Fostering public value through a collaborative organizational culture in highly vulnerable social contexts

The concept of public value has gained increasing relevance in public management theory and practice, yet it remains the subject of multiple—and sometimes conflicting—interpretations. Initially conceptualized by Moore (1995), public value refers to what is considered valuable by society and created through legitimate and effective public action. Moore's framework introduced a managerial perspective centered on strategic alignment between value creation, political legitimacy, and operational capacity. Over time, however, the concept has expanded to include normative, ethical, and deliberative dimensions (Benington and Moore, 2011; Nabatchi, 2018), reflecting a shift toward participatory and network-based governance models (Osborne, 2006).

The paper explores the relationship between public value creation and collaborative culture, defined as a relational process grounded in mutual respect, shared vision, and trust (Sørensen and Torfing, 2011). A growing body of research highlights the positive relationship between a collaborative culture and the capacity to effectively address public needs (Bommert, 2010; Sutton and Shouse, 2016; Torfing, 2019). Scholars emphasize the role of collaborative culture in network building, public sector innovation, and, more broadly, in the creation of public value (Sancino et al., 2018; Van Gestel and Grotenbreg, 2021). From this perspective, collaboration—built on mutual respect, trust, and a shared vision—is central to the creation of public value. Within this framework, the concept of co-production has been analyzed as a collaborative practice and an essential tool for generating public value (Bovaird and Loeffler, 2012; Cordella et al., 2017). When adapted to the public sector, co-production can help overcome bureaucratic constraints and more effectively meet social expectations. The theoretical argument driving this study is that a collaborative culture serves as a foundation for two essential organizational dynamics: first, it strengthens collaboration with community stakeholders, ensuring that objectives and activities align with local social and economic needs; second, it supports the development of an internal organizational model based on distributed leadership and professional knowledge-sharing (Liguori et al., 2012). The decision to focus on collaborative culture as the starting point of the proposed framework is grounded in the nature of public value outcomes in education, as conceptualized in public management (Benington and Moore, 2011; Bryson et al., 2014; Moore, 1995). Public value in adult education does not result from the isolated action of individual teachers, but from complex organizational outcomes and impacts—such as skills development, inclusion, employability, and social integration—that require coordinated and coherent collective action (OECD, 2016; Osborne et al., 2016). These outcomes presuppose alignment among teachers, continuity across educational pathways, and integration with external stakeholders, making organizational cohesion a necessary condition for public value creation (Bryson et al., 2015; Talbot, 2009).

A collaborative culture provides the organizational foundation for this cohesion by enabling distributed leadership, in which teachers assume roles beyond classroom instruction, including coordination and mentoring (Gronn, 2002; Harris and Jones, 2018; Spillane, 2006). At the same time, collaborative culture supports the development of knowledge-sharing systems that allow educators to collectively reflect on student needs and adapt teaching strategies in response to complex and heterogeneous learner profiles, fostering organizational learning and continuous improvement (Fullan, 2018; Kools and Stoll, 2016). Moreover, public value in adult education extends beyond internal educational processes to include the external usability of acquired competencies in the labor market and in society. This external dimension of public value requires collaborative governance arrangements with local institutions and associations, consistent with network-based and co-productive models of public service delivery (Bryson et al., 2015; Emerson et al., 2012). Such outward-oriented collaboration is sustainable only when it is rooted in a strong collaborative culture, which enables organizations to engage coherently and strategically with their institutional environment (Ansell and Gash, 2008; Osborne et al., 2016).

However, collaborative culture, distributed leadership, collaborative governance practices and knowledge-sharing systems do not generate public value directly. Their contribution materializes through the adoption of student-centered teaching processes, which represent the operational mechanism through which organizational conditions are translated into educational outcomes. Student-centered teaching emphasizes responsiveness to learners' needs, personalization of learning pathways, active engagement, and recognition of students' social and experiential backgrounds—features that are particularly critical in adult education contexts characterized by vulnerability and marginalization (Hargreaves and Fullan, 2012; OECD, 2016; Ryan and Deci, 2000). When supported by collaborative organizational conditions, student-centered teaching enables the coherent development of skills, inclusion, and employability-oriented competencies, thereby constituting the proximate mechanism through which public value is effectively created in educational institutions.

In this sense, internal organizational dynamics shape the conditions under which student-centered teaching processes can be enacted consistently across the organization, making public value creation observable in everyday professional practices and assessments.

These elements are crucial for fostering an adaptive and inclusive environment, which is particularly necessary in contexts characterized by high levels of social and economic marginalization (Bracci et al., 2019).

Despite the growing theoretical interest, several gaps persist in literature. First, public value is often treated as an abstract, normative ideal, with limited empirical investigation into its concrete generation within public organizations, despite the existence of emerging measurement frameworks and case-based studies (e.g. Deidda Gagliardo et al., 2025; Papi et al., 2018, 2025; Spano, 2014). Second, while collaborative governance has been widely discussed, the internal organizational conditions that enable effective co-production—particularly the role of collaborative culture—are underexplored. Third, public value creation has rarely been examined in the domain of adult education, especially within institutions serving marginalized populations.

This study addresses these gaps by investigating how collaborative culture shapes the capacity of public educational institutions to generate public value. The analysis focuses on Centri Provinciali per l’Istruzione degli Adulti (CPIAs) in Sicily (Italy), which plays a critical role in promoting lifelong learning, social inclusion, and workforce integration for disadvantaged groups. Drawing on literature in public management, organizational learning, and educational leadership (Fullan, 2018; Bryson et al., 2015; Harris and Jones, 2018), we theorize that collaborative culture enables distributed leadership, organizational learning systems, and collaborative governance practices, which in turn support the adoption of student-centered teaching strategies—ultimately contributing to public value creation.

Accordingly, the research question guiding this study is:

The study adopts a quantitative research design and applies Structural Equation Modeling (SEM) to test the hypothesized relationships. Data were collected through structured questionnaires administered to teachers in CPIAs across Sicily. By contextualizing public value creation within the daily practices of adult education institutions, this research contributes to the theoretical debate on value co-creation in public services, while offering actionable insights for policy and management in education systems serving vulnerable communities.

