Entrepreneurial intentions are made, not born: the active role of institutional, educational and economic support

The entrepreneurial landscape has undergone a profound transformation in the post-pandemic era, compelling scholars to reconsider the fundamental mechanisms that drive individuals, particularly young people, toward venture creation. This renewed urgency stems not merely from academic curiosity but from a pressing societal need: as traditional career paths become increasingly uncertain, entrepreneurship emerges as both an economic imperative and a personal salvation for many (Fikri and Newman, 2024). Yet despite decades of research, a paradox persists: while we understand that entrepreneurship is a deliberate, planned behaviour (Guzmán-Alfonso and Guzmán Cuevas, 2012), we continue to treat the environmental forces shaping these plans as peripheral rather than foundational. The dominant theoretical lens through which scholars examine entrepreneurial intentions, Ajzen’s (1991) Theory of Planned Behavior (TPB), has yielded invaluable insights into how personal motivators such as attitude, self-efficacy, and risk propensity drive entrepreneurial action. However, this framework has inadvertently created a blind spot in our understanding. By positioning contextual factors, institutional support, training infrastructure, economic conditions, as mere moderators that amplify or dampen pre-existing motivations (Prabowo et al., 2022; Nițu-Antonie et al., 2023), we may be missing a more fundamental truth: that these environmental elements do not simply influence entrepreneurial intentions but actively create them. This oversight carries significant policy implications. Consider the current Italian context, where recent Eurostat (2025) data reveal a compelling yet troubling narrative: business registrations have surged by 3.3% year-over-year, signalling renewed entrepreneurial vigour, yet bankruptcy declarations have simultaneously risen by 19%. This juxtaposition suggests that while more individuals are answering the entrepreneurial call, they are doing so in an ecosystem that may be inadequately preparing them for success. Such patterns underscore the critical need to understand not just whether people intend to start businesses, but how their surrounding environment shapes the very psychological foundations upon which these intentions rest.

Within this broader landscape, university students occupy a particularly pivotal position. They stand at the threshold of career formation, uniquely positioned to be shaped by their institutional environment yet sufficiently unencumbered to pursue entrepreneurial paths (Liu et al., 2025; Bardales-Cárdenas et al., 2024). The university context offers a natural laboratory for understanding how systematic interventions, from formal education to mentorship programs, can cultivate entrepreneurial mindsets. Recent scholarship has increasingly recognized that student entrepreneurship represents more than a subset of general entrepreneurship research; it constitutes a distinct phenomenon requiring specialized theoretical frameworks that account for the formative power of the university ecosystem (Villegas-Mateos and Amann, 2025; Henry et al., 2024). The evolution of this literature reveals a progressive sophistication in understanding how universities shape entrepreneurial development. Fischer et al. (2019) demonstrate that well-designed incubation programs do more than provide resources, they fundamentally alter students’ beliefs about their own capabilities. Gonzalez Tamayo et al. (2024) extend this insight, showing that institutional support operates through complex mediating pathways, particularly self-efficacy and role model exposure. Yet even these advances treat contextual factors in isolation, examining how individual environmental elements influence specific psychological outcomes rather than considering how the ecosystem as a whole might function as an integrated, formative force.

This fragmented approach becomes particularly problematic when we consider how universities have embraced what Bardales-Cárdenas et al. (2024) term their “third mission”, socio-economic development through entrepreneurship. As institutions increasingly position entrepreneurship education as core to their value proposition, they create comprehensive ecosystems designed to systematically intervene in students’ developmental trajectories. The literature suggests these interventions address multiple needs simultaneously: skill acquisition, network formation, confidence building, and opportunity recognition (Liu et al., 2025; Villegas-Mateos and Amann, 2025). Yet our theoretical models continue to parse these elements rather than examining their synergistic effects.

Our study addresses this theoretical gap by proposing a fundamental reconceptualization: that the entrepreneurial ecosystem does not merely moderate the relationship between personal characteristics and intentions but serves as a formative antecedent that actively constructs the psychological foundations of entrepreneurship. We introduce Entrepreneurial Ecosystem Support (EES) as a higher-order construct that captures the integrated influence of institutional support, training infrastructure, and economic environment. This perspective shifts the theoretical conversation from asking “how do environmental factors influence entrepreneurial intentions?” to “how do environmental factors create the very capacity for entrepreneurial thinking?” This reconceptualization finds theoretical grounding in both Shapero’s Entrepreneurial Event Model (Shapero and Sokol, 1982), which emphasizes how environmental “displacements” trigger entrepreneurial consideration, and an extended interpretation of TPB that recognizes how external forces shape the formation of attitudes and perceived behavioural control (Urban and Kujinga, 2017). By synthesizing these perspectives, we argue that exposure to supportive institutions, quality training, and favourable economic conditions does more than facilitate entrepreneurship, it fundamentally alters individuals’ self-conceptions, risk tolerances, and career aspirations.

To empirically test this theoretical proposition, we employ a sophisticated methodological approach using Partial Least Squares Structural Equation Modelling (PLS-SEM) with the two-stage disjoint procedure recommended by Becker et al. (2019). Drawing on data from 492 participants at Gabriele d’Annunzio University who engaged with the Ingenium project’s entrepreneurship initiative (Work Package 8), we examine how EES influences entrepreneurial intentions both directly and through its effects on self-efficacy, entrepreneurial attitude, and willingness to take risks. This approach allows us to disentangle the complex pathways through which environmental factors shape entrepreneurial development while maintaining the conceptual integrity of ecosystem thinking. Our findings promise to advance entrepreneurship theory by demonstrating that intentions are not merely influenced by context but are fundamentally constructed by it. For policymakers and university administrators, this insight suggests that fostering entrepreneurship requires more than providing resources or removing barriers, it demands creating integrated ecosystems capable of reshaping how individuals conceive of themselves and their possibilities.

