As social capital is a phenomenon that has conceptual origins in network approaches (Borgatti and Halgin, 2011; Yu et al, 2017), the configuration of a project, e.g., a regional development project, in which the social capital emerges, must be understood as a specific network, which becomes connected with surrounding networks of the region (Beckham et al, 2023; Cooke et al, 2005; Sørensen, 2016). The stakeholders (individuals or organisations, represented by individuals) interact within networks incentivised by a particular interest which is of relevance for the project and can be supported through the project (Eskerod et al, 2015; Freeman, 2010). More specifically, during projects, such as regional development projects, stakeholders establish contacts, plan and share activities over a given period of time. These activities do not only support the participation of regional stakeholders in decision processes during the project work, but also the development of social capital. Hence, we assume that each stakeholder in a project shapes the social capital through relationship-building, which takes place embedded in the project activities (Bizzi and Langley, 2012; Brunet et al, 2021). It can thus be argued that, the higher the level of stakeholder engagement will be—as induced by the project management—the more likely social capital is being developed, which, in return, supports a knowledge transfer into the region (Bartsch et al, 2013; Greenwood, 2007; Sørensen, 2016; Yu et al, 2017).

As a point of departure to conceptualise social capital, this paper relates to the broad discussion in the literature on social capital and networks: Borgatti and Halgin (2011) argue that classical network theories, such as the social capital approach, treat networks as an exogenous variable, but that emerging network structures will also be shaped internally by stakeholders with their interaction in networks. Hence, contrary to this view by network scholars, the agency of stakeholders (in the wording of network theories: “actors”) in terms of building new relationships and developing trust will intentionally support both the quantity and quality of the social capital emerging between these stakeholders (Falk and Kilpatrick, 2000).

In fact, the emergence of social capital can be understood as a process which is shaped by the interaction among stakeholders in the form of activities, initiated by the leadership of project management and/or due to the agency of its participants (Falk and Kilpatrick, 2000; Prokkola and Jakola, 2024). In the case of regional development projects, these stakeholders are the project manager(s) and the participating stakeholders. Undoubtedly, the agency in managing regional development projects with a focus on relationship management supports the emergence of “good”, lasting social capital (Harmaakorpi and Niukkanen, 2007; Iyer et al, 2005; Lundin and Söderholm, 1995; Nilsen et al, 2023). The more frequently, on the one hand, and the more friendly and constructively, on the other, the activities in the project will take place, the more likely it will be that such lasting social capital will accumulate, which benefits the project and, indirectly, the region.

However, this does not necessarily mean that the social capital is positive as such (Falk and Kilpatrick, 2000; Woolcock, 1998), because the teams of stakeholders that interact in projects are different in their levels of stakeholder engagement, as shown during the project work. Various individual or team-related mechanisms exist, such as leadership and communication style (Turner and Müller, 2005), conflict-resolution capabilities (Pinto, 2000; Stepanova, 2015), etc., which will influence both the process and quality of emerging social capital. Since project teams show different forms of stakeholder participation and are moreover associated with various levels of stakeholder engagement, emerging social capital from project work can be positive in terms of being productive in a targeted way, or not. Notably when mechanism that moderate the process of emerging social capital, such as bridging social capital are missing during project work, it can happen that the emerging social capital is less productive, i.e., does not generate benefits for the region.

As defined above, and in line with social capital theory (Burt, 2005; Coleman, 1990), emerging social capital is influenced by the underlying social structures that shape the availability of resources, which are residing in the relationships of stakeholders (Maak, 2007). Social capital can be further categorised as bridging and bonding social capital, which are two forms of emerging social capital that complement each other and may be beneficial for regional development projects (Callois and Aubert, 2007; Ceci et al, 2020; Sørensen, 2016; Woolcock and Narayan, 2000; Woolcock, 1998). While the so-called bonding social capital (Coleman, 1990; Sørensen, 2016) mainly stabilises the interaction of stakeholders due to the existence of rules, norms, sanctions, trust, or specific institutions, the so-called bridging social capital (Burt, 2005; Sørensen, 2016) supports the expansion of relationships in projects and the acquisition of information, knowledge, and other resources from wider circles of stakeholders.