Educational institutions have traditionally been assessed in terms of their contribution to human capital formation and individuals' economic prospects. In recent years, however, an expanding strand of literature has highlighted that education systems—particularly when embedded within public policy frameworks—also perform broader social functions related to social cohesion, inclusion, and collective well-being. Beyond economic and performance-oriented perspectives, some scholars have explicitly conceptualized schools as generators of public values in normative and societal terms. Ozoliņš (2010), for instance, describes schools as “moral habitats” in which the formation of moral virtues contributes to the development of shared public values in pluralistic societies. In this sense, education plays a foundational role in shaping civic dispositions and sustaining the moral infrastructures upon which democratic life depends. From a public management perspective, this implies that educational performance cannot be adequately assessed through financial or output-based indicators alone, but must instead be evaluated in relation to broader public value outcomes that reflect societal goals and collective benefits (Benington and Moore, 2011; Bryson et al., 2014). In line with this view, the public value literature conceptualizes performance as a multidimensional construct encompassing collectively desirable outcomes, social legitimacy, capability enhancement, and societal cohesion (Meynhardt, 2009; Meynhardt and Jasinenko, 2021).

At the same time, it must be specified that this multidimensional nature leads Public Value being more difficult to measure empirically. Indeed, several scholars have highlighted important conceptual and methodological challenges related to the definition, operationalization, and assessment of Public Value (Rhodes and Wanna, 2007; Brown, 2021; Prebble, 2021). Furthermore, no universally accepted framework for measuring Public Value currently exists, as different approaches emphasize distinct dimensions and evaluation criteria depending on the organizational and institutional context considered (Faulkner and Kaufman, 2018). More recently, scholars have also distinguished between objective and subjective perspectives on Public Value measurement, further enriching the debate on how Public Value can be assessed empirically (Papi et al., 2025; Deidda Gagliardo et al., 2025). Despite these challenges, a growing body of empirical research has sought to operationalize Public Value and develop tools for its assessment. In particular, subsequent empirical research has developed and validated measurement scales of perceived public value, clarifying its core evaluative dimensions and strengthening its operationalization (Meynhardt and Bartholomes, 2011; Meynhardt and Jasinenko, 2021). This body of work has also demonstrated that organizational public value operates across levels: it shapes institutional performance at the macro level while influencing employee engagement and work experience at the micro level (Meynhardt et al., 2024). Recent scholarship has further problematized the conceptualization of value in public services, emphasizing that public value consists of multiple interrelated elements that emerge within service ecosystems and are experienced across different stakeholders (Cui and Aulton, 2023). Together, these contributions emphasize the relevance of examining how public value is generated and recognized within educational organizations.

To study the determinants of public value creation in education, we draw on a multidisciplinary body of literature spanning public management, organizational learning, and educational leadership (Bryson et al., 2015; Fullan, 2018; Harris and Jones, 2018). These fields provide complementary perspectives for understanding how educational institutions can align internal practices and governance mechanisms with broader societal goals. Specifically, public management contributes insights into public value concept and governance mechanisms; organizational learning emphasizes the role of adaptive processes and knowledge-sharing systems; and educational leadership highlights the importance of vision, collaboration, and capacity-building in schools. Together, these perspectives form the conceptual foundation for analyzing how collaborative cultures, leadership distribution, and learning systems contribute to the implementation of student-centered practices and, ultimately, to the creation of public value. The creation of public value in education is realized through schools' ability to generate tangible and meaningful impacts on society, improving both individual and collective opportunities for students (Moore, 1995, 2013). In the case of CPIAs, the concept of public value takes on particular relevance: these institutions aim to create value through inclusive access to education, improved individual skills and better social inclusion (Fullan, 2018; OECD, 2016). Education contributes to the common good when it responds to the needs of society, promoting equity and reaching out to all individuals, especially the most vulnerable, to actively engage in social and economic life (Bryson et al., 2014; Osborne et al., 2016). The acquisition of basic skills, the development of transversal competencies, the promotion of inclusion, support for socio-emotional skills, and the strengthening of connections with the labor market are all elements that help transform education into a driver of equity and social development (Dumont and Istance, 2010; Marsick and Watkins, 1999). These dimensions reflect direct organizational outcomes over which educational institutions exercise meaningful influence, as opposed to longer-term societal impacts—such as labor market integration or social mobility—which emerge over time and are shaped by multiple external factors beyond the direct control of individual organizations (Osborne et al., 2016; Talbot, 2009).

Public value creation is the outcome of a sequential process initiated by learner-centered teaching, which depends on the capacity of educators to design instruction around students' real needs and contexts. According to Hargreaves and Fullan (2012), when teaching is based on active, participatory, and personalized methodologies, students develop greater autonomy and long-term learning abilities. If the teaching process values students and responds to their needs, they will be more motivated and achieve better learning outcomes (Harris and Jones, 2018; OECD, 2016; Ryan and Deci, 2000). As argued by Moore (2013, 1995), public value results from activities that combine social relevance, political legitimacy, and operational capacity. Student-centered teaching aligns with these three dimensions, as it promotes inclusion, empowerment, and the development of life and work skills. Furthermore, consistent with Ryan and Deci (2000), authentic learning occurs when students are motivated, engaged, and recognized as active agents. These conditions emerge only in environments where educational processes are designed around the real needs of individuals, rather than abstract standards. Consequently, placing the student at the center is not only a sound pedagogical practice-it is also a key lever for creating public value. This logic reflects a broader evolution in public management, which shifts the focus from top-down service delivery to co-created value, where users are seen as partners in the design and implementation of public services (Alford and O'Flynn, 2009; Osborne, 2010). Student-centered teaching does not arise in isolation but is nurtured through a synergistic interplay of distributed leadership, robust knowledge-sharing mechanisms, and collaborative governance practices. This approach is supported by key contributions in public management literature, which highlight the importance of distributed leadership in promoting professional collaboration (Harris and Jones, 2018; Spillane, 2005), the role of knowledge systems in enabling inclusive decision-making (Schildkamp et al., 2013) and the ability of collaborative governance practices to link schools with community resources and policy networks (Bryson et al., 2014; Crosby et al., 2017). Building on these foundations, recent research (e.g. Bryson et al., 2015; Crosby et al., 2017; Sørensen et al., 2021) argues that public value emerges through the synergistic interplay of culture, governance, and leadership. In this perspective, organizational culture serves as an enabling condition that allows distributed leadership and open governance to translate into student-centered educational practices. This approach not only enhances learning outcomes but also contributes to the development of a school ecosystem that generates value for the broader community. From this perspective, student-centered teaching emerges as a product of systemic coherence: when leadership, organizational learning, and external collaboration converge, schools become adaptive and responsive ecosystems. Fullan (2018) and the OECD (2016) affirm that only through this convergence can institutions effectively address the diverse and often vulnerable profiles of adult learners. Ultimately, this theoretical approach redefines teaching as a dynamic, student-centered process—one that not only improves learning outcomes but actively contributes to the creation of public value within and beyond the school.