Entrepreneurial intentions have emerged as a critical focus in entrepreneurship research primarily because they serve as a robust predictor of subsequent new venture creation. Indeed, the decision to launch a business is not a reflexive act but rather a conscious, deliberative process influenced by how individuals perceive and interpret situational cues (Krueger et al., 2000). As such, intention-based models offer researchers a theoretically grounded approach to understand why and how individuals transition from merely recognizing entrepreneurial opportunities to actively pursuing them. Two theoretical pillars continue to structure this field. Ajzen’s (1991) TPB explains EI through attitudes, subjective norms and perceived behavioural control, whereas Shapero’s Entrepreneurial Event Model (Shapero and Sokol, 1982) emphasises perceived desirability, feasibility and a propensity to act. Despite their different vocabularies, the two frameworks converge on the idea that intention is formed when an opportunity is evaluated as both attractive and personally feasible. Meta analytic evidence confirms their predictive power across diverse settings but also reveals considerable residual variance that demands contextual explanation (Kautonen et al., 2015). Moreover, examining intentions provides researchers with insights into the antecedents of entrepreneurial behaviour. By measuring constructs such as entrepreneurial attitude, social norms, and perceived feasibility, scholars can capture much of the variance that situational or demographic variables alone fail to explain (van Gelderen et al., 2008). Another valuable dimension of intention-based models is their ability to incorporate exogenous factors such as prior entrepreneurial exposure and educational interventions. For instance, exposure to role models can bolster perceived desirability, whereas entrepreneurship education has been shown to enhance perceived feasibility through the development of relevant skills and knowledge (Zhang et al., 2014; Liñán, 2008). Recent research also suggests that compulsory or elective entrepreneurship training may shape its impact on entrepreneurial intentions, highlighting the nuanced ways in which institutional contexts influence motivational processes (Karimi et al., 2016; Sánchez, 2011). Furthermore, understanding intentions helps elucidate the cognitive and affective mechanisms involved in the entrepreneurial process.

University students have become the focal population for testing and extending these models, because they are simultaneously forming career aspirations and are often exposed to systematic entrepreneurship training. Within this cohort, research shows that classic TPB variables interact with a wider set of dispositional and affective drivers. Risk taking propensity, innovativeness, proactiveness and an internal locus of control all raise the likelihood that students will form entrepreneurial intentions (Zollo et al., 2017). Entrepreneurial self-efficacy (SE), the belief in one’s ability to perform entrepreneurial tasks, emerges as a particularly strong predictor (Chien Chi et al., 2020; Litzky et al., 2020). Moreover, emotional competence amplifies this mechanism by reinforcing self-efficacy and persistence in the face of uncertain outcomes (Chien Chi et al., 2020). These findings extend TPB by highlighting the centrality of self-regulatory and emotional resources at a life stage when students are still experimenting with professional identities.

A second research stream has documented how the university environment and the broader ecosystem shape those personal motivators. Formal entrepreneurship education, whether embedded in curricula or delivered through extracurricular projects, consistently enhances entrepreneurial attitudes and intentions by providing skills, role model exposure and opportunity recognition tools (Egerová et al., 2017). Perceptions of institutional support inside the university, incubators, mentoring, seed funding and networking events, further reinforce positive attitudes (Zollo et al., 2017). Outside the campus, family business background, peer endorsement and informal collaborative networks supply resources and legitimacy that bolster students’ confidence to act entrepreneurially (Scuotto and Morellato, 2013). Yet these contextual effects are far from uniform: cross cultural studies show that the salience of self-efficacy and entrepreneurial capital is stronger in individualistic societies than in collectivist ones, and that the relevance of traits such as risk taking depends on the quality of local institutions (Litzky et al., 2020). Crucially, most empirical work analyses context in a fragmented fashion. Gonzalez Tamayo et al. (2024), for example, demonstrate that perceived institutional support affects EI only indirectly through self-efficacy and parental role models, but their model omits economic and training considerations. A recent systematic review by Gallegos et al. (2025) confirms that contextual variables are usually introduced as single proxies or as moderators, rather than as integrated antecedents that actively generate personal motivators. This piecemeal approach limits our ability to capture the synergistic effects that arise when institutional, educational and economic subsystems interact. The present study addresses this gap by conceptualising EES as a second order formative construct that aggregates institutional, training and economic support. By positioning EES as an antecedent, rather than a moderator, of attitude, self-efficacy and risk propensity, we test the proposition that the ecosystem itself creates the psychological foundations of student entrepreneurship, offering a more comprehensive explanation of why some students progress from opportunity recognition to actionable entrepreneurial intention.

The fragmentation problem identified in entrepreneurial intentions research finds its most compelling solution in the emergence of the entrepreneurial university, a paradigm shift that transforms higher education institutions from knowledge repositories into active architects of entrepreneurial mindsets (Syed and Spicer, 2025). Etzkowitz’s (2003) Triple Helix model revolutionized this understanding by positioning universities as co-equal partners with industry and government, not merely in knowledge transfer but in the fundamental construction of entrepreneurial capacity. What distinguishes entrepreneurial universities is their seamless integration of education, research, and socio-economic development into a unified ecosystem (Syed and Spicer, 2025). Unlike traditional institutions where entrepreneurship support exists as peripheral add-ons, these universities embed entrepreneurial thinking into their institutional DNA. Technology transfer offices, incubators, and industry partnerships become integral expressions of the academic mission, creating what we term an “ecosystem field effect”, where multiple support dimensions generate emergent properties exceeding their individual contributions. This integration becomes particularly evident in University-Based Entrepreneurial Ecosystems (UBEEs), which Villegas-Mateos and Amann (2025) characterize as interconnected systems developing ventures within university frameworks. Yet this definition understates their transformative power. UBEEs represent dynamic environments where formal structures (incubators, accelerators) and informal elements (cultural norms, peer networks) interact to fundamentally reshape participants’ self-conceptions and aspirations. When students simultaneously engage with entrepreneurship courses, university incubators, and venture capitalists through campus events, these experiences create multiplicative rather than additive effects, transforming core beliefs about what is possible.

Despite these advances, a critical gap persists. While the entrepreneurial university literature masterfully documents ecosystem components, it insufficiently addresses the psychological mechanisms translating ecosystem exposure into enhanced self-efficacy, transformed attitudes, and increased risk tolerance. This is precisely where our study contributes, bridging macro-level ecosystem theory with micro-level psychological change through the EES construct. By operationalizing EES as a formative, higher-order construct integrating institutional support, training infrastructure, and economic environment, we capture how entrepreneurial universities actually function, not as collections of discrete resources but as coherent environments systematically reconstructing entrepreneurial psychology. This reconceptualization suggests that effective entrepreneurial universities must orchestrate all elements into synergistic wholes, making them both the solution to theoretical fragmentation and ideal laboratories for understanding how ecosystems create, rather than merely influence, entrepreneurial intentions.

