement of local actors includes the population at large, economic and social interest groups, and representatives of public and private institutions in upholding cultural, social, political, and ecological values and the possible assessment of social costs and long-term effects [1].

Five editions of the LEADER Initiative have been implemented over the past 30 years, with improved schemes based on previous experience. They have been consolidated with Community-Led Local Development (CLLD). Although this approach is mostly focused on qualitative and structural interventions rather than just using quantitative and monetary measures as criteria to assess its success [2], EUR 19.81 billion have been allocated through this programme for rural development in the period of 1991–2020 to give local communities the means to develop their potential and their territories [3]. LEADER Initiatives beneficiaries have made investments that raised the living standards of their territories, enriched rural life, and improved physical infrastructure and tourist attractiveness as a result of protecting their rural environment and cultural heritage [4].

The idea underlying the bottom-up approach is that local people are the best experts in the development of their territories. The co-design of a development project can increase stakeholders’ self-awareness, the ability to develop a shared vision, and the capacity to generate novel ideas and produce value-adding initiatives, eventually benefiting the region as a whole [5] In this vein, Li et al. [6] highlight that rural development is strongly shaped by the ability of the community to efficiently respond to external changes. In this regard, for sustainable rural development, it is important to strengthen the internal capacities of the communities that represent their resilience potential. This enables communities to withstand any external influence and even unexpected shocks.

The European Network for Rural Development conceives the LEADER Initiative as a participatory democracy tool, and the allocated funds provide the engine for stakeholder interactions [7]. This is a source of another crucial endogenous resource: community social capital. When the social capital of a group is increased, cooperation is, in turn, positively influenced. Thus, social capital changes the cost–benefit ratio for individuals and mitigates the problem of possible free-riding in policy programs [8]. The practice of rural development partnership not only empowers people and strengthens local capacity but also develops peoples system thinking skills. System thinking skills or big picture thinking mean that people see the complex picture of the issue and comprehend the interdependence and interconnectedness of various components comprising the issue. This skill helps to better navigate complexity and find efficient solutions for a sustainable future [9,10,11].

Moreover, as rural development partnerships are multi-stakeholder in nature, the main goal is to create a win–win scenario for all parties involved, where they combine their resources and skills to successfully address shared social or environmental goals [12,13,14,15].

The definition of partnership differs in the literature depending on its purpose, the actors involved, the stage of development, the spatial dimension, and the implementation mechanisms [16]. Here, we refer to a territory-based concept of partnership, as defined by Biderman et al. [17], which characterises formal long-term collaborations between governmental, corporate, and volunteer organisations who agree to share responsibilities, risks, costs, and benefits in achieving shared goals in a particular geographic location. This process involves a combination of consultation, negotiations, and bargaining, heavily dependent on a shared vision and on the ability to identify appropriate trade-offs both between and within interest groups [18].

With respect to their origins, local partnerships in EU rural developments relate to three approaches. Local partnerships are created (i) by individuals, often “local leaders”, who would like to actively intervene in the local reality, not only economically but also socially and culturally; (ii) by businesses that claim a privileged position in economic decisions; (iii) and by public authorities, which can compensate for a rare or failing private initiative in problem areas. The nature of the process and outcomes are, therefore, different for each of the three cases [19].

Within the context of LEADER, a partnership is considered the formation of a network of relationships and solidarity at the level of a certain rural area, called a Local Action Group (LAG). Often it is expected to be more than just joint economic activity; rather, it may be a local entity representing a will to build a social identity [5].

Following the introduction and implementation of the first edition of LEADER, rural partnerships created over the course of this initiative have been scrutinised from many perspectives by numerous scholars in different European countries, including new member states and post-socialist countries [4,20,21,22,23,24,25,26,27]. Some of them have emphasised the programme’s beneficial influence [4,28,29], while others have criticised it and suggested improvements [8,21,24,30,31] Ray describes the LEADER Programme as contributing to rural development, as it initiated many local-level projects. Hoffman and Hoffman [4] mention the positive impact of programmes that have built strong social and human capital and cooperation skills and improved the level of living and the quality of life in the countryside.