Collaborative culture represents an essential aspect of organizations, allowing for the creation of an inclusive and flexible environment (Fullan, 2018; Senge, 1990). It strengthens engagement by creating conditions in which all participants feel responsible and motivated to contribute to decision-making processes (Ansell and Gash, 2008; Klijn and Koppenjan, 2016). This is particularly important in complex contexts such as the educational system, where collaboration among schools, local communities, and policymakers requires not only structural mechanisms, but also a shared culture that promotes co-responsibility and long-term improvements (Sicilia et al., 2021).

Building on this foundation, numerous studies demonstrate that collaborative culture is the key driver for distributed leadership. When collaboration is embedded in organizational practice, it encourages shared responsibility and active participation, which in turn enhances the quality of decision-making and boosts motivation among staff members (Leithwood et al., 2008, p. 200; Spillane, 2006). Bezzina and Paletta (2022) further stress that meaningful and lasting school leadership emerges from communities grounded in reflection, teamwork, and a collective sense of purpose — all of which are essential features of a collaborative environment. Leadership is no longer the domain of a single figure at the top, but a shared endeavor shaped by relationships, interdependence, and collective purpose (Leithwood and Riehl, 2003; Benington and Moore, 2011). In the context of CPIAs, the role of leadership is even more delicate, as it operates in an environment characterized by greater vulnerability and more complex educational needs (Bolden, 2011; Harris, 2014; Spillane, 2006). Beyond organizational management, leadership actively shapes the collaborative culture within which professional development unfolds. It drives teacher engagement and fosters continuous learning (Spillane, 2005; Harris et al., 2018). As Fullan (2018) notes, school leaders serve as facilitators of professional learning—not just strategic managers—encouraging environments where educators grow together. Scholars such as Gronn (2002) and Spillane et al. (2001) emphasize that distributed leadership emerges as a response to complexity, dispersing influence across roles and teams. In education, this model strengthens innovation and responsiveness, encouraging co-creation of value and inclusive decision-making (Paletta, 2020; Robinson et al., 2008). Crucially, collaborative culture is not just a supportive backdrop—it actively shapes how leadership is distributed. Shared values, trust, and participation define the context for distributed leadership to be effective (Spillane, 2005; Harris and Jones, 2018). As Hargreaves and Fullan (2012) argue, schools grounded in collaboration enable leaders to extend influence authentically, leading to lasting transformation. Without this cultural foundation, distributed leadership risks becoming superficial, lacking the trust and mutual recognition required for meaningful cooperation (Schein, 2010). In CPIAs, establishing this link is vital: only through a collaborative foundation can leadership adapt to adult learners' diverse pathways and create responsive, equitable structures (Fullan, 2018).

Understanding an educational system requires more than examining teaching practices and student performance—it demands attention to the organizational structures that support and shape the entire learning experience. These structures include development and improvement plans, along with instruments for data collection, internal communication, and knowledge exchange (Watkins and Marsick, 1999). These aren't merely bureaucratic tools: they operate as organizational mechanisms that translate institutional goals into coordinated actions, fostering accountability and adaptability (Bryk et al., 2010; Fullan, 2001). When integrated into everyday operations, systems like data-informed planning and feedback loops enhance the quality of teaching and learning. Strategic use of evidence allows schools to personalize learning pathways and promote equity, particularly in diverse educational settings (Schildkamp et al., 2013). Collaborative culture is a determining factor in shaping organizational learning mechanisms. If we analyze how an organization functions, it becomes clear that its operations are oriented toward the achievement of a common goal. However, what guides behaviors and internal dynamics in practice is not only the shared objective, but also — and above all — the organizational culture (Amin, 2022; Marreiros et al., 2023; Schein, 2010). Consider, for example, that all decisions made by members of an organization are deeply influenced by the organization's own culture (Denhardt and Denhardt, 2000; Schein, 2010). Consequently, collaborative culture not only guides strategic choices within an organization (Amin, 2022) but also influences the capacity for adaptation and innovation of these systems (Marreiros et al., 2023). In the educational context, several empirical studies have examined how a school culture that fosters collaboration and continuous learning is associated with increased knowledge sharing and stronger community engagement (Senge, 1990). On the other hand, a more hierarchical and bureaucratic culture leads to rigid organizational systems that are less adaptable to community needs (Ansell and Torfing, 2021; Bouckaert et al., 2010). As emphasized by Senge (1990) and Watkins and Marsick (1999), learning organizations are built upon networks of trust-based relationships and environments of continuous feedback. These dynamics can only emerge when the organizational culture promotes interdependence, openness to dialogue, and responsible autonomy. In this framework, culture not only influences leadership but also determines the quality of learning systems, which are the mechanisms through which organizations manage information (Marsick and Watkins, 1999). Particularly in the CPIAs as education to vulnerable context, the effectiveness of organizational systems is essential for the development of personalized learning pathways, given the diverse backgrounds of students.