Both categories of social capital seem to be important for regional development projects. It can be argued that the mere focus in the project management on bonding social capital bears the risk of lock-in-effects, so that the current capabilities united in the project will not be adapted to future trends (Bresnen, 2007; Eklinder-Frick et al, 2011; Pillai et al, 2017). Therefore, the complementary bridging social capital can help projects with existing bonding social capital to update the capabilities and resources in the project by means of drawing wider circles, activating new relationships, and acquiring new ideas and resources from stakeholders outside the focal setting of the project (Eklinder-Frick et al, 2012). This may be considered as a means of increasing the likelihood that regional solutions can be adapted to future trends and even turn into innovation that may provide the region with competitive advantages (Leick and Gretzinger, 2018; Maak, 2007).

Numerous empirical studies have demonstrated that the outcomes of regional development projects, such as specific project-related or regional goals (Horlings, 2012; Nicholds et al, 2017), including the social capital that is established during such projects in terms of specific (types of) relationships, are contingent upon the quality of the project management with an agency provided by the project manager(s) to lead the group of project participants. In fact, leadership skills, such as general networking capabilities (Merkaj et al, 2020), and proactive leadership approaches that initiate relevant project activities, which pertain to the gatekeeping and brokering of contacts, knowledge, and information (Devine-Wright et al, 2001; Gretzinger and Leick, 2017), form part of the agency (Nilsen et al, 2023) that characterises the quality of project managers (Harmaakorpi and Niukkanen, 2007).

Social capital represents an important outcome that is commonly strived for in regional development projects, as it is considered to be supportive of achieving pre-defined goals based upon the relationships built in the projects, such as, for instance, innovation and economic growth (Muringani et al, 2021; Pihkala et al, 2007; Westlund and Larsson, 2021), or environmental and social goals (Andersson, 2009; Rius-Ulldemolins and Gisbert, 2019). The literature also acknowledges that the social capital strived for in regional development projects is often burdened or constraint by specific regional conditions (for example, rurality, peripherality, or organisational thinness, Leick and Gretzinger, 2020; Isaksen and Karlsen, 2016) and thus requires shared understandings of the goals in the project (Bentley et al, 2017; Tietjen and Jørgensen, 2016), which, again, points to the agency of key project stakeholders, who are supposed to create these shared understandings.

Furthermore, in the project management literature, the involvement of stakeholders through participatory approaches is described as a means of supporting social-capital building in projects (Baker et al, 2010; Holman, 2008). Besides agency based upon project leadership, the importance of governance through the involvement of individuals or teams among the project participants is also acknowledged (Biesenthal and Wilden, 2014; Derakhshan et al, 2019). With regard to regional development projects, stakeholder involvement is confirmed to affect the project outcomes in a positive way (Lukman et al, 2009; Prager and Freese, 2009). However, the specific modes or levers of how this participation of stakeholders is achieved in order to build social capital and which qualities of social capital this involvement may imply—or not—remain unclear in this literature.

During regional development projects, stakeholders interact, plan and share activities. These activities are triggered by both stakeholder engagement and stakeholder participation. Bonding and bridging social capital play an important role here, because this social capital is not only inherent in the social structures of projects but emerges as well due to agency in the course of time. Put differently, activities initiated and managed by key stakeholders and the way how these activities are embedded in the projects have an impact on the emergence of new social capital.

The Danish-German cross-border region connects the county of Southern Jutland in Denmark with the federal county of Schleswig Holstein, Germany, and is home to approximately 700,000 citizens. The major industries are logistics, tourism and food energy, clean-tech, and mechanical engineering (Gretzinger and Ingstrup, 2022). The concept of the chosen regional development project, which was based upon the participation of stakeholders from both countries, was to form nine project teams that would enhance relationship-building and knowledge exchange by running targeted activities with the goal to build solutions for various regional challenges. Originally, nine project teams were planned in the funded EU project, but only eight of them managed to finalise all or most of their targets (project team 3 did not manage to move beyond the planning phase but is still included in the empirical part of this paper).

Each project team worked on a specific regional development issue in the following areas: food, tourism, CO₂-neutral heating, funding, accessibility, recruitment of labour, broadband, networking and exchange, digital infrastructure and mobility. Moreover, each team was guided by a project manager, who oversaw the facilitation of the process by agreeing upon goals, developing activities, and implementing action plans to render the identified problems manageable (Gretzinger and Ingstrup, 2022).