Esparcia [28], although acknowledging the overall positive impact of the programme, note that it has certain institutional and structural issues that prevent the wider community from being involved in and benefiting from it: “LEADER has become a political, social and economic instrument” [28]. Similar arguments are put forward by Navarro et al., who point out that actively involved stakeholders mainly represent those who have the time, resources, and aspirations to take part, notably, local elites. The poor and other disadvantaged groups remain marginalised (Navarro et al., 2015) [24].

Most scholars agreed that the nature and degree of partnership vary significantly depending on the characteristics of each country, such as the political regime, the degree of government centralisation, the strength of civil society and local participation, and the presence of a strong tradition of cooperation.

Mosley et al. [21] provide a comprehensive examination of 24 case studies across several European countries. According to this research, common factors serving to improve or weaken partnerships in rural development have been observed. Among the positive ones, there is the existence of coherent and relevant aims based on the recognition of common needs, strong but not over-dominant leadership, good administration and technical support, good user-friendly communication, the early achievement of visible benefits, mutual trust, the equal participation of parties in the decision-making process, well-developed informal networking, and others. Among the constraints for successful partnerships were the stated centralisation of decision-making, the dominance of a certain small group of stakeholders, limited financial resources, excessive bureaucracy, short time perspectives of development programmes, and others.

Since the predominant sector in rural areas is still agriculture, most studies focus on this sector. In this regard, research carried out by Guerrero-Ocampo [27] presents a special interest. He investigated the social structures of multi-actor partnerships involved in interactive innovation processes in agricultural innovation systems. The results of 17 case studies that he examined based on social network analysis (SNA) and descriptive statistics show that the composition of the innovation networks is diverse, but when the frequency of connections is examined, there is a tendency to develop more interactions between organisations of the same type. Farmers and research institutions are critical members in the “core” of innovation networks.

Despite the abundance of research analysing the LEADER Programme, its structural specifics, and contributing role of LAGs in the growth of cooperation in rural areas, there has been little discussion at a deeper level focusing on the very structure of partnerships concerning how partnerships are formed, how stakeholders are connected, who are the most influential ones, and how decisions are made in their networks.

This paper provides a contribution in this direction, as it investigates, with respect to two territorial case studies, whether there are (R1) significant differences between the various stakeholder categories in influencing rural development partnerships and (R2) which categories of stakeholders are more empowered in directing the process of sustainable rural development.

To study these aspects, we adopted an analytical approach based on SNA to investigate the Italian case. Italy is a country that has been dealing with partnership formation in a rural context since the very first edition of the LEADER Initiative. These cases relate to two major agricultural regions, namely, Veneto and Apulia.

Section 2 illustrates the materials and methods employed. It focuses on case studies, a review of the stakeholder categories, and the analytical approach used. In Section 3, the results of the analysis are presented. Section 4 returns to the research question and provides several policy implications and some concluding remarks.

2. Materials and Methods

2.1. Case Study Regions

Our work is focused on the Veneto and Apulia regions, two Italian territorial administrative units. These two regions were selected based on several factors: distinct geographical locations, climatic differences, ecological specifics, and economic and social contrasts. Veneto is considered one of the most developed regions of Italy, whereas Puglia is still lagging behind. For the purposes of its convergence objective, the EU Rural Development Scheme has distributed greater financial resources to the Apulia region in the implementation of the LEADER Programme. A specific focus on these two territories is provided in what follows.