In contemporary public management, collaborative governance practices (CGPs) are a pivotal concept. It is a crucial aspect to consider as an organization no longer operates solely in pursuit of its own internal goals, but rather within an open and interconnected system, where external dynamics directly influence both decision-making and operational processes (Osborne, 2006; Rhodes, 1996). In a context characterized by uncertainty and constant evolution, organizations are driven to adopt a collaborative and adaptive approach, engaging with a network of actors to ensure effective governance and the creation of public value (Ansell and Gash, 2008; Bovaird, 2007). When we refer to CGPs, we are describing a set of relationships among public, private, and third-sector actors that enable the co-creation of context-specific educational solutions (Ansell and Gash, 2008; Emerson et al., 2012). As demonstrated by Page et al. (2015) and Brandsen et al. (2018), governance models that actively engage stakeholders enhance both democratic legitimacy and the effectiveness of public policies—especially in sensitive areas such as adult education and migrant inclusion. In the context of Sicilian CPIAs, the collaborative governance model is particularly significant, as it enables a more effective response to the complex educational, social, and cultural needs of learners from marginalized and migratory backgrounds (IOM, 2022; UNHCR, 2021). It is important to consider that, although the classroom is the primary environment where teachers spend most of their professional lives (Lortie, 1975), it is not their only workplace. Educational institutions do not function in isolation but rather operate as open systems; this dynamic requires interactions with the local community to enhance the quality of education (Harris et al., 2018). In CPIAs, developing territorial networks is essential for ensuring student inclusion and education, as well as increasing employment opportunities (Bryson et al., 2014). These collaborations improve schools' effectiveness in preparing students for both the workforce and social life (Fullan, 2018). This assumption is supported by public management literature, which highlights how a culture oriented toward collaboration fosters the ability to build effective and interconnected governance networks (Schein, 2010). According to Hargreaves and Fullan (2012), a school with a strong collaborative culture is better equipped to create and maintain partnerships with the local community, as it facilitates dialogue with local stakeholders (Silins and Mulford, 2004), adapts more flexibly to socio-economic changes, and has a greater capacity to engage external partners (Watkins and Marsick, 1999). In this perspective, Emerson et al. (2012) argue that collaborative governance practices emerges when there is a coherent alignment between shared values, organizational arrangements, and operational practices—elements they consider essential for enabling sustained and effective collaboration. Building upon this internal cultural foundation, collaborative governance practices emerge as a structural catalyst for inclusive participation and meaningful engagement. By integrating the voices of teachers, families, and local institutions (Silins and Mulford, 2004), the educational system becomes more attuned to community needs, facilitating social and economic impact beyond the provision of formal education (Osborne, 2006; Stoker, 2006, p. 200). Through well-established networks of collaboration with external stakeholders, educational centers can create spaces for dialogue in which innovative practices are developed, and inclusive decision-making processes are strengthened (Hargreaves and O'Connor, 2018).

The previous section presented a theoretical exploration aimed at constructing a conceptual framework for analyzing public value creation within educational contexts. Starting with the foundational insights of Moore (1995) and Stoker (2006), we delve into the dynamic nature of public value as a context-sensitive construct shaped by continuous organizational learning.

To test this framework, we formulate the following research hypotheses (Figure 1):

Our model, therefore, follows a sequential structure based on hypotheses derived from the literature. In the next section, we will specifically analyze the methodology used to examine the effect that exists between these variables.

The empirical investigation was conducted in Italy, within the network of CPIAs operating across the Sicilian region. These public institutions are specifically designed to provide adult education, including literacy programs, foundational academic instruction, and vocational training. They serve a highly diverse population, comprising migrants, individuals seeking professional requalification, and adults returning to formal education.

In the Italian context, CPIAs also play a relevant role in relation to the NEET phenomenon, although they are not institutions exclusively designed for this target group. Young people and adults who are neither in employment nor in education or training often experience fragmented educational trajectories, weak formal qualifications, limited basic skills, and difficulties in accessing labour market opportunities. CPIAs address several of these risk factors by offering second-chance education, literacy and language courses, basic and transversal competence development, and pathways aimed at re-entering formal education and training. Their contribution is therefore particularly relevant in vulnerable territories, where educational disadvantage, unemployment, migration, and social exclusion tend to reinforce one another. From a Public Value perspective, CPIAs may contribute to reducing the risk of long-term exclusion by strengthening individual capabilities, supporting employability, and connecting learners with local educational, social, and institutional networks.

Sicily offers a particularly meaningful context in which to examine the role of CPIAs. The region hosts one of the largest CPIA networks in Italy and faces significant social and demographic challenges, including the highest number of unaccompanied foreign minors nationwide (IOM, 2022).

As also shown in Table 1, Sicily is affected by a considerable unemployment problem, which engages the CPIAs in the region in a markedly different way compared to the rest of the country. In particular, the table below presents the percentage of students who were unemployed at the time of enrolment in CPIAs, the proportion of detained students, and the share of students with low educational attainment both at entry and in the 2024/2025 school year.

The table highlights higher levels of educational and social vulnerability among CPIA students in Sicily compared to both Southern Italy (excluding the islands) and the national average. In Sicily, 89.9% of students enter CPIAs with low educational attainment, compared to 82.2% in Southern Italy and 70.6% nationwide. The proportion of unemployed students (46.8%) is broadly in line with the national average (49.5%), as is the share of detained students (5.3 vs 3.7%).

These data underscore the strategic role of CPIAs in supporting social inclusion and employability, particularly in contexts characterised by educational fragility. The high proportion of learners with low educational attainment, especially when compared to the national figure, highlights the specific profile of the target population, which largely consists of individuals with interrupted or vulnerable educational pathways.

This context calls for a proactive and inclusive educational approach, grounded in internal collaboration and tailored to learners' needs. By placing students at the centre of the educational process and strengthening work-oriented competences, CPIAs can contribute to preventing unemployment and early school leaving. In this perspective, Sicilian CPIAs function as local educational hubs, linking individual learning needs with territorial resources and institutional networks, thereby promoting social inclusion and collective well-being.

For the data collection phase, a structured questionnaire was administered to schoolteachers across various CPIAs in Sicily (see  Appendix, Tables A1–A6). A total of 711 teachers were invited to participate in the survey, corresponding to the total number of teachers employed across the various CPIAs (see Table 2). Of these, 317 valid responses were retained for analysis, resulting in an effective response rate of 44.6%. However, it should be noted that the number of respondents varies considerably between CPIA centres, ranging from around 70% in the Agrigento CPIA to 13% in the Messina CIPIA. The fact that some CPIA centres are more represented than others may influence the results of the statistical model, and this should be taken into account when interpreting the results.