A growing body of theory and evidence demonstrates that institutional, educational and economic environments do more than merely moderate individual drivers of entrepreneurship: they generate the very beliefs and attitudes that propel entrepreneurial action. Institutional theory, for example, shows how clear property rights regimes, supportive financial institutions and predictable regulations reduce perceived barriers and cultivate a sense of institutional support (Estrin et al., 2012; Dai and Si, 2018; Hansen, 2019). Entrepreneurial ecosystem research reaches a similar conclusion from a resource perspective, arguing that training programmes, mentoring and networking infrastructures foster self-efficacy and risk tolerance (Fuerlinger et al., 2015; Theodoraki and Messeghem, 2017). Empirical studies in emerging economies confirm that business training interventions strengthen entrepreneurs’ confidence and performance (Kuzilwa, 2005), while university-based incubation initiatives increase students’ readiness to act (Fischer et al., 2019).

The economic climate constitutes a third pillar of contextual influence. When macro indicators signal growth, opportunity abundance and manageable risk, individuals display higher willingness to invest effort and capital in new ventures (Schoon and Duckworth, 2012). Survey evidence from Lebanese business students links strong social support, and, by extension, a favourable socioeconomic setting, to heightened entrepreneurial intentions (Dabbous and Boustani, 2023). Cross-country analyses of the BRICS economies reach the same conclusion: predevelopment policies and innovation oriented public spending stimulate venturing activity (Ayoungman et al., 2023).

Yet most empirical work still treats these contextual levers in isolation. Gonzalez Tamayo et al. (2024) show that perceived institutional support raises entrepreneurial intention only indirectly through self-efficacy and parental role models, but their model omits training and economic conditions. A comprehensive systematic review by Gallegos et al. (2025) corroborates this fragmentation, noting that contextual variables are usually modelled as controls or moderators rather than as integrated, formative antecedents of personal motivators. Such compartmentalisation hides the synergies that arise when institutional, educational and economic subsystems reinforce one another. To capture those synergies, we aggregate the three domains into a higher order construct, EES, specified as formative because each dimension supplies a distinct, noninterchangeable resource. Drawing simultaneously on institutional theory, the entrepreneurial ecosystem perspective and socioeconomic analyses, we contend that EES shapes (1) students’ confidence in their entrepreneurial abilities, (2) their overall attitude toward starting a venture and (3) their tolerance for calculated risk. Formally:

By positing EES as an antecedent, we extend Shapero’s notion of environmental “displacement” triggers and Ajzen’s TPB: a well-developed ecosystem mitigates structural barriers, supplies role models and capital, and thereby constructs the cognitive and affective foundations of entrepreneurial intention (Estrin et al., 2012; Fuerlinger et al., 2015; Schoon and Duckworth, 2012).

Self-efficacy, entrepreneurial attitude, and risk-taking propensity are frequently identified as key antecedents of entrepreneurial intention. Within the TPB, attitudes and perceived behavioural control (conceptualized as self-efficacy) constitute the motivational foundation of intention (Ajzen, 1991). Numerous studies focusing on students have consistently demonstrated that entrepreneurial attitude is typically the most significant predictor, followed, at a considerable distance, by self-efficacy and risk propensity (Wach and Wojciechowski, 2016; Doanh and Bernat, 2019). Social-cognitive research further suggests that high entrepreneurial self-efficacy mitigates perceived barriers and maintains effort amidst uncertainty (Bandura, 2004; Bachmann et al., 2020), while favourable entrepreneurial attitudes emerge from perceived desirability, social approval, and exposure to role models (Douglas and Prentice, 2019; Vamvaka et al., 2020). A willingness to take calculated risks complements these cognitive factors by encompassing tolerance for ambiguity, a trait associated with opportunity pursuit and growth-oriented strategies (Stewart and Roth, 2001; Kerr et al., 2017), although its influence varies across cultural and institutional contexts (Antoncic et al., 2018).

Although hypotheses concerning self-efficacy, attitude and risk propensity have been empirically supported, our interest lies in examining how strongly they operate when the entrepreneurial ecosystem, captured by our new formative construct EES, is modelled as an antecedent that creates these motivators. If EES truly forges the individual level cognitions on which intention rests, the magnitudes and mediation patterns reported in earlier work may shift. Retaining the canonical hypotheses therefore serves two purposes: it provides a validity check against the extant literature and allows us to assess whether ecosystem support attenuates, amplifies or redirects their effects.

Accordingly, and consistent with TPB, Social Cognitive Theory and prior findings on risk propensity, we state:

These hypotheses acknowledge that confidence in one’s capabilities, a favourable evaluation of entrepreneurship and tolerance for uncertainty jointly underpin the transition from opportunity recognition to entrepreneurial commitment; the novelty lies in testing them within an ecosystem centred causal architecture.

While entrepreneurial intention literature often views contextual factors as moderators of personal motivators (Prabowo et al., 2022; Anjum et al., 2024), we posit these factors as antecedents that shape individual beliefs and attitudes. This perspective aligns with Shapero and Sokol’s (1982) work on entrepreneurial events, which shows how conditions influence perceptions of feasibility and desirability, paralleling perceived behavioural control and attitude in Ajzen’s (1991) TPB. Krueger and Brazeal (1994) reframed TPB by emphasizing how external context shapes individual cognitions, noting that contextual facilitators affect the desirability-feasibility dynamic driving entrepreneurial actions. Supporting this view, Ayob et al. (2014) show how empathy and exposure to social entrepreneurship function as antecedents of personal motivators in venture creation.