Apulia Region. This region is located in the southeast of Italy; it is the 7th region in terms of area (19k Km²) and population (+4 million). According to the Italian National Institute of Statistics (ISTAT), Apulia can be classified as a lagging-behind region. Its GDP accounts for EUR 69.5 billion. Its economy has a strong agricultural tradition, and this sector contributes 3.6% to the regional GDP. The key agricultural products are wheat, olive oil, and vegetables. These products are crucial in the Apulian export balance, which has accounted for +11% since 2008. A large number of traditional products are presented, which are included in MIPAAF (Italian Ministry of Agriculture, Food, and Forestry Policies), and Slow Food does represent a valuable resource in promoting the overall area and its local production [32]. Strong growth in the service sector accounts for 24.3% of the total GDP. Tourism contributes 3.6% of the regional GDP with EUR 9 billion. Due to its strategic position in the Mediterranean basin, Apulia is well connected both externally and internally by air and by sea, with 10 airports, 12 commercial and civil major ports in the region, and another 34 touristic ports. The rail network is the least developed infrastructure network in the Puglia region at 20% below the Italian average. The Apulia region is part of the Natura2000 EU Network, with 21% of its total area, 92 sites, 44 habitats, 90 bird species, and 81 species of EU interest. In Apulia, there are 4 UNESCO-designated sites. The post-graduate employment rate is 37.1% below both the South Italy benchmark (42%) and Italy overall (62.8%). This metric shows how strong the brain drain phenomenon is, which could really compromise the region’s future competitiveness. The 23 LAGs present in Apulia have oriented their Local Development Strategies (LDS) toward the innovation of local food, crafts, or manufacturing production systems; the development of energy supply chains; social promotion; and urban requalification, including the enhancement of cultural heritage and sustainable tourism [33].

Veneto Region. This region is located in the northeastern part of Italy. It corresponds to the most developed region category referred to in the EU Rural Area Classification of the 2014 Italian Rural Development Programme. It covers an area of 18,399 sq. km; is divided into 579 municipalities; and is 56% flat, 29% mountainous, and 15% hilly. According to the classification made by CORINE-Land-Cover 2006, there is a prevalence of agricultural lands (57.2%) and a significant portion of forested lands and/or semi-natural environments (29%), while 4% of the territory is claimed by water bodies, and 1.5% is affected by wetlands. The territory’s urban and industrial infrastructure, on the other hand, accounts for 8.2% of the regional territory [32]. The main economic activities are tourism and agriculture. Undoubtedly, the wine sector stands out in the region, and it is currently the largest producer and exporter of quality Italian wine, with almost all (91%) of protected designations of origin or geographical indications. LAGs operate in about 70 per cent of the Veneto region, within rural areas, and/or with an agricultural vocation, affecting about 35 per cent of the population. The main objectives of the LEADER Programme implementation in this region were defined as supporting participatory approaches; improving the capacities of local partnerships; promoting cooperation between territories; and stimulating the harmonious, endogenous development of rural areas [34].

2.2. Partnership Categories

For the purpose of identifying and categorising stakeholders in rural development partnerships, we employed a model presented by the LEADER Laboratory in 1997 [19]. This model introduces three levels of classification. At the first, more general level, the stakeholders are grouped into three macro-categories based on their backgrounds and missions: (1) public institutions, (2) private companies, and (3) civic society (i.e., people or associations of people). The model further classifies these groups into a second, deeper level based on the nature of their operations, for example, the agro-food sector, the finance sector, and others. These sections have different subgroups (third level of classification) formed based on stakeholder interests. Working with this model, we have found some disparities, in particular, in the second and third levels, where some categorisations are related to the geographical specifics of stakeholders and some are based on the characteristics of stakeholder operations. To fix these issues, we modified the second and third levels, making classifications more accurate, clearer, and better suited to the rural development context. Since the present research is focused on sustainable rural development, the categorisation of partnership dimensions developed by [35] was used as a reference to elaborate our second-level classification for the purpose of coding stakeholders in our analysis. Our final classification method is reported in

Table 1. Stakeholder categorisation method.

First Level	Second Level	Third Level
	Local Authorities	Town Halls
		Municipal Associations
		Regional or Territorial Government
Public Institutions	Public Services	Economic Services
		Social Services
		Research and Educational Services
		Infrastructural Services
	Government Agencies	Economic Agencies
		Environmental Agencies
		Other Agencies
	Agro-Food Sector	Processing
		Farmers
		Cooperatives or Producer Organisations
		Union of Cooperatives
Private Companies	Financial Sector	Bank
		Agricultural Mutual Savings Bank
	Other Producing Sectors	Business and Industrial
		Import–Export
		Other Services
	Economic Associations	Professional Association
		Agricultural Association
		Labour/consumer
		Development Agencies
Civil Society	Social Associations	Cultural Association
		Leisure, Sport, or Recreational Association
		Social or Religious Association
	Environmental Associations	Protection
		Study
		Use

.