Table 3 reflect the characteristics of the surveyed teachers with the overall teacher population in Sicilian CPIAs. Overall, the sample closely reflects the reference population across the main demographic and professional variables. The age distribution is highly consistent with the population structure, with only marginal differences across categories. In particular, teachers aged 55 and over represent the largest group in both the sample (48%) and the reference population (47%), followed by those aged 45–54 (34 vs 36%).

Regarding years of service in the same CPIA, the sample shows a slight overrepresentation of teachers with ≤1 year of service (18 vs 9%) and a modest underrepresentation of those with more than five years (51 vs 57%). However, the remaining categories are closely comparable. Finally, the distribution by contract type is nearly identical between the sample and the reference population (91 vs 92% permanent contracts; 9 vs 8% fixed-term contracts).

Overall, these comparisons indicate a satisfactory level of representativeness of the sample with respect to the underlying population.

The aim of this study is to investigate how the creation of public value can be theoretically conceptualized and empirically detected within the school system, with reference to Adult Education Provincial Centres (CPIAs). While public value theory has become a central paradigm in public management scholarship, its application to educational organizations—and especially to adult education institutions—remains conceptually underdeveloped and methodologically fragmented. This study seeks to contribute to this gap by proposing a theoretically grounded operationalization of public value tailored to the institutional mission of CPIAs.

The measurement scale (see Table 4) was developed through a theory-driven process rooted in the public value tradition (Benington, 2011; Bozeman, 2007; Meynhardt, 2009; Moore, 1995), which conceptualizes public value as the capacity of public organizations to generate collectively desirable outcomes, enhance social legitimacy, and strengthen societal capabilities. Within this framework, public value is not limited to efficiency or performance outputs, but encompasses broader dimensions related to equity, inclusion, well-being, and collaborative governance.

To contextualize these dimensions within the educational domain, the analytical framework proposed by Kools et al. (2020) was integrated as a complementary lens. Their systemic perspective on educational change and organizational learning provides a bridge between public management theory and the internal processes through which schools generate socially relevant outcomes. By combining these approaches, the study conceptualizes CPIAs as institutional actors embedded in territorial ecosystems, whose value creation processes operate simultaneously at the individual (capability development), social (inclusion and cohesion), economic (employability), and relational (network building) levels.

The resulting instrument thus reflects a multidimensional understanding of public value as an emergent property of educational action. It positions CPIAs not merely as service providers, but as public organizations contributing to the expansion of human capabilities, the reduction of structural inequalities, and the strengthening of local governance networks. In this sense, the scale represents an attempt to translate the normative and strategic foundations of public value theory into an empirically testable construct within the field of adult education.

In carrying out our analysis, we organized the research process into two main phases, following the methodological frameworks (Churchill, 1979; Steenkamp and Van Trijp, 1991). In the first phase, we defined the Measurement Model by conducting Exploratory Factor Analysis (EFA) using SPSS, followed by Confirmatory Factor Analysis (CFA) (see “ Supplementary Material: Descriptive Statistics” file for further details). In the second phase, we applied Structural Equation Modeling (SEM) in LISREL to validate the proposed model. The subsequent sections provide a comprehensive description of each phase.

To define our measurement model, it is crucial to assess the reliability of the constructs. Initially, a Cronbach's α test was performed to ensure that the selected items accurately reflected the hypothesized constructs, with all group scores meeting the established thresholds (Churchill, 1979; Hulland, 1999; Steenkamp and Van Trijp, 1991). Following, Exploratory Factor Analysis (EFA) was employed to determine the minimum number of factors that could effectively summarize the original set of items while maintaining the condition m > n (Hulland, 1999). EFA is a multivariate statistical technique that is frequently employed prior to Structural Equation Modeling (SEM) to uncover the latent structure underlying the items in a dataset. Specifically, EFA is used to identify the possible existence of unobservable constructs, examine correlations among individual items, and reduce data dimensionality by eliminating cross-loadings. Given that EFA may result in some loss of information, it is essential to maximize the amount of variance retained to ensure an accurate representation of the underlying constructs. As shown in  Appendix, the questionnaire initially consisted of a total of 58 items. The EFA process identified six factors from the original set, retaining 39 items and eliminating 19. This objective was achieved by employing Maximum Likelihood Estimation with VARIMAX rotation, which explained 69% of the total variance. However, we mention that EFA results indicate that item Coll_Cul_3 loads both on Leadership construct (0.510) and Collaborative Culture (0.459). This cross-loading can be justified by the existing literature (e.g. Demir, 2008) as the construct of Leadership is typically conceptualized within an organizational context, particularly when examining Collaborative Culture. This finding should be validated through a Confirmatory Factor Analysis (CFA) evaluating whether these measures are empirically distinct from other constructs, using LISREL. The CFA results were satisfactory and met the accepted threshold values for Average Variance Extracted (AVE) and Construct Reliability, as reported in Table 4. Specifically, the Leadership construct exhibits satisfactory reliability and convergent validity, with a Construct Reliability of 0.917 and an AVE of 0.592. Similarly, Collaborative Culture demonstrates strong measurement properties, with a CR of 0.891 and an AVE of 0.732. All reported values exceed the commonly accepted threshold levels (CR > 0.70; AVE >0.50), indicating adequate construct reliability and convergent validity.

The final step of our analysis involves evaluating the structural equation model (SEM) to assess the significance of the hypothesized cause-and-effect relationships and to evaluate the overall explanatory power of our research model.

Initially, our model showed a good fit with the data, as all relevant indicators met the required criteria for SEM analysis, yielding excellent results that significantly exceeded commonly accepted thresholds (Churchill, 1979; Steenkamp and Van Trijp, 1991). Specifically, the following fit indices were obtained: χ² (degrees of freedom) = 1751.131(692); Global Fit Index (GFI) = 0.757; Adjusted Goodness of Fit Index (AGFI) = 0.727; Comparative Fit Index (CFI) = 0.984; Root Mean Square Error of Approximation (RMSEA) = 0.0765, and Standardized RMR = 0.0306. Table 5 presents the results of the Structural Equation Modeling (SEM) analysis.

