When context shapes gender diversity: women in management and the performance of academic spin-offs and innovative startups

Academic spin-offs (ASOs) play a major role in promoting academic research from a commercial perspective and transferring new technologies from publicly funded research to industry (Di Gregorio and Shane, 2003; O'Shea et al., 2008; Meoli et al., 2013; Bozeman et al., 2015; Mathisen and Rasmussen, 2019). Being founded by faculty members as well as PhD, graduate and undergraduate students, an ASO is a new venture that benefits from university affiliation and aims to commercially exploiting the technology developed within universities (Rasmussen and Borch, 2010; Colombo et al., 2010). While prior literature has extensively examined the determinants of ASO creation (Lockett et al., 2005; Rasmussen and Borch, 2010; Fini et al., 2020; Veltri et al., 2022), comparatively less attention has been devoted to understanding the drivers of their performance. This gap is particularly relevant because, despite their high technological profile, ASOs often exhibit lower growth than other innovative ventures (Ensley and Hmieleski, 2005; Mustar et al., 2008).

In line with upper echelons theory, the composition of top management teams is a critical determinant of firm outcomes (Hambrick and Mason, 1984; Hambrick et al., 1996; Carpenter et al., 2004; Hambrick, 2007). Within this stream of research, the role of women in management has received increasing attention. Research has examined the role of female executives across a variety of contexts (Jennings and Brush, 2013; Acs et al., 2011), including both established firms and startups, as well as gender differences in academic entrepreneurship (e.g. Rosa and Dawson, 2006; Hayter, 2015; Abreu and Grinevich, 2017; Miranda et al., 2017; Di Paola, 2020; Lauto et al., 2022; Civera and Meoli, 2023).

Studies suggest that gender diversity may enhance decision-making quality, innovation and firm performance. However, empirical evidence remains mixed, with positive, negative and nonsignificant effects reported across different contexts.

A growing body of literature suggests that these inconsistent findings may be explained by contextual and institutional factors. In particular, the institutional environment may shape the extent to which diversity translates into organizational outcomes. Academic spin-offs represent a particularly relevant setting in this respect, as they are deeply embedded within university systems characterized by specific norms, structures and constraints. Although universities may offer legitimacy, knowledge resources and supportive infrastructures, they may also reproduce hierarchical dynamics and gender stratification, which influence how diversity operates within management teams (Corley and Gaughan, 2005; Muscio and Vallanti, 2024).

Research on academic entrepreneurship has increasingly examined gender differences in entrepreneurial intentions, participation, constraints and opportunities (Rosa and Dawson, 2006; Croson and Gneezy, 2009; Hayter, 2015; Abreu and Grinevich, 2017; Miranda et al., 2017; Di Paola, 2020). Yet only a limited number of studies have focused on the performance implications of women's involvement in academic spin-offs (e.g. Rodríguez-Gulas et al., 2018; Sciarelli et al., 2021; Prencipe et al., 2023) or explicitly compared academic and nonacademic innovative ventures (Civera and Meoli, 2023; Francois and Belarouci, 2022). In parallel, most studies on innovative new ventures tend to focus on female presence in founding teams, rather than on broader managerial involvement (Clarysse and Moray, 2004; Lim and Suh, 2019; Liao et al., 2023).

Taken together, these gaps leave open the important question of how context shapes the relationship between female involvement in management and firm outcomes. Addressing this issue, our study investigates whether and how the effect of female presence in top management teams differs in academic spin-offs compared to innovative startups not embedded in university systems. Accordingly, we ask: How does context shape the effect of gender diversity in academic entrepreneurship? Does female managerial presence affect firm performance differently in ASOs relative to nonacademic innovative startups?

To address our research question, we designed a large-scale quantitative study, gathering data on 1,581 Italian ASOs (corresponding to about 77% of the country population of ASOs) and 2,980 Italian innovative startups in the period 2015–2023. In line with a recent route in research on academic entrepreneurship (Lauto et al., 2022), we go beyond the dichotomy between fully men-controlled versus fully women-controlled firms. Instead, by focusing on the proportion of women within the management team, we examine women's role in terms of degree of involvement rather than presence/absence. In this study, firm performance is operationalized as sales growth, which represents a more appropriate proxy for performance in early-stage and innovation-driven ventures, where growth rather than profitability is typically the primary objective (Ensley and Hmieleski, 2005; Clarysse et al., 2010; Rodríguez-Gulas et al., 2018). Our analyses show that the relationship between female involvement in management and growth performance differs between academic spin-offs and innovative startups not anchored to universities.