Studies focusing on emerging economies show how contextual elements directly influence personal motivators. Hughes and Mustafa (2016) note that in developing regions, cultural and institutional realities shape how individual attributes manifest in corporate entrepreneurship. Research indicates the environment does more than “moderate” individual drives; it actively fosters or constrains them (Schmutzler et al., 2018). The literature reveals contradictions: while Anjum et al. (2024) view business incubation centres as moderators strengthening entrepreneurial intentions, others argue such structures may determine whether individuals develop attitudes or self-efficacy (Yurtkoru et al., 2014). Hmieleski and Baron (2008) found that under high environmental dynamism, entrepreneurial self-efficacy might have detrimental effects on firm performance, showing personal motivators interact complexly with contextual parameters. Despite these complexities, a recurrent theme is the significance of external context in shaping personal motivators. Institutional environments that offer transparent regulations, accessible funding, or supportive training programs can lead individuals to feel more efficacious, hold more positive attitudes toward entrepreneurship, and tolerate higher levels of risk (Urban and Kujinga, 2017). Conversely, cultural barriers or economic volatility may erode this sense of feasibility or desirability.

This study posits that contextual factors serve as antecedents that either foster or impede critical personal motivators, rather than merely moderating the interaction between motivators and intentions. With the integration of institutional support, training and support, and the general economic environment into a single higher-order construct, termed EES, the theoretical framework was adjusted to reflect this aggregated construct. EES represents broader contextual factors influencing entrepreneurial dynamics, encompassing regulatory frameworks, financial resources, training opportunities, and economic conditions conducive to entrepreneurship. It is proposed that EES impacts EI through its influence on SE, EA, and WTR. These personal motivators function as mechanisms through which the external environment shapes individuals’ propensity to engage in entrepreneurial activities. The following mediation hypotheses were proposed:

By consolidating contextual factors into a single formative-formative higher-order construct, the multidimensional nature of external entrepreneurial support and its holistic influence on individuals’ entrepreneurial mindsets are accounted for. A conceptual framework is proposed based on the preceding discussion, as illustrated in Figure 1.

This study was conducted as part of the Ingenium project, in which Gabriele d’Annunzio University (Ud’A) participated as a partner. Specifically, Work Package 8 (WP8) of the Ingenium Project focuses on promoting entrepreneurship among students, faculty, researchers, and administrative staff. In line with Ud’A’s objective of fostering entrepreneurial motivation and supporting students in launching their own ventures, this study had two primary aims: (1) to assess the entrepreneurial propensity of Ud’A students and (2) to identify the personal and social factors that may influence such propensity. To ensure a diverse sample, data were collected via an email campaign distributed to all university constituents with the intention of obtaining responses from a wide range of academic programs. The email clearly stated the purpose of the study and participation was voluntary. Additionally, data collection was conducted anonymously using digital questionnaires administered using Microsoft Forms. The final sample comprised 492 participants, including students, faculty, researchers, and administrative staff, as shown in Table 1.

The survey was conducted between January and February 2025, targeting the entire mailing list of Gabriele d’Annunzio University, which comprises approximately 21,108 students and 1,436 faculty and staff members, utilizing a non-probabilistic sampling method. A total of 492 questionnaires were returned, resulting in a response rate of approximately 2.18%. An email detailing the study’s objectives, an anonymity assurance, and a link to Microsoft Forms was sent once, with no monetary or grade-related incentives provided. Harman’s single-factor test and full collinearity variance inflation factors (VIFs) of less than 3.9 suggested that common method variance was unlikely to affect the estimates. Specifically, Harman’s test showed that the first factor explained 0.252 of the total variance, well below the 0.50 threshold (Kock, 2020), indicating that no single factor accounts for the majority of variance in our data.

A structured questionnaire was used to evaluate various constructs pertaining to entrepreneurial intention among university students, faculty, and staff. Each construct was operationalized through multiple items measured on a 5-point Likert scale, ranging from strongly disagree (1) to strongly agree (5). The constructs examined in this study include Institutional Support (INST), Training and Support (TS), General Economic Environment (GEE), Self-efficacy (SE), Entrepreneurial Attitude (EA), and Willingness to Take Risks (WTR). The specific items for each construct were adapted from scales established in the literature to ensure validity and relevance to the context of entrepreneurial intentions, as presented in Table 2. All measurement instruments were adapted following a twostep procedure.

1. For each latent construct we started from the most widely validated scale (e.g. Urban, 2013 for Institutional Support) and discarded items with redundant semantic content or poor fit with the university context (e.g. questions on export subsidies); (ii) the remaining items were translated into Italian using Brislin’s (1986) backtranslation technique. Because the sample comprises Italian university students, faculty and staff, items referring to national SME policy were reframed at the regional or university level.

Entrepreneurial ecosystem theory views the environment for entrepreneurship as a configuration of interdependent subsystems, regulatory quality, resource endowments, and sociocultural or educational infrastructures, all of which are crucial for stimulating entrepreneurial activity (Stam, 2015; Spigel, 2017). This systemic conceptualization draws from General Systems Theory (von Bertalanffy, 1968), which posits that complex phenomena exhibit emergent properties arising from the interactions among components rather than from the components themselves. In entrepreneurial contexts, this principle has evolved into the notion of innovation systems (Lundvall, 2010; Nelson, 1993) and, more recently, entrepreneurial ecosystems where the whole transcends the sum of its parts (Isenberg, 2010).

From this systemic perspective, we define EES as a second-order construct comprising three first-order dimensions: INST, TS and GEE. The formative specification of this construct reflects the fundamental nature of how ecosystems operate. Consistent with systems theory’s emphasis on hierarchical organization and emergent properties (von Bertalanffy, 1968), causal priority flows from the dimensions to the ecosystem, not the other way around. Improvements in the legal framework, availability of mentoring programs, or macroeconomic stability enhance the ecosystem; a “good” ecosystem cannot independently modify statutory tax incentives or establish incubators. As Jarvis et al. (2003) argue, when indicators are causes rather than manifestations of the latent variable, they become non-interchangeable: a vibrant venture capital market cannot compensate for absent entrepreneurial education, nor can extensive training replace dysfunctional credit institutions. Removing any dimension would fundamentally alter the conceptual domain of EES, meeting MacKenzie et al.’s (2011) criterion for formative constructs.