The results overall address Collaborative Culture as a positive and significant influencing factor for Distributed Leadership (H1. γ = 0.850, p < 0.01), Knowledge sharing and learning system (H2, γ = 0.706, p < 0.01) and Collaborative Governance practices (H3. γ = 0.762, p < 0.01). The Hypothesized model results also show that overall, there is a positive and significant relationship between Distributed leadership, Knowledge sharing and learning system, Collaborative Governance and Student-Centered Teaching Processes (H4.I, β = 0.304, p < 0.01; H4.II, β = 0.303, p < 0.01; H4.III β = 0.354 p < 0.01). In conclusion, the hypothesized model validates the positive and significant relationship between Student-Centered Teaching Processes and Creation of Public Value (H5: β = 0.556, p < 0.01).

The ultimate objective of the model is to identify the antecedent conditions leading to the Creation of Public Value. To further validate the hypothesized relationships, coherently with the literature (e.g. Geyskens and Steenkamp, 2000) we operationalized an additional performance variable as an alternative to Creation of Public Value. The underlying rationale is that this variable captures the extent to which CPIAs contribute to students' personal, social, and professional development. However, this construct is measured subjectively, as it relies on teachers' perceptions rather than on objective and quantifiable outcomes.

To address this limitation, we operationalized an objective variable using data from “Scuola in Chiaro” ^[1] the national database of Italian CPIA performance. Specifically, we considered the following indicators: (1) digital competence, (2) multilingual competence, and (3) personal, social, and learning-to-learn competence, as defined by the European competence framework.

The European Key Competences Framework for Lifelong Learning ^[2] is an objective and normatively grounded benchmark. The framework identifies digital competence, multilingual competence, and personal, social, and learning-to-learn competence as essential capabilities for active citizenship, social inclusion, and employability. These dimensions reflect collectively endorsed societal goals and therefore represent institutionalized expressions of public value within the European policy space. By measuring the extent to which CPIAs enable students to achieve at least a basic level in these competences, we operationalize Public Value not merely as perceived educational quality, but as the enhancement of citizens' capabilities that generate broader social and economic benefits.

The introduction of the “objective” PV variable was intended to complement the perceptual measure through official administrative data at the CPIA level. Although the questionnaire was administered to teachers, respondents were explicitly asked to evaluate organizational characteristics and practices of their respective CPIA rather than individual experiences or personal outcomes. Accordingly, the level of analysis in the model has consistently remained the CPIA in both models, as teachers' responses were conceptualized as organizational assessments referring to the institution to which they belong. The additional “objective” PV variable does not shift the analytical focus to the teacher level; rather, it further strengthens the CPIA-level perspective by complementing perceptual evaluations with institution-level evidence.

As stated above, objective Measures of Public Value is defined by 3 items: (1) digital competence, (2) multilingual competence, and (3) personal, social, and learning-to-learn competence. Each of these dimensions is reported as a percentage of students distributed across four proficiency levels: (1) initial, (2) basic, (3) intermediate, and (4) advanced. The objective PV items were operationalized as the percentage of students achieving at least the “basic” level of competence, under the assumption that the “initial” level does not yet reflect a sufficiently consolidated educational outcome. Consequently, for each competence dimension, the percentages of students reaching the basic, intermediate, and advanced levels were aggregated to obtain a single institutional performance indicator. In order to ensure comparability with the other latent variables included in the SEM framework and to facilitate integration into the LISREL covariance structure, these percentages were subsequently transformed into a 4-point Likert-type scale through proportional rescaling. The resulting institutional-level score was then assigned to all respondents belonging to the corresponding CPIA. This procedure allowed the inclusion of objective organizational performance indicators within the individual-level dataset, thereby integrating subjective perceptions with externally validated educational outcome measures.

Once computed, the performance score of each CPIA was assigned to all respondents affiliated with that institution. In this way, we were able to test the hypothesized model using this alternative variable of CPIA Public Value, within our cross sectional data. First, the model demonstrates again, statistically significant fit to the data, with χ² (degrees of freedom) = 1441.141 (584); Global Fit Index (GFI) = 0.798; Adjusted Goodness of Fit Index (AGFI) = 0.769; Comparative Fit Index (CFI) = 0.986; Root Mean Square Error of Approximation (RMSEA) = 0.0682; and Standardized Root Mean Square Residual (RMR) = 0.0419. All these values fall well within the thresholds commonly accepted in the literature for the SEM model to be significant and acceptable.

With respect to the structural effects, the results indicate that Collaborative Culture is a positive and statistically significant antecedent of Distributed Leadership (H1: γ = 0.721, p < 0.01), Knowledge Sharing and Learning Systems (H2: γ = 0.623, p < 0.01), and Collaborative Governance practices (H3: γ = 0.606, p < 0.01). The results of the hypothesized model further show that Distributed Leadership, Knowledge Sharing and Learning Systems, and Collaborative Governance practices are all positively and significantly associated with Student-Centered Teaching Processes (H4.I: β = 0.321, p < 0.01; H4.II: β = 0.340, p < 0.01; H4.III: β = 0.450, p < 0.01).

Finally, the hypothesized model confirms a positive and statistically significant relationship between Student-Centered Teaching Processes and CPIA Public Value (H5: β = 0.134, p < 0.05). Results are summarized in Table 6 below.

Overall, these findings provide robust empirical support for the hypothesized model, demonstrating that the identified organizational antecedents indirectly contribute to CPIA Public Value through student-centered teaching processes, thereby strengthening the validity of the proposed framework beyond subjective measures of public value.