We used the classifications presented in Table 1 as a baseline to categorise the stakeholders in our case study analysis.

2.3. SNA and Relational Data

We used SNA to analyse the ability of the various stakeholder categories to exert some kind of influence and power over rural development activities. The idea is that an important part of stakeholders’ power stems from their peculiar position and connectedness in the partnership network. This is captured by some specific network indexes, as explained below in this section.

To represent and analyse the collaboration interactions between the stakeholders from the regions we selected as case studies, we used a technique called the affiliation network approach. In general, this approach allows us to pass from a two-mode network, representing the relationships between two sets of entities, called (i) agents and (ii) events, to a one-mode network that depicts the relationships between only one set of agents. A link in the two-mode network is established between the events and the agents attending those events. Finally, a link in the one-mode network is established between the agents attending the same event.

In this work, the events are represented by the LAGs and the agents by their partners. The partnerships analysed are those in force in the programming period 2014–2020. During the analysis of the results, it should be borne in mind that the current structure of partnerships is affected by the experience gained from previous editions of the LEADER Programme, so it reflects an evolutionary process that began with the LEADER II edition and lasted several decades.

The LAGs representing the case study were selected by identifying, for each region, a convenient area formed by many neighbouring municipalities, characterised by the high presence of rural development projects. The data used to obtain the affiliation networks were collected based on the documentation retrieved from the official websites of the selected LAGs. We surfed these web resources to collect the documents reporting the partnerships of each LAG. This allowed us to build two affiliation networks, one for each region. Moreover, we searched for information on the activity of each partner to identify its stakeholder category according to the classification provided in the previous section. For the purposes of this analysis, we grouped the stakeholders into five main categories: (i) public institutions, (ii) the agro-food sector, (iii) other producing sectors, (iv) economic associations, and (v) social associations.

The networks were then analysed using network and punctual indexes [36]. The former, along with a visual representation, allowed us to grasp information on the whole web of stakeholder collaborations. We employed some indexes as the number of nodes and connections, the density (i.e., the between the number of actual connections and the maximum number of possible connections), the centralisation index (i.e., the sum of differences between the degree of the most central node and the degrees of all other nodes, divided by the largest theoretical sum; this reflects the extent to which the network is characterised by the presence of one or more very central nodes), the average degree (i.e., the sum of the degrees of all stakeholders divided by the number of stakeholders in the network), and the average distance (i.e., the average number of steps along the shortest paths for all possible pairs of network nodes).

Moreover, we used four punctual indices to measure various aspects of stakeholder power derived from their peculiar position and connectedness in the network. These measures are briefly summarised below.

Degree centrality. This is the number of connections the stakeholder has with others and accounts for a relevant positional advantage, proving that a stakeholder with many connections can directly influence the resources that flow through the network and hold these resources without intermediation.

Betweenness centrality. Another aspect of power arising from the network position relates to the possibility of a stakeholder acting as an intermediary between the others. The betweenness centrality is a measure of this brokerage role of the stakeholder. as it is the sum of links connecting other stakeholders passing through himself [37].

Closeness centrality. This is simply the reciprocal of the farness of a given node, where the farness is the sum of the lengths of the shortest paths to every other node. The ratio of this index lies in the fact that the closer a stakeholder is to all the others, the higher its influence is likely to be.

Eigenvector. This is a more refined measure of closeness, as it is calculated using a factor identifying the “components” of distances among actors. The first component accounts for the “global” distance between stakeholders rather than the immediate closeness [37].

We treated these measures using statistical nonparametric procedures in order to verify if significant differences exist in the amount of power belonging to the various categories of stakeholders. Following [38], two tests were implemented. The Kruskal–Wallis H test was employed to verify whether a global difference between the stakeholders’ centrality measures arises overall. The Mann–Whitney U test was performed to assess which inter-stakeholder categories are significantly different in the centrality measures.

3. Results

Based on the case study identification explained in Section 2, in our investigation, we considered five LAGs of the Apulia region—namely, Gargano, Daunofantino, Daunia Rurale, Piana del Tavoliere, and Meridaunia—and seven for the Veneto case—specifically, Delta Po, Adige, Vegal, Baldo Lessina, Alta Marca, and Prealpi Dolomiti. Figure 1 and Figure 2 represent, respectively, the affiliation networks and the derived one-mode networks of the case studies.