In contrast to traditional approaches to measure social capital through structural data and network analysis (Borgatti and Halgin, 2011), this paper does not count stakeholders and their ties, but instead illustrates and describes the process through which social capital emerges, based upon a combination of quantitative data that explore the use of relationships and qualitative interview findings, which aim to develop a rich understanding of the phenomena studied (Holman, 2008).

To study the research question addressed in this paper, two approaches were combined for the data sampling. Firstly, a bilingual online questionnaire (in Danish and German) was used that asked the project members, based upon scaled variables, to measure issues related to the project work and project management, such as the mutual support, their level of commitment, their stakeholder engagement and stakeholder participation, and the emerging social capital from their perspective. A total of 50 stakeholders were initially invited to fill in the questionnaire during September 2019, of which a total of 39 stakeholders answered it, which corresponds to a response rate of 78 per cent. Secondly, additional in-depth interviews were conducted with n = 15 project participants in 2019 that explored in more detail the various activities and processes connected to these activities during the project period. The interviews were assembled and conducted in line with a purposive sampling strategy (Miles et al, 2014). The goal of the in-depth interviews was to uncover problematic but also positive examples, such as specific contexts, problems or results of project activities. The interviewer was an independent person from a regional consultancy agency, a fact that safeguarded a high level of neutrality. These qualitative data, which were gained through personal interviews, complemented the data sample based upon the questionnaire, because they allowed for an interpretation and evaluation of the leadership style, notably regarding the question of how it supported, or hindered, the translation of stakeholder participation and/or stakeholder engagement into relationship-building during the project work. The data from the interviews were coded on a lower-theme level (Saldaña, 2021), as they are primarily used in this paper as additional explanations that aim to reach in-depth understandings of why and how the activities and processes supported or hindered stakeholder engagement/stakeholder participation and how this interacted with the emerging social capital.

More generally, combining these two methodological approaches offers two advantages (Cameron et al, 2015; Kelle, 2006). Firstly, the initial quantitative analysis, which was based upon a survey, measures the subjective and broad evaluation of various issues by the project participants, and it serves as an outside perspective. For example, social capital is approximated by differentiated variables that respondents evaluated, ranging from frequent (but not necessarily intense) contacts during the project and professional networking relationships (e.g., the use of relationships for receiving consultancy services) to friendship relationships, which are likely to be sustained. Another example is that the survey enabled a precise evaluation of how “good” or “bad” the context of the project work and project management were experienced by the respondents, for example, the support by the leader of the team or by project members. The subsequent interviews allowed the respondents, who are project participants, to report in more detail about their inside view on the project collaboration, including the emerging social capital, and explain specific activities conducted.

Secondly, the combination of these two empirical approaches (quantitative surveys and qualitative interviews) has been utilised to actively engage stakeholders in the evaluation of the project. It was communicated that, through the empirical fieldwork, stakeholders could gain a voice. Beyond the purely academic research, the empirical materials were thus used for reflections on the project which had been discussed internally on project meetings. While the quantitative survey offered the stakeholders the possibility of grading activities, the qualitative in-depth interviews provided some space for reporting and sharing the individual experiences. In addition to these advantages, the validation of the data could be improved through the combination of methods used in this paper (Kelle, 2006).

For the qualitative data sampling through interviews, sensitising definitions in line with Bowen (2006) and Blumer (1954) (see Section 1. Introduction) were used that offered room for interpretations of social capital, notably with regard to the difference between stakeholder participation and stakeholder engagement and the understanding of social capital by the respondents. However, for the quantitative data analysis, these sensitising definitions were translated into definitive operationalisable concepts that express the phenomena studied in terms of well-tested scales and proxies.

To this aim, the concept of stakeholder participation was applied in line with the definition presented in Section 1, and following Lehtinen and Aaltonen (2020), it was operationalised as the perceived participation in the case project team. More specifically, the respondents were shown a figure with the five ladders of citizen participation (adapted from Arnstein, 1969; Luyet et al, 2012) with the following steps included: (1) partnership (defined as a high interdependence on the local and trans-border level), (2) high participation (with a high interdependence on the project level), (3) participation (with a low interdependence on the project level), (4) low participation (with information-exchange, but no real impact by the respondent), and (5) non-participation (defined as a casual acquisition of information). Stakeholder engagement was operationalised based upon the degree to which the respondent found the project study work meaningful and purposeful. More specifically, according to the definition used (Lehtinen and Aaltonen, 2020), the 17 items of the Utrecht Work Engagement Scale (Schaufeli et al, 2006) were adapted to the context of the investigation.