In line with the concept of Public Value Management (Stoker, 2006), the study confirms that CPIAs do not merely function as knowledge transmitters but as dynamic organizations that respond to societal challenges through shared governance, and inclusive educational strategies (Kools and Stoll, 2016; Moore, 1995). By analyzing teachers' perceptions across multiple provinces, the study explores how these institutions cultivate collaborative and adaptive learning environments that not only deliver formal education but also foster institutional effectiveness, social integration, and community engagement (Fullan, 2006; Kools and Stoll, 2016). SEM results show that collaboration positively influences distributed leadership, knowledge-sharing systems and collaborative governance practices. These findings reinforce previous studies that highlight how shared leadership and participatory decision-making improve institutional performance and responsiveness to community needs (Leithwood et al., 2004; Spillane, 2006). Furthermore, the study demonstrates that student-centered teaching processes contribute significantly to the creation of public value. CPIAs that emphasize individualized learning paths and active student engagement show a greater ability to meet community needs, promote social integration, and improve student employability. The study also sheds light on how collaborative governance practices within CPIAs contributes to better institutional performance and stronger social impact. CPIAs that encourage participatory governance, where decision-making is shared between teachers, students, and local institutions, appear to be more effective in addressing the educational and social needs of their communities. This supports the Public Value Management (PVM) framework, which argues that public institutions create value not only through the direct provision of services, but also through engagement with stakeholders and the co-creation of solutions (Moore, 1995; Stoker, 2006). Given that Sicily has a high number of unaccompanied minors and a diverse migrant population, this collaborative governance practices model is particularly relevant, as it allows CPIAs to adapt their educational strategies to the specific challenges of their students (UNHCR, 2021).

This study contributes to understanding how Public Value Management (Bryson et al., 2014; Meynhardt, 2009; Stoker, 2006) is relevant in socially vulnerable contexts. First, we contextualised the theory of public value in Provincial Centres for Adult Education (CPIAs). The creation of public value emerges as a multifaceted concept that extends beyond traditional educational outcomes. CPIAs generate public value by equipping students with essential basic skills, providing them with the core competences necessary for active citizenship and lifelong learning. However, their role goes beyond literacy and numeracy, as they also strive to promote transversal and socio-emotional skills, which improve adaptability, communication and emotional intelligence, all of which are fundamental for personal and professional growth. A defining characteristic of the public value of CPIAs is their commitment to inclusive education. By recognising and addressing the diverse social and cultural backgrounds of students, these institutions promote equal opportunities and social cohesion, ensuring that education is accessible to all. This inclusiveness contributes directly to broader social benefits, such as reduced inequalities and greater integration into the community (Alford and O'Flynn, 2009; Bryson et al., 2014). Moreover, public value management in CPIAs play a key role in aligning education with labour market demands. By developing key competencies that respond to evolving labour market needs, they improve employability and economic participation, fostering individual empowerment and social mobility (Benington and Moore, 2011; Talbot, 2009). Their commitment to building local networks to support adult skills development reinforces their role as catalysts for community growth, strengthening collaboration between education, employment and social services (Osborne et al., 2016; Page et al., 2015).

Second, in this paper we have shown that a collaborative culture creates public value indirectly by shaping governance, leadership and teamwork (Bryson et al., 2015). In educational institutions, a collaborative culture serves as a catalyst for public value, not by directly producing measurable outputs, but by shaping the way governance, leadership, and internal collaboration function.

From a practical perspective, these findings do not imply direct evidence of improved student well-being or social outcomes. Instead, they highlight organizational and managerial levers—such as collaborative culture, distributed leadership, knowledge sharing, and student-centered teaching—that public managers and policymakers can strengthen to enhance the value-generating capacity of adult education institutions, particularly in contexts of social vulnerability.

When adult education centers embrace collaborative culture, they create an ecosystem where educators, staff, and external stakeholders work together toward shared educational and social goals.

At the heart of this transformation is a culture of shared responsibility. When staff actively participate in decision-making and professional development, they become more engaged in the institution's mission, fostering innovation and continuous improvement. System figures, who emerge through peer recognition, play a vital role in facilitating this process by mentoring colleagues, promoting knowledge exchange, and ensuring a coordinated, trust-based environment. Their leadership is not about hierarchy but about empowering others to take on responsibilities, ensuring that collaboration becomes a structural, rather than incidental, feature of the institution (Crosby et al., 2017).

Collaboration extends beyond distributed leadership—it is embedded in how schools operate daily. Systematic dialogue, evidence-based decision-making, and continuous learning enable teachers to improve their practices dynamically. The use of student data, peer feedback, and shared reflections on both successes and failures fosters an adaptable educational system that responds effectively to student needs. Additionally, communication strategies and shared platforms ensure that knowledge and best practices circulate efficiently across teams, locations, and disciplines, reinforcing an interconnected learning environment (Meynhardt, 2009; Stoker, 2006).

Beyond internal collaboration, external partnerships amplify public value creation. Schools and educational centers are not isolated entities but integrated into broader community networks (Paletta, 2012; Paletta et al., 2009). By working with local services, research institutions, training centers, and social organizations, they extend their impact beyond the classroom, aligning education with real-world needs. Engaging with social welfare initiatives and combating educational poverty further highlights the role of schools as agents of social inclusion and empowerment (Bryson et al., 2014).

Ultimately, a collaborative culture transforms education into a driver of systemic change. By fostering shared leadership, knowledge exchange, and collective problem-solving, schools enhance both their internal effectiveness and their broader social mission. This dynamic process indirectly creates public value, as a well-governed, inclusive, and forward-thinking educational institution contributes to social cohesion, workforce readiness, and lifelong learning—foundations of a resilient and equitable society.

While this research provides valuable insights into the organizational dynamics of CPIAs, several limitations should be acknowledged:

1. Geographical limitations. One of the main limitations of this research is its focus on CPIAs in Sicily. While this region provides a valuable multiple case study, especially given its high number of MSNA and diverse student population, the findings may not fully reflect the reality of adult education centers in other Italian regions or international contexts. Socioeconomic conditions, institutional policies, and local governance structures can vary significantly, potentially affecting how collaboration and leadership function in different educational settings (UNHCR, 2021).

2. Cross-Sectional Nature of the Study. This study captures data at a single point in time, which means it provides a snapshot of the relationships between collaboration, leadership, and public value creation, but it cannot track how these relationships evolve over time. The effectiveness of collaborative practices and leadership structures may change as schools adapt to new policies, technologies, or student needs.