This formative hierarchy captures the synergistic nature that ecosystem scholars identify as the hallmark of successful entrepreneurial regions, aligning with von Bertalanffy’s (1968) principle of equifinality, whereby similar outcomes emerge from different initial conditions through systemic interactions. Both Feld’s (2012) “Boulder thesis” and Vedula and Kim’s (2019) complementarity argument emphasize how institutional, educational, and economic subsystems collaborate to reduce transaction costs and enhance perceived feasibility. The systemic approach recognizes that entrepreneurs experience their environment holistically, as an integrated field of affordances and constraints rather than as discrete policy instruments (Autio et al., 2018). Analysing these dimensions in isolation would miss this configurational logic and compromise our ability to develop actionable composite indices for policy benchmarking.

While these dimensions operate interdependently, they maintain clear conceptual boundaries, reflecting what systems theorists call “subsystem differentiation” (Luhmann, 1995). Institutional Support encompasses rule of law, regulatory transparency, and public incentives; Training and Support includes education, mentoring, and incubation resources; the General Economic Environment comprises demand conditions, capital availability, and macroeconomic stability. This differentiation enables each subsystem to process complexity according to its own logic while contributing to overall ecosystem functionality. Our empirical validation confirms this structure: heterotrait–monotrait ratios and Fornell–Larcker diagnostics demonstrate sufficient discriminant validity at the first-order level, while variance inflation factors below 1.5 indicate that the dimensions, though correlated, avoid redundancy (Becker et al., 2019).

The formative measurement approach also acknowledges that ecosystems exhibit what Boulding (1956) termed “organized complexity”, where excluding seemingly minor elements can disrupt systemic functioning. Some indicators, such as specific tax incentives within INST or access to university incubators within TS, show non-significant weights yet retain loadings above 0.40 and represent facets deemed essential by ecosystem theory and recent policy debates. Their exclusion would risk construct under specification and distort the contribution of remaining indicators (Hair et al., 2022). Our sensitivity checks confirm that retaining these theoretically critical items does not materially affect structural path estimates.

To operationalize this systemic approach, we employ the two-stage disjoint PLS-SEM procedure (Becker et al., 2019), obtaining latent variable scores for INST, TS, and GEE in the first stage and using them as formative indicators of EES in the second stage. This method prevents interpretational confounding while allowing reliability and validity assessment at both hierarchical levels, a practice particularly recommended for formative–formative models with heterogeneous dimensions (Guenther et al., 2023). Through this systemic, hierarchical approach, we preserve both the analytical tractability required for empirical testing and the holistic perspective necessary to understand entrepreneurial ecosystems as integrated wholes.

Structural Equation Modelling (SEM) analyzes relationships between observed and latent variables (Stein et al., 2017). It extends multiple regression and factor analysis, enabling examination of direct and indirect effects within a unified framework. SEM has been applied across entrepreneurship research, social sciences, psychology, tourism, and epidemiology (Amorim et al., 2010; Holbert and Stephenson, 2002). This study used Partial Least Squares Structural Equation Modelling (PLS-SEM), a variance-based approach suitable for theory building and prediction with smaller sample sizes (Hair et al., 2011). The analysis used SmartPLS 4 (Ringle et al., 2022), which estimates first-order and higher-order constructs while examining mediating effects. A key feature is the incorporation of a higher-order formative construct approach. We conceptualized contextual factors as dimensions of the broader construct EES. This second-order construct aggregates three first-order formative constructs, INST, TS, and GEE, to capture external conditions influencing entrepreneurial behaviour. This approach evaluates how institutional, educational, and economic factors shape entrepreneurial motivations. Using the two-stage disjoint approach proposed by Becker et al. (2019), this study addresses multicollinearity and measurement errors. The first stage assesses the measurement model for reliability, validity, and discriminant properties. The second stage evaluates structural model relationships between EES, personal motivators, and Entrepreneurial Intentions. This approach examines how contextual factors influence personal motivation and entrepreneurial intention.

Table 3 presents descriptive statistics for the latent variables, providing insights into the entrepreneurial landscape of the participants. EI exhibited a moderately high mean score (M = 3.31, SD = 1.03), indicating a general inclination toward entrepreneurship. EA yielded a slightly higher mean (M = 3.56, SD = 0.94), suggesting that the participants viewed entrepreneurship favourably. SE was comparable (M = 3.13, SD = 0.85), implying that respondents generally demonstrated confidence in their entrepreneurial abilities. The WTR variable recorded a moderate mean (M = 3.14, SD = 0.92), suggesting that the participants were reasonably open to uncertainty. Contextual variables were rated lower. INST had the lowest mean (M = 2.37, SD = 0.73), indicating perceived weak institutional structures that facilitate entrepreneurial activities. TS has a low mean (M = 2.59, SD = 0.82), reflecting the perception that entrepreneurship education needs enhancement. GEE also scored low (M = 2.32, SD = 0.77), suggesting economic conditions are not perceived as highly conducive to business creation. Examination of skewness and kurtosis values indicated that most of the variable distributions approximated normality. EI and EA demonstrated slight negative skewness (−0.38 and −0.63, respectively), implying a tendency toward higher scores. TS exhibited nearly zero skewness (−0.01), INST showed modest positive skewness (0.18), and GEE demonstrated small positive skewness (0.29). These statistics suggest that, while participants display a positive orientation toward entrepreneurship, external factors, particularly institutional, educational, and economic support systems, may play a pivotal role in shaping their entrepreneurial intentions.

This investigation employs a higher-order formative-formative construct approach, wherein three first-order formative latent variables are aggregated into a second-order formative construct termed EES. This methodology facilitates a comprehensive assessment of how institutional, educational and economic factors collectively influence entrepreneurial conditions. Furthermore, the study incorporated reflective latent variables as mediators, with EI as the dependent variable.

The reliability and validity of the first-order constructs were evaluated following guidelines for formative and reflective measurement models (Chin, 2010; Hair et al., 2011). Table 4 shows that the reflective constructs (EA, SE, WTR, and EI) demonstrated robust composite reliability (CR values of 0.939, 0.916, 0.913, and 0.944), Cronbach’s alpha (CA values of 0.918, 0.885, 0.881, and 0.925), and average variance extracted (AVE values of 0.755, 0.685, 0.678, and 0.77, respectively), indicating satisfactory internal consistency and convergent validity (Bagozzi and Yi, 1988). For formative first-order constructs (INST, TS, GEE), variance inflation factor (VIF) values were examined to assess multicollinearity, ensuring that all indicators remained below the critical threshold of 5.0 (Hair et al., 2011). Table 4 shows that the VIF values ranged from 1.579 to 3.849, indicating that multicollinearity was not a concern. Outer weights of some indicators were statistically non-significant (INST5, TS3, TS8, GEE6, GEE8); however, outer loadings above 0.4 justify their retention if conceptually relevant (Chin, 2010).