Social capital was measured based upon the frequency of contacts with other project team members and two types of relationships: a consultancy relationship (the question of whether the relationship between the project team members is considered to be a professional relationship) and a friendship relationship (the question of whether the relationship with others in the project could be described as a friendship; Weiler and Hinz, 2019). To explore how well supported the stakeholders felt by the leadership and the other collaborating stakeholders, further variables were included that denote specific characteristics of the collaboration in the project teams, such as mutual support, communication, cohesion and effort (Hoegl and Gemuenden, 2001), and, in addition, variables that measure social skills (or the general social ability of project stakeholders) and their general self-efficacy in order to operationalise the respondent’s self-perceived project management skills (Ferris et al, 2001; Schwarzer and Jerusalem, 1995). All responses were measured through five-point Likert scales (values ranging from 1 = strongly agree/never to 5 = strongly disagree/always).

Finally, because the exploration also aimed to understand how different constellations of leadership and stakeholder interaction within the various project teams influence the phenomena studied, the affiliation to the project team was implemented as a supplementary variable “project team”.

The quantitative data were analysed by means of a JCA in a deductive manner and in line with Greenacre (1988, 1984). A JCA is a procedure to analyse quantitative data according to three consecutive steps: firstly, in the case of this paper, the 15 in-depth interviews were transcribed. Secondly, they were clustered according to the following themes: participation and engagement of stakeholders (low/high/characteristics); the activities of stakeholders in terms of social or professional activities (frequency/characteristics); and the drivers and barriers of processes. Thirdly, the clustered data were interpreted according to Bizzi and Langley (2012) by using the patterns of actors, resources and activities that matter for social networks and by studying how these three fundamental elements that constitute networks sequence each other regarding the phenomena of interest. All data from the interviews were treated anonymously, and the project teams were assigned numbers to avoid an identification of their members.

The general idea of a correspondence analysis is related to the main component analysis and enables the presentation of the correlations or, more precisely, the correspondence between the categories of cross-classified tables. A JCA is applied to empirical data that are assigned to contingency tables involving three or more variables or sets of categories. For both correspondence analysis, in general, and JCA, in particular, in the graphical overview, the common position of the categories in distance from the center is to be interpreted as the correspondence of the categories. The closer the categories are located to the center, the higher the correspondence of the categories will be, and vice-versa. The dimensions (or axes in the graphical overview) present the variations of the input data with decreasing amounts of explanation. To perform the JCA, the values of the presented scales were recoded from mean scores into integer values (1: high, 2: medium, 3: low). Stakeholder participation remained at its original scaling, ranging from 1: high participation to 5: very low/non-participation. As a result, various clusters of stakeholder engagement could be identified.

As a first step, and based upon the results of the JCA, a two-dimensional solution is established (Fig. 1), which explains 79.4 per cent of the principal inertia. The first axis (dimension 1) absorbs 64.3 per cent of the variance, and the second axis (dimension 2) 14.9 per cent. The colour of the data in Fig. 1 points to the contributions that the variables make to the dimension: orange constitutes a high, yellow a medium and blue a low contribution.

Fig. 1.

Participation and networking as two dimensions. PP, Stakeholder participation: PP++ very high, PP+ high, pp medium, pp- low, pp– very low; SE, Stakeholder engagement: SE+ high, SE medium, se- low; F, Frequency of contact: F+ high, F medium, f- low; CR, Consultancy relationship: CR+ high, CR medium, cr- low; FR, Friendship relationship: FR+ high, FR medium, fr- low; Q, Quality of teamwork: Q+ high, Q medium, q- low; SK, Social skills: SK+ high, SK medium, sk- low; E, General self-efficacy: E+ high, E medium, e- low; cs1-cs9, Case Study in which the stakeholders participated (supplementary variable). Dim 1 is named “low social capital”; Dim 2 is named “strong social capital”, see the findings in Table 1.