The partnerships are formed by 202 and 235 stakeholders, respectively, in the Apulia and Veneto cases.

Table 2. Network metrics of the two case studies.

Metrics	Apulia	Veneto
No. of nodes	202	235
Public Institutions	69 (0.34)	156 (0.66)
Agro-Food Sector	54 (0.26)	16 (0.07)
Other Producing Sectors	43 (0.21)	15 (0.06)
Economic Associations	17 (0.08)	29 (0.12)
Social Associations	19 (0.09)	19 (0.08)
No. of ties	12,944	10,186
Avg. Degree	64.08	43.34
Deg. Centralisation	0.69	0.28
Density	0.32	0.19
Avg. Distance	1.68	2.40

helps in identifying some key features. The most represented category in both cases is public institutions. It represents most of the stakeholders (66%) in Veneto. Moreover, the private sector (agro-food and other producing sectors) is well represented (26% and 21%) in the Apulia region, while it is less represented in Veneto. Finally, the economic association is the least represented sector in Apulia. In general, the representativeness is more balanced in Apulia than in Veneto.

Apart from this first descriptive analysis, the visual representation reveals that, though it is underrepresented in the Apulia region, the economic association category exhibits high centrality, and several actors from this category bridge structural holes between the various parts of the network. At the same time, two public institutions and a private company also show the same characteristics. In the case of Veneto, this bridging role is played by four major actors, two public institutions, a private company, and an economic association, with four poles formed of town halls and other actors gravitating around them.

Table 2 summarises other key characteristics, showing higher connectivity in the case of Apulia (which outperforms the Veneto network in all the network indexes). In particular, the density of the network reaches 32% in all the possible connections. This index is also rather high in the case of Veneto, as it tops out at 19%. Moreover, the network of the Apulia region is highly centralised (69%), signalling that there are only a few actors collecting the majority of all the existing connections.

Figure 3 reports the average centrality measures of each stakeholder category. It supports the initial hint suggested by the visual representation related to the high connectivity of the economic association category in the case of the Apulia region. This category seems to be the most central with respect to the other measures. This difference appears to be at its highest in the case of betweenness. The Apulia network is also highly connected, and this is reflected in the high closeness of all the stakeholder categories. Apart from the economic associations, another strong category is represented by the public institutions as it shows high eigenvector. In the case of Veneto, the centrality seems more equally distributed between the stakeholder categories, as anticipated by the lower centralisation measure (29%) reported in Table 1. However, in terms of betweenness, an important role is played both by the public institutions and the other producing sectors. On the other side, economic associations and the agro-food sector have the highest degree of centrality in the mean.

To test if these remarkable positions depend on a significant difference in the centrality of the different stakeholder categories, we performed the Kruskal–Wallis H test (

Table 3. Kruskal–Wallis H test of differences in the centrality measures of stakeholder categories.

Centrality Measures	Apulia	Veneto
Degree	0.00017 ***	0.02548 **
Betweenness	0.00005 ***	0.08206
Closeness	0.00033 ***	0.04726 **
Eigenvector	0.00175 ***	0.01853 **

** Significant at the 95% level; *** significant at the 99% level.

). It revealed that there is a significant difference in all the centrality measures in the case of Apulia (significant at the 99% level) and in all except the betweenness in the case of Veneto (significant at the 95% level). In order to identify the sources of these differences, a pairwise comparison was performed employing the Mann–Whitney U test (

Table 4. Mann–Whitney U test of differences in the centrality measures between stakeholder categories.