Discriminant validity (Table 5) assesses whether a construct is sufficiently distinct from other constructs in the model (Chin, 2010; Hair et al., 2011). According to the Fornell-Larcker criterion, discriminant validity is established when the square root of the AVE for each construct exceeds its correlation with other constructs (Fornell and Larcker, 1981). As shown in Table 5, the square root of AVE for each construct (EA = 0.869, EI = 0.878, WTR = 0.823, SE = 0.828) surpassed the correlations with the other constructs. This confirms that each latent variable captures a distinct concept and is not excessively correlated with others, thus fulfilling the discriminant validity requirement. The correlations between constructs indicate the expected relationships: EA and EI exhibit a strong correlation (0.811), suggesting that attitude toward entrepreneurship is a significant predictor of intention. SE and WTR displayed moderate correlations with EI (0.603 and 0.685, respectively), reinforcing their theoretical roles in influencing entrepreneurial intentions. These findings confirm the robustness of the first-order constructs, ensuring that the model accurately differentiates between latent variables.

Figure 2 presents the structural model after generating the second-order construct EES, illustrating the relationships among contextual factors, personal motivators, and entrepreneurial intentions. The model confirms EES significantly influences SE, EA, and WTR, which mediate its effect on entrepreneurial intentions. The R² values indicate the model’s explanatory power, with entrepreneurial intention displaying the highest explained variance (R² = 0.682). This suggests the three mediators account for a substantial portion of variability in participants’ entrepreneurial propensity (Cohen, 1988). Among personal motivators, entrepreneurial attitude exhibits the strongest direct effect on EI, reinforcing its role as a key determinant of entrepreneurial decision-making. Self-efficacy and willingness to take risks also contribute, reflecting their complementary influence on shaping entrepreneurial behaviour. The results demonstrate that while EES fosters a conducive environment for entrepreneurship, its impact is predominantly mediated by individual perceptions and motivations, underscoring the significance of cognitive and attitudinal factors in the entrepreneurial process.

Following the generation of Entrepreneurial Ecosystem Support as a second-order formative construct, we assessed its validity (Table 6). The weights of the first-order constructs (INST = 0.302, TS = 0.352, GEE = 0.619) indicate that each component contributed significantly to the higher-order construct. These values demonstrate that the aggregated EES construct effectively captures the combined influences of institutional, educational, and economic factors on entrepreneurship. Furthermore, the VIF values for the first-order constructs were below 1.5, confirming the absence of multicollinearity (Hair et al., 2011).

The Fornell-Larcker criterion results confirm that all constructs maintain discriminant validity, as the square root of the AVE for each construct is greater than its correlation with the other constructs (Table 7). This indicates that EA, EI, SE, and WTR were empirically distinct, reinforcing the integrity of the measurement model. The strong discriminant properties ensure that each latent variable captures a unique aspect of entrepreneurial behaviour without excessive overlap. In particular, the high discriminant validity of EA and EI highlights the distinct role of entrepreneurial attitudes in shaping intentions, while SE and WTR, although correlated, remain conceptually separate. These findings validate the theoretical soundness of the model and provide a solid foundation for further structural analysis.

Figure 3 and Table 8 present the results of the structural model analysis derived from bootstrapping with 5,000 subsamples. The findings demonstrate that EES has a statistically significant positive effect on self-efficacy (β = 0.317, p < 0.01), entrepreneurial attitude (β = 0.303, p < 0.01), and willingness to take risks (β = 0.372, p < 0.01), thus supporting H1a, H1b, and H1c. These results substantiate the proposition that contextual factors play a crucial role in shaping the personal motivators for entrepreneurship.

Among the mediators, EA demonstrated the strongest direct effect on EI (β = 0.632, p < 0.01, f² = 0.539), underscoring its central role in the decision to pursue entrepreneurship, and providing substantial support for H3. SE and WTR also contributed to EI, albeit with smaller effect sizes (β = 0.115, p < 0.01, f² = 0.022; β = 0.147, p < 0.01, f² = 0.027), thus supporting H2 and H4, respectively. The effect sizes (f²) were evaluated to determine the magnitude of each predictor’s influence, with values of 0.02, 0.15, and 0.35 considered small, medium, and large effects, respectively (Cohen, 1988). The effect of EES on EA (f² = 0.101) and WTR (f² = 0.16) fell within the medium range, while the effect of EES on SE (f² = 0.112) approached a medium threshold. By contrast, EA’s effect of EA on EI is substantial (f² = 0.539), demonstrating its dominant role in explaining entrepreneurial intentions, while SE and WTR exhibit small effect sizes (f² = 0.022 and 0.027, respectively). Overall, the model exhibited strong predictive power, with EI explaining 68,2% of the variance (R² = 0.682), indicating that the combination of self-efficacy, entrepreneurial attitude, and willingness to take risks effectively captured participants’ entrepreneurial intentions. These findings emphasize the significance of fostering a supportive entrepreneurial ecosystem as it indirectly influences EI through its impact on key psychological motivators.

Table 9 presents the results of the mediation analysis, confirming the indirect effects of entrepreneurial ecosystem support on entrepreneurial intention through self-efficacy, entrepreneurial attitude, and willingness to take risks. These results provide strong empirical support for the hypothesized mediation pathways.