In general, dimension 1 (named: low social capital) distinguishes the non-engaged stakeholders with low social capital (measured through their frequency of interaction, friendship and/or consulting relationships), who perceive no participation among the project team members, from the highly engaged stakeholders, who perceive the project work in the team as participatory. More precisely, the positive side of dimension 1 (in mathematical terms) is characterised by a low frequency of contact among the case study members, a low perception of both friendship and consultancy relationships among them, a medium level of general self-efficacy and social skills, and a low level of the project work quality in the team. Not surprisingly, this side of dimension 1 is also characterised by both a low stakeholder engagement and low stakeholder participation. By contrast, the negative side of dimension 1 (in mathematical terms) is characterised by a high perception of the collaboration quality, high perceptions of friendship relationships among the project team members, and both a high stakeholder engagement and high stakeholder participation.

The results for dimension 2 (named: strong social capital) in the JCA are more ambiguous, compared to dimension 1. Dimension 2 distinguishes the non-engaged project team members with strong social capital but with both high and low stakeholder participation from the medium-engaged members with mediocre networking activities and mediocre stakeholder participation. In more detail, the positive side of dimension 2 (in mathematical terms) is characterised by project team members with both low and very high stakeholder participation, a high frequency of contact, high perceptions of friendship and consultancy relationships in their team, a low level of perceived teamwork quality, and a low stakeholder engagement. The negative side of dimension 2 (in mathematical terms) is characterised by project team members with a medium level of both stakeholder participation and stakeholder engagement, a medium frequency of contact, a medium perception of friendship and consultancy relationships among the project team members, and a perceived teamwork quality on medium level.

To complement the JCA results with in-depth insights for the individual project teams, the following findings are derived from the personal interviews that shed light on specific associations within the teams.

Project team 1: In project team 1, many social and professional activities were initiated. The professional activities were targeted to the demands of the stakeholder so that any stakeholder could receive some benefit from the project collaboration. All stakeholders in this team were very engaged, as the leadership involved them from the very beginning. The leadership style (the leader was Danish) in project team 1 was described as collegial, based upon dialogue, common ground and showing a common sense about the project management. Because of this, most of the stakeholders perceived that they reached high levels of participation in the project. A tremendous volume of targeted activities initiated by both the leadership and the most active stakeholders shaped manifold communications channels and supported the commitment of individuals. This happened early in the project process, as one interviewee in this team acknowledges: “we did quite soon manage to shape plenty of communication channels with and around the key actors from the municipality, important regional institutions and the German partners”.

The stakeholders in team 1 were not only highly engaged, but also achieved high levels of participation, and their formal commitment to the project supported their participation within the team as well. Even those stakeholders who were not formally committed to the project reached high levels of engagement, but not high levels of participation. However, both the stakeholders that achieved high levels of engagement and participation and those, who were engaged but did not manage to formally participate in decision processes, contributed greatly to the development of emerging social capital. Within team 1, strong relationships were established that are likely to continue after the formal project end.

Project team 2: This team was also committed to many customer-oriented activities (targeted to defined goals and a defined target group, respectively; the leader was a Dane), and the stakeholders were involved in these activities from the very beginning of the project work. As a result, many relationships could be established, and possible future collaborations were discovered that offered the potential to benefit the stakeholders due to synergies. These potentials identified supported the relationship-building during the project work. Most of the stakeholders in project team 2 achieved high levels of both stakeholder participation and stakeholder engagement. Emerging social capital can be identified for this project team, similar to project team 1. Due to the future potential for a long-term collaboration, it is likely that the relationships established will last beyond the formal project duration. Those stakeholders, who were not formally involved in the project, however, were mostly not involved in the project activities of team 2 and thus showed no engagement, which is a difference to team 1. One interviewee stated this as follows: “even if a few network partners got motivated to be engaged, …the network partners were often inactive and not really integrated into the team”.

Project team 6: In this case, the goals of the project team were not developed in a bottom-up process but mostly defined by the project leader. This led to a situation in which two sub-groups emerged. The Danish stakeholders worked in one sub-group that was interested in developing specific tools to support recruitment in rural areas, whereas the German stakeholders were in another group that was more interested in receiving an overview of various tools that could be developed. One interviewee from project team 6 described these reasons for this development as follows: “…some participants were thrilled, but, all in all, too many practitioners were not committed. … For me as leader, it was difficult to find out who would support the project and who would not”. Finally, a lack of commitment and dialogue lead to a fragmentation of the team and a missing team spirit. Only a few formally committed stakeholders showed higher levels of stakeholder engagement in project team 6. Notwithstanding this, even if there was no cohesion in this team and the level of engagement was low, many contacts and relationships emerged. This was apparently due to the manifold activities that were put into action. Hence, social capital emerged in project team 6, but the team did not develop a common theme that could represent a linking pin for future activities.