		Apulia				Veneto
Pairwise Comparison	Degree	Betweenness	Closeness	Eigenvector	Degree	Betweenness	Closeness	Eigenvector
Pub-Agr	0.02088 **	0.61006	0.02382 **	0.02444 **	0.0394 **	0.23404	0.02202 **	0.01046 **
Pub-Pro	0.4009	0.67448	0.50286	0.63122	0.28014	0.4654	0.03156 **	0.02382
Pub-Eco	0.00036 ***	<0.00001 ***	0.00072 ***	0.01174 **	0.01552 **	0.30772	0.00988 **	0.00244 ***
Pub-Soc	0.30302	0.89656	0.3843	0.23404	0.37886	0.20766	0.04136 **	0.0455 **
Agr-Pro	0.0251 **	0.9681	0.0251 **	0.01878 **	0.06576	0.76418	0.59612	0.72786
Agr-Eco	0.00012 ***	<0.00001 ***	0.00016 ***	0.0003 ***	0.8493	0.02382 **	0.77948	0.96012
Agr-Soc	0.83366	0.83366	0.83366	0.81034	0.0164 **	0.98404	0.96012	0.61006
Pro-Eco	0.0005 ***	<0.00001 ***	0.0008 ***	0.00544 ***	0.15854	0.08726	0.76418	0.92034
Pro-Soc	0.67448	<0.00001 ***	0.67448	0.5287	0.77948	0.75656	0.67448	0.53526
Eco-Soc	0.00128 ***	0.00012 ***	0.002 ***	0.0139 **	0.0477 **	0.01684 **	0.98404	0.5552

** Significant at the 95% level; *** significant at the 99% level. Legend: Pub = public institutions, Agr = agro-food, Pro = other producing sectors, Eco = economic associations, Soc = social and environmental associations.

).

As shown, there are significant differences between several pairs of stakeholder categories in both case studies. In the case of Apulia, these pairs are (1) public institutions and agro-food companies; (2) public institutions and economic associations; (3) agro-food and other producing sectors; (4) agro-food and economic associations; (5) other producing sectors and economic associations; (6) other producing sectors and social associations; (7) and economic associations and social associations.

In the case of Veneto, these pairs are (1) public institutions and agro-food companies; (2) public institutions and other producing sectors; (3) public institutions and economic associations; (4) public institutions and social associations; (5) agro-food and economic associations; (6) agro-food and social associations; and (7) economic associations and social associations.

In the case of Apulia, the empowerment of the economic associations category is confirmed. It significantly outperformed all the other categories in all the indicators. This means that the stakeholders belonging to this category are able to directly govern the flow of information between the other stakeholders, play a high brokerage role, and are close to all other partners in general, covering all the dimensions of power accounted for by the network indexes used for this analysis. Surprisingly, the most marginalised category in this area is represented by the agro-food sector, which is systematically subject to the power of the other categories. This means that the stakeholders in this category depend on others, especially economic associations, to gain relevant information related to rural development and need to use their brokerage role to obtain distant partners. Another weak position is covered by the stakeholders of the social and environmental associations, which are in marginal positions in the network.

In the case of Veneto, the power distribution is more balanced. In particular, and also in this region, economic associations play a significant networking role, but the actors from this category are not always essential for the other actors to interact and receive information, as other private stakeholders, both from the agro-food and other producing sectors, can serve as connecting hubs. In this case, the most marginalised category is represented by public institutions. This is due to the fact that the majority of the public institutions in the Veneto LAGs are municipalities, which, because of their territorial competencies, belong to only one LAG. This means that this kind of actor remains locally focused and has little influence outside its territory.

4. Discussion and Conclusions

The aim of this work was to contribute to the analysis of the formation and structure of partnerships formed to implement rural development plans in the context of EU agricultural policies. In particular, the analysis was focused on two case studies, which were used to verify the existence of practical disparities in the empowerment of diverse stakeholder categories.

The analysis presented confirmed that the empowerment of the stakeholders follows unequal paths (R1). This is due to the fact that, depending on the LAG’s context, some categories assume central roles in connecting other partners. While some categories, such as town halls, have very local natures and belong to only one LAG, others play a coordination role beyond the territorial LAG boundaries.

The main difference between the two cases investigated is that this central role is clearly played by the economic associations in Apulia, whereas this power is more evenly distributed among different kinds of stakeholders in the case of Veneto. This also reflects the centralisation characteristics of the two networks, as the Apulia results were highly centralised (R2).