3. Reliance on Self-Reported Data. The study relies on questionnaires completed by teachers, which, while valuable, come with inherent biases. Respondents might provide socially desirable answers rather than fully objective assessments of their institutions. Additionally, personal experiences and perceptions can vary widely, making it difficult to obtain a completely uniform picture of collaboration and leadership across CPIAs. Furthermore, the elementary unit of analysis is the individual teacher rather than the school as an organizational entity. As a result, the findings reflect trends at the provincial education system level rather than providing specific insights into each individual CPIAs.

4. A further limitation of the study is the absence of direct evidence on students' perceived well-being, social outcomes, or counterfactual comparisons with contexts where CPIAs are not present. Future research should adopt multi-actor and longitudinal designs, integrating teachers' assessments with students' perspectives and comparative approaches, in order to better capture the full range of public value outcomes generated by adult education institutions.

By addressing these gaps, future research can further deepen our understanding of how CPIAs—and similar institutions—can maximize their potential as engines of lifelong learning, social inclusion, and workforce development (OECD, 2016).

This study provides empirical evidence to support the role of CPIAs in Sicily as institutions to generate public value through student-centered teaching approaches and collaboration. By analyzing teachers' perceptions across provinces, the research highlights the essential role of CPIAs in addressing the educational needs of diverse and often marginalized populations, including young workers, the unemployed, migrants, and individuals in socially vulnerable situations.

The findings confirm that a strong collaborative organizational culture significantly enhances the capacity of CPIAs to generate public value. CPIAs that prioritize inclusive and adaptive teaching methods show a greater ability to meet community needs, support social integration, and enhance their students' employability prospects (Bryk et al., 2010; Darling-Hammond, 2010). The hypothesized model demonstrates that distributed leadership, knowledge sharing, and collaborative governance practices significantly contribute to student-centered teaching processes, which in turn positively influences public value creation.

From a theoretical standpoint, this study contributes to a deeper understanding of the concept of public value within the educational domain, offering an empirical perspective that highlights its concrete generation within complex organizational settings. Specifically, the findings demonstrate that collaborative culture acts as an enabling condition for distributed leadership, organizational learning, and participatory governance—elements that prove essential for the adoption of student-centered teaching practices. In doing so, the study extends the Public Value Management framework by integrating it with theories of educational leadership and organizational learning, and proposes an innovative interpretation of co-production, understood not merely as external collaboration but as a process rooted in organizational culture.

Finally, the theoretical contribution is situated within a critical and context-sensitive perspective, emphasizing that the generation of public value is particularly significant in vulnerable contexts, where educational institutions are called upon to respond to complex and evolving needs. In such settings, the capacity to activate collaborative dynamics emerges not only as a managerial lever but also as an ethical and social responsibility.

The concept of collaborative culture in this study is measured using the scale derived from the TALIS 2018 Teacher Questionnaire (OECD, 2018), which assesses the extent to which teachers engage in professional collaboration, knowledge-sharing, and collective decision-making within their institutions. The scale has been further contextualized to align with the specific organizational structures of Italian CPIAs, emphasizing interdisciplinary collaboration, teacher networks, and cooperative governance models that enhance institutional effectiveness.

The questionnaire reflects the following dimensions: staff participation, mutual support among teachers, shared innovation, and stakeholder involvement.

The scale on distributed leadership was developed by integrating the public management literature on governance and organizational coordination with the educational research on distributed leadership. In particular, studies on shared governance and collective leadership in public institutions are combined with the conceptualization of distributed leadership proposed by Spillane (2006) and Leithwood et al. (2009). The SLO framework (Kools et al., 2020) informed the operational translation of these constructs within the specific organizational setting of CPIAs.

The scale captures the institutionalization of knowledge-sharing and learning processes as core governance mechanisms within schools. In line with the public management literature on performance management and evidence-based governance, public organizations are conceived as adaptive systems that rely on structured evaluation practices, systematic data use, and collective reflection to improve outcomes over time.

The items reflect formalized self-evaluation procedures (e.g. PTOF and Improvement Plans), the routine collection and analysis of performance evidence, the professional development of teachers in data literacy, and the existence of structured communication and knowledge-sharing infrastructures. Together, these elements operationalize the school's capacity to function as a learning-oriented and performance-driven public institution, capable of continuously aligning its strategies and practices with intended objectives.

This dimension captures the extent to which schools operate as networked public organizations within a broader governance system. It is grounded in the public management literature on collaborative governance, network management, and public value co-creation (Ansell and Gash, 2008; Bryson et al., 2015; Osborne, 2010; Paletta, 2012), which emphasizes that public value emerges through structured inter-organizational collaboration.

The scale operationalizes this perspective by measuring environmental scanning, formalized partnerships, inter-institutional cooperation, and structured stakeholder engagement. In the CPIA context, particular attention is given to network agreements, collaboration with local service providers and universities, and coordinated actions aimed at addressing educational disadvantage and social vulnerability.

The dimension of student-centred learning processes in this study is adapted from the TALIS Teacher Questionnaire (OECD, 2018). TALIS data highlights the importance of using student feedback, adapting teaching strategies and fostering collaboration among teachers to improve student engagement and learning outcomes. This dimension has been contextualised for the CPIAs system, incorporating vocational guidance, personalised learning arrangements, abandonment prevention mechanisms and targeted interventions for marginalised students.

This study conceptualizes public value creation within education by grounding it in the public value and public management literature (Moore, 1995; Bozeman, 2007; Benington and Moore, 2011; Meynhardt, 2009; Meynhardt et al., 2024). Within this perspective, public value is understood as a multidimensional construct encompassing the generation of socially desirable outcomes, the enhancement of institutional legitimacy, the development of individual capabilities, and the strengthening of social cohesion.

Applied to the CPIA context, this conceptualization highlights the role of adult education institutions in promoting social inclusion, expanding equitable access to learning opportunities, supporting employability and skill development, and contributing to local governance networks. Public value is therefore interpreted not only as strategic alignment or stakeholder mobilization, but as the measurable contribution of CPIAs to individual, social, and economic development within vulnerable territories.