Project teams 4, 5, 7, 8: In these four project teams, there are some common characteristics. Various challenges affected the project work in these teams. Some teams started late and had not enough time to establish a structure that was goal-oriented and based upon a clear strategy. Other teams developed such a structure but lacked stakeholders with an interest in the topics and goals discussed. A common denominator with the four project teams is that the level of activities was low with all of them. In a similar vein, the level of stakeholder engagement that the teams reached remained low, and only very few stakeholders were observed with high levels of stakeholder engagement, who, however, rather played their own game, instead of being really committed to the project team, as one interviewee reports: “it seems to be that the Danish colleagues are embedded in other structures as we. Often, I had the feeling they would have liked to participate, but due to other priorities, the planned activities needed to be cancelled”. In these four cases, social capital did not emerge on the team level, and social relationships are only based upon few and isolated relationships among the stakeholders. The leaders of these teams were in one case Danish, and in 3 cases German.

Project team 9: In team 9, no common ground emerged that could lay the foundation of how to interact in the project team. There was no dialogue about identifying and achieving goals or developing activities to achieve these goals. The leader, a German, managed the team in a top-down manner. The team, again, divided itself in two sub-groups (similar to project team 6): those who were capable to live up to the targets given by the leader, and those who were not. The following quotation by an interviewee summarises this situation: “the problem was that the team-members apparently had not the feeling of being recognised by the leader, and some were overdoing with paying respect”. Due to some advice provided by the leader, the team managed to deliver output in terms of solutions to the problems addressed. However, the level of stakeholder engagement was overall low, even in the active sub-group, and it was close to nothing with the other sub-group. One of the interviewees reported that some stakeholders had even developed fear for the leader. In project team 9, some new relationships emerged on the individual level, but no social capital emerged that could support relation-building in line with the business-related topics that were at the center of this project team.

In summary, and comparing the findings from the individual project teams, the interviews revealed that the level of stakeholder engagement is influenced by the quality of leadership, the number of proactive stakeholders, and the quantity and/or quality of the activities performed. Notably those activities that interested the stakeholders led to their high levels of engagement. To reach a range of such activities that seem to benefit many stakeholders in the project team, an early involvement of stakeholders by leaders during the project duration seems to be necessary. The creation of activities was also key to engage those stakeholders, who were not formally appointed to the project (they were called “network partners” and did not receive any resources for, e.g., travels or compensation of their work hours invested in the project).

However, and more generally, the interviews also showed that high levels of stakeholder participation require well-designed and stakeholder-oriented activities, based upon commitment through formal appointments of and delegation of tasks to stakeholders, because both the engagement and participation of stakeholders in the teams affect the emergence of social capital. The most important driver to achieve this is leadership, which needs to be based upon dialogue between leader and project stakeholders, including those non-formally appointed ones. These non-formally involved project stakeholders needed to be motivated by the leader through the activities designed. Leadership was, moreover, paramount regarding the design of goals and activities that commit all involved stakeholders. These activities encompassed both targeted activities according to the stakeholders’ interests and social activities that support relationship-building (see project teams 1 and 2, in particular). By contrast, a purely target-oriented leadership that only orchestrates project goals, instead of developing them so that the stakeholders may relate to them, was found to be rather destructive, because the project outcomes will then only depend upon the good-will of stakeholders (see project team 9, in particular).

For the emergence of social capital, i.e., to achieve frequent and close stakeholder interaction in the project teams, leadership was, again, crucial: the leader had the task of designing activities according to the interests of the various stakeholders and involve them from the start, for instance, by allocating responsibilities to those stakeholders, who were both capable of and willing to take them over. While the mere existence of relationships that were developed in the project teams determine the bridging social capital, the quality of the solutions developed through these relationships—which denotes the output of the project teams that could be used by the stakeholders and their identification of future possibilities (see teams 1 and 6)—represent the bonding social capital.