The indirect effect of EES on EI through SE was significant (β = 0.036, p < 0.01, 95% CI [0.015, 0.065]), supporting H5. This finding indicates that while institutional, training, and economic support contribute to entrepreneurial intentions, their impact is partially mediated through individuals’ self-efficacy regarding their entrepreneurial capabilities. However, given its relatively small beta coefficient, SE demonstrates a modest mediating role in the relationship between EES and EI. EA emerged as the strongest mediator with a significant indirect effect of EES on EI (β = 0.191, p < 0.01, 95% CI [0.127, 0.257]), thus supporting H6. This result underscores the critical role of a positive perception of entrepreneurship in translating external ecosystem support into concrete entrepreneurial intentions, substantiating the notion that beyond structural and economic conditions, attitudes toward entrepreneurship function as a key psychological mechanism driving the motivation to initiate a business venture. The mediating effect of WTR was also significant (β = 0.055, p < 0.01, 95% CI [0.023, 0.092]), thus supporting H7. This result suggests that a broader entrepreneurial ecosystem influences risk tolerance, which fosters entrepreneurial intention. However, similar to SE, the effect size is relatively small compared to EA, indicating that, while risk-taking propensity is an important factor, it plays a less prominent role in shaping entrepreneurial intentions than attitude. Overall, these findings confirm that EES influences EI both directly and indirectly through key psychological motivators, particularly entrepreneurial attitudes. The results emphasize the necessity for policies and educational programs that not only strengthen institutional and economic support, but also cultivate favourable attitudes toward entrepreneurship, as this pathway appears to be the most influential in shaping the intention to initiate a business venture.

Given our theoretical proposition that entrepreneurial ecosystems function as formative forces, a critical empirical question emerges: do these ecosystems exert uniform influence across the diverse constituencies of a university? To address this, we employed the rigorous three-step MICOM procedure (Henseler et al., 2016), comparing our largest stakeholder groups, students (n = 376) and lecturers/administrative staff (n = 116). The measurement invariance tests yielded reassuring yet revealing results (Table 10). All constructs achieved both configural and compositional invariance, with correlations between group-specific and pooled composites remaining statistically indistinguishable from unity (p > 0.05). This establishes partial measurement invariance, confirming that our constructs carry equivalent meaning across groups and that structural comparisons are methodologically sound.

However, the mean comparisons unveiled a compelling pattern that speaks directly to our theoretical framework. Students demonstrated significantly higher scores on entrepreneurial attitude, willingness to take risks, and entrepreneurial intentions, precisely the psychological outcomes our model predicts should emerge from ecosystem exposure. Most intriguingly, variance equality was rejected only for EES itself, indicating that students perceive the university’s entrepreneurial resources through a more heterogeneous lens than their faculty and staff counterparts. This dispersed perception suggests that students, as primary targets of ecosystem interventions, experience these resources in more varied and perhaps more personally meaningful ways.

The multi-group structural analysis deepened these insights (Table 11). Within the student subsample, the ecosystem’s formative power emerged with striking clarity: EES exerted substantial effects on SE (β = 0.357), EA (β = 0.300), and WTR (β = 0.392), all highly significant (p < 0.001). The pathway from attitudes to intentions proved particularly robust (β = 0.641), confirming that for students, ecosystem-shaped attitudes powerfully drive entrepreneurial aspirations. The faculty/staff subsample revealed a subtly different psychological architecture. While EES continued to shape personal motivators in the expected directions, the pathways from self-efficacy and risk-taking to intentions failed to achieve conventional significance (p = 0.132 and 0.333, respectively). Only the attitude-intention link remained robust, suggesting that for established university members, entrepreneurial intentions flow primarily through evaluative beliefs rather than through confidence or risk tolerance.

Though the multi-group analysis found no statistically significant differences in path coefficients (two-tailed p ≥ 0.144), the numerical patterns tell an important story. The ecosystem’s influence on self-efficacy and risk-taking proved stronger for students (Δβ = +0.160 and + 0.094), as did the attitude-intention relationship (Δβ = +0.114). These patterns suggest that entrepreneurial ecosystems achieve their formative effects most powerfully among those still forming their professional identities, students who encounter ecosystem resources not as established professionals but as individuals actively constructing their career possibilities. This differential impact illuminates a crucial theoretical insight: while entrepreneurial ecosystems can influence all university stakeholders, their capacity to fundamentally reconstruct psychological foundations appears strongest among those most open to identity transformation. The ecosystem functions not merely as a universal resource but as a developmentally sensitive intervention, resonating most powerfully with those navigating the liminal space between education and career. This finding reinforces our conceptualization of ecosystems as formative forces while highlighting the importance of considering stakeholder heterogeneity in ecosystem design and assessment.

The present inquiry advances entrepreneurial-intention theory on four, mutually reinforcing fronts. First, by operationalising EES as a second-order formative construct, the study converts the configurational logic of ecosystem scholarship (Stam, 2015) into a tractable measurement model. In doing so, it demonstrates empirically that contextual forces act upstream of the canonical TPB variables: EES influences intention almost exclusively through its effects on self-efficacy, entrepreneurial attitude and risk propensity, rather than as a distal moderator. This finding extends the TPB by inserting a theoretically grounded antecedent layer that captures the systemic quality of entrepreneurial environments. Second, the relative strength of the mediation paths revises the prevailing hierarchy of cognitive determinants. When contextual variance is modelled holistically, entrepreneurial attitude, an evaluative “want-to” cognition, absorbs a larger share of the ecosystem effect than self-efficacy, the operative “can-do” belief traditionally deemed dominant (Wach and Wojciechowski, 2016). This shift suggests that favourable ecosystem signals first alter desirability perceptions and only subsequently reinforce perceived behavioural control. Third, the multi-group analysis shows that the structural paths do not differ significantly between students and academic staff, implying that a mature university ecosystem may homogenise entrepreneurial mind-sets across otherwise heterogeneous campus constituencies (Brentnall et al., 2023). This null difference tempers assumptions that students are uniquely sensitive to ecosystem cues and indicates that shared institutional routines can level entrepreneurial cognition among diverse stakeholder groups. Fourth, the data reveal a compensatory mechanism that nuances the resilience argument posited by Schmutzler et al. (2018). Within a setting characterised by objectively modest ecosystem scores, respondents nonetheless report moderate-to-high intentions, signalling reliance on dispositional resources such as risk tolerance. However, the mediation analysis shows that this compensation is partial: structural deficits still attenuate the total effect on intention, implying boundary conditions to the resilience of personal motivators. Collectively, these contributions integrate ecosystem theory with intention models, recalibrate the relative salience of cognitive antecedents under varying contextual quality, delineate how institutional maturity can equalise stakeholder cognitions, and articulate limits to individual agency in adverse environments.