Going more in-depth, the influence of the economic associations in Apulia relates mainly to agricultural associations, which are in almost all the groups. Their fundamental mission is to support and represent farmers to ensure their participation in the formulation and implementation of agricultural development policies and actions. Their central role in the Apulia region stems from both historical and territorial characteristics. From a historical perspective, the aggregation of farmers is a long-term process that went through alternating phases of restructuring agricultural organisations between profitable and efficient professional units, freely managed by farmers, and new types of organisations directed at representing farmers in decision-making contexts. From the territorial point of view, the super-representativeness of the agricultural associations arises from the fact that, being a lagging-behind region, Apulia is still an agriculture-based economy with many entrepreneurial resources within the primary sector. The representative role of the agricultural association in this region represents a strength in the extent to which these cooperation bodies play an umbrella role, providing valuable coordination activity in a densely connected rural development network and promoting information flow and collective action. On the other hand, the weakness of this configuration lies in the fact that the rural development agenda can receive strong pressure from the agricultural sector, pushing more sectoral strategies in turn. This poses a possible risk to the social sustainability of the development strategies that, according to CLLD principles, should prioritise a complex approach focused on the balanced valorisation of endogenous territorial resources in an interconnected fashion.

In the case of Veneto, representativeness is shared between diverse categories of stakeholders. Private companies play a brokerage function, bridging the network holes between more clustered local groups. This is due to the more advanced economy and brings some interesting opportunities for increasing pluralism and inclusion as more agents become active in decision-making; this contributes to a more comprehensive rural development agenda and eases the problem-solving process as more diverse resources are pulled together. On the other hand, the network is sparser and shrinking, posing possible challenges in terms of extended coordination and collective action.

This analysis has shown that partnership configurations can be very different with respect to stakeholder empowerment and that a suitable tool to reveal these differences is SNA. Indeed, it is a technique that allows for in-depth analysis of partnership compositions by deriving useful insights for policy designers. In particular, it can reveal biases even in cases of networks that seem very dense and well connected. The analysis of different aspects of power, such as those captured by the indicators used in this paper, can be useful in identifying imbalances in the distribution of power and, ultimately, in the representativeness of interests within the partnership. In other words, even partnerships that seem well balanced in terms of the representativeness and numerosity of stakeholder categories may be unequal in terms of empowerment.

This is a valuable indication for policymakers committed to future rural development processes at higher hierarchical levels, such as the Managing Authority of the LEADER Program, represented by the Regional Administrations in the case of Italy. For such actors, it is useful to know not only who the strongest and most marginal actors are but also how unequal the power is between these categories. This power distribution can give rise to different configurations. From the analysis presented in this paper, for example, two were discovered: one in which all the power over information brokering and control is concentrated in the hands of one category and one in which—while there is some level of inequality—several categories may play information-sorting roles. These two configurations are very different. In the first case, actors will have only one channel available for updates. In the second case, while there is a concentration of power, there are several alternatives in terms of information channels. This obviously reduces the risk of marginalisation even when there is a disagreement between stakeholders. Policymakers also see their informational and training roles restricted in the first case rather than in the second.

This analysis also provides indications of adjustment interventions: in the case of a strong concentration of relational power, as in the case of Puglia, policymakers will necessarily have to involve the strongest category as a priority and use it to play an aggregating role. Then, they can design appropriate engagement strategies targeting the categories at the greatest risk of marginalisation. These may be differently achieved depending on the categories targeted. This action can be implemented through pilot camps and entrepreneurial visits in the case of the agri-food sector, or it can be represented by initiatives addressed to the citizens with regard to the third sector and civil associations.

In territories where the distribution of power is not as concentrated, on the other hand, it is possible to foster cross-sectoral collaborations and focus more on the involvement of the most marginalised categories in the initial phases.

Some limitations of this research relate to the fact that it analysed the structure of the partnerships but not the reasons underlying the surveyed networks’ construction (e.g., trust between specific segments of the stakeholders, economic interests, informal connections, etc.). To answer these questions, a combination of approaches should be carried out, including historical analyses of past rural and agricultural local projects, in-depth interviews, and ad hoc surveys with stakeholders. Moreover, since the analysis presented was focused on the power differences between the stakeholder categories, it would be useful to refine the grain of the analysis by extending it to the assessment of power differences between actors belonging to the same category. Finally, the domain of the research could be extended by including comparisons between European case studies.