The pattern that emerges from our data, modest entrepreneurial intentions coexisting with notably low evaluations of institutional support and local economic prospects, suggests that strengthening a university centred ecosystem need not begin with large injections of capital. Rather, the most powerful levers appear to be those that alter how potential founders see the environment: their sense that launching a venture is both desirable and realistically achievable. An obvious starting point lies in the university’s own rules of engagement. When procedures for recognising a spinoff, allocating intellectual property rights, or gaining access to shared laboratories are opaque or slow, they blunt enthusiasm before it can crystallise into concrete plans. Streamlining these pathways, and making the revised steps highly visible, would signal institutional commitment and increase perceived feasibility, a precursor of the positive entrepreneurial attitudes that our model identifies as pivotal.

Perceptions of capability can be reinforced when structured learning is coupled with the tangible resources required to act. Integrating modest seed fund opportunities into experiential courses, hackathons or incubator cycles not only equips participants with technical skills but also demonstrates that promising ideas will not stall for lack of initial finance. By nesting funding within a learning context, the university converts training from a largely cognitive exercise into a bridge toward real market engagement. Yet capability without opportunity is rarely persuasive. A more buoyant sense of demand can be cultivated through outward facing collaborations, for example, initiatives in which regional agencies announce challenge briefs in areas aligned with public priorities such as digital transition or sustainable manufacturing. When students can trace a direct line from a societal problem to a procurement call and on to a prototype developed inside the university, the abstract notion of “market opportunity” becomes immediate and concrete. Finally, attitudes toward risk are shaped as much by cultural signals as by actuarial probabilities. Policies that normalise iteration, clear guidelines on how a venture may pivot or restart without punitive loss of intellectual property, mentoring that frames early setbacks as expected learning moments, reduce the perceived severity of failure. In doing so, they temper risk aversion while respecting prudent decision-making, a balance that our findings show to be associated with higher intentions. Taken together, these measures, procedural transparency, resource infused experiential learning, externally anchored opportunity pathways and an institutionalised second chance ethos, offer a coherent strategy for universities operating in resource constrained regions. By targeting precisely those ecosystem dimensions that our respondents judged weakest, they maximise the likelihood that incremental improvements will translate into the attitudinal and motivational shifts required for a more vibrant culture of student and staff entrepreneurship.

The study’s evidence that relatively modest gains in institutional transparency and student oriented training programmes translate into markedly stronger pro entrepreneurial attitudes carries important consequences for regional development policy. It indicates that governments seeking to curb youth outmigration and invigorate local innovation systems need not rely exclusively on largescale financial incentives; targeted improvements in procedural clarity and skill building infrastructures can yield comparable attitudinal dividends at substantially lower cost. At the same time, the results caution against overreliance on promotional campaigns that showcase entrepreneurial “success stories” without corresponding reforms to the underlying ecosystem. Absent tangible enhancements in the rules, resources and learning opportunities available to nascent entrepreneurs, narrative strategies alone appear unlikely to shift the attitudinal calculus of potential founders.

While our study advances understanding of how integrated ecosystem support shapes entrepreneurial intentions, several limitations warrant acknowledgement and point toward productive research directions. First, our mixed sample of students, faculty, and staff, while reflecting the university’s entrepreneurial community, may hide important heterogeneity in how different stakeholder groups experience ecosystem support. As Henry et al. (2024) compellingly argue, entrepreneurship programs require “systematic inclusivity-proofing” to ensure equitable access and impact. Students seeking career development, faculty pursuing research commercialization, and staff supporting institutional advancement likely respond differently to the same ecosystem characteristics. This heterogeneity, compounded by demographic differences across gender, ethnicity, and socioeconomic background, suggests that our aggregate findings may mask significant variation in EES effectiveness (Villegas-Mateos and Amann, 2025). Second, our cross-sectional design captures only a snapshot of what is inherently a dynamic process. Entrepreneurial ecosystems evolve, as do the individuals within them. Longitudinal research tracking how ecosystem changes influence the trajectory of entrepreneurial development would provide crucial insights into the temporal dynamics our static model cannot capture (Syed and Spicer, 2025). Such studies could illuminate critical periods when ecosystem support matters most and identify how different support elements gain or lose salience over the entrepreneurial journey. Third, the single-university, single-country context limits generalizability. While Italy’s complex institutional environment provides a compelling case for studying ecosystem effects under challenging conditions, testing our model across diverse economic and cultural contexts would strengthen its theoretical robustness. Comparative studies spanning established entrepreneurial economies and emerging markets could reveal how cultural and institutional factors moderate the ecosystem-intention relationship (Estrin et al., 2012; Schmutzler et al., 2018). Finally, our focus on intentions leaves unexplored the crucial transition to action. While intentions predict behaviour, the entrepreneurial journey from intention to venture creation involves numerous barriers and facilitators that ecosystem support may differentially address (Kautonen et al., 2015). Future research should extend our model to examine how EES influences not just the formation of intentions but their translation into entrepreneurial action.

Despite these limitations, our demonstration that entrepreneurial ecosystems function as formative forces rather than mere facilitators opens important theoretical and practical avenues. The challenge now is to develop more nuanced models that preserve the insight of ecosystem integration while accounting for the rich diversity of entrepreneurial experiences within university communities. Such work promises to advance both our theoretical understanding and our ability to design ecosystems that nurture entrepreneurial potential across all university stakeholders.

This research study solely involved the collection of fully anonymous questionnaire data, and it did not gather any identifying or sensitive information that could lead to participant identification. In accordance with our institution’s guidelines and national regulations, studies that collect anonymous data and present minimal risk to participants do not require formal approval from an ethics review board. Therefore, no specific clearance was obtained for this research.

All participants received an introductory information sheet before completing the questionnaire, which clearly outlined the study’s objectives, the voluntary nature of participation, and the anonymity of the data collected. By choosing to proceed with and submit their responses, participants provided implicit informed consent. The data were processed solely for academic research purposes, in compliance with D.Lgs.196/2003 as amended by D.Lgs.101/2018, the EU General Data Protection Regulation (Reg. UE, 2016/679), and the relevant institutional and international ethical guidelines.