This paper contributes to the literature in three main ways. First, we provide large-scale empirical evidence on the relationship between female involvement in management and firm growth in the context of academic entrepreneurship. We thus further our understanding of how women's participation in management can impact the performance of ASOs. Specifically, our analysis complements prior studies on the effect of female ownership on the ability to acquire financial resources (e.g. Lauto et al., 2022) and the resource endowments of ASOs (Rodríguez-Gulas et al., 2018), by shifting the focus on the relationship between female management and the growth performance of ASOs. Second, we adopt a comparative perspective between academic spin-offs and innovative startups, highlighting the role of institutional embeddedness in shaping the gender diversity–performance relationship. In doing so, we underscore the uniqueness of ASOs and the critical role of the context in explaining why the presence of women in the management team may be beneficial or detrimental to firm performance. Third, we offer contextually grounded interpretative mechanisms, namely as “constrained diversity” and “double-layered balance”, that help explain when and why gender diversity may generate different outcomes across organizational settings. In the context of our study, these mechanisms are intended as theoretically informed interpretations rather than directly tested causal channels. Specifically, we posit that in academic spin-offs, gender diversity may configure as a “constrained diversity”, since male and female (academic) managers typically share a similar scientific background and institutional experience, rooted in the university context. In contrast, in innovative start-ups, the presence of women in management is more likely to introduce heterogeneity of backgrounds and experiences, which are generally considered to contribute to a greater variety of strategic perspectives and ultimately to diversity benefits. Furthermore, we introduce the concept of “double-layered balance”, required of academic women entrepreneurs: beyond the well-documented challenge of reconciling professional and family responsibilities, academic women must simultaneously manage the dual roles of scholar and entrepreneur. These multiple tensions represent an additional layer that may reduce female managers' commitment to firm growth in ASOs, relative to innovative startups.

Our findings also offer interesting implications for policymakers attempting to foster technology transfer and the commercial exploitation of academic research, showing that “one-size-fits-all” gender policies may be ineffective. While ASOs require interventions that mitigate the constraints of academic environments, innovative startups may benefit more from policies that support gender diversity advantages.

The paper is organized as follows: in Section 2, we present our theoretical framework and develop our hypotheses. Section 3 focuses on our research methodology and Section 4 presents the results of our empirical analysis. Section 5 discusses the findings and implications of the study for both theory and practice and outlines future research directions.

Women have traditionally faced difficulties in accessing top management positions (Rosener, 1995). However, over the last 2 decades, the percentage of women who have reached the upper echelons of organizations has gradually increased (Oakley, 2000; Terjesen et al., 2009). Such change has raised an increasing interest in the analysis of the distinctive traits of female managers, including their contribution to firm performance. This line of inquiry is rooted in the upper echelons theory, which posits that decision-makers’ demographic characteristics, personality traits, values and experiences strongly influence organizational decisions and performance (Hambrick and Mason, 1984; Hambrick et al., 1996; Carpenter et al., 2004; Hambrick, 2007). Specifically, studies indicate that the characteristics and composition of the entire management team may better explain organizational decisions and outcomes than those of a single decision-maker, such as the CEO (Carpenter et al., 2004; Bauweraerts et al., 2022). Focusing on the team composition, research suggests that, relative to men, women can offer different, often more creative, perspectives on problem-solving, potentially extending the range of strategic options available to firms (Wiersema and Bantel, 1992; Kakabadse et al., 2015; Dezsö and Ross, 2012) and leading to positive outcomes in terms of creativity, entrepreneurship and innovation (Torchia et al., 2011; Lyngsie and Foss, 2017; Foss et al., 2021). Several studies also highlight that a greater presence of women increases group diversity, which in turn enhances the quality of brainstorming and fosters deeper and more fruitful information search, discussion and the exchange of ideas (Torchia et al., 2015; Prencipe et al., 2023). For instance, research indicates that greater gender diversity promotes more open work environments (Nielsen and Huse, 2010), streamlining communication, stimulating knowledge sharing and spurring innovation (Eagly and Carli, 2003; Bauweraerts et al., 2022).

Evidence has also shown that female directors tend to be more receptive to change and more stakeholder-oriented than their male counterparts (Adams and Funk, 2012). Other studies highlight that women managers possess conflict resolution skills that enhance their relationships management capabilities (Stroebe et al., 2017; Brewer et al., 2002; Eagly and Karau, 1991). Moreover, the multiple roles that women perform in their personal lives may equip them with skills that enhance leadership and managerial competences (Ruderman et al., 2002).

Despite these arguments, empirical research on the effect of female participation in top management teams on firm performance has yielded mixed evidence, indicating a lack of consensus on whether the involvement of women in top executive teams improves performance. Some studies report a positive effect (Krishnan and Park, 2005; Mahadeo et al., 2012; Lückerath-Rovers, 2013), while others find negative (e.g. Inmyxai and Takahashi, 2012; Pathan and Faff, 2013) or nonsignificant relationships (e.g. Dezsö and Ross, 2012; Jia and Zhang, 2013; Zhang et al., 2013).

Among the possible interpretations for these mixed findings, two considerations deserve particular attention. First, while women bring attributes that may enhance firm performance, such advantages may materialize only when women reach a critical mass within the management team (Torchia et al., 2011). A limited presence may not be sufficient to make women's voices heard in the decision-making process, thereby limiting their impact on team cognition and the realization of diversity benefits. Second, empirical evidence suggests that the effect of female involvement in management on firm performance is contingent upon both firm-level contextual factors and top management team characteristics (Mensi-Klarbach, 2014). In particular, entrepreneurship research underscores the importance of context in shaping the emergence and success of female entrepreneurship (Terjesen et al., 2009; Welter, 2011; Civera and Meoli, 2023), highlighting the role of elements such as financial resources, social capital, policies and university support (Jennings and Brush, 2013; Foss et al., 2013; Lauto et al., 2022; Muscio and Vallanti, 2024). The venture's innovativeness and technological environment have also emerged as key contextual factors influencing the outcomes associated with female presence (e.g. Liao et al., 2023).

Following this research route, in the next sections, we examine the potential impact of female management on performance in the context of ASOs and develop a theoretical framework to explain why and how the presence of women may display varying effects across different entrepreneurial contexts.

Research highlights that the characteristics of the work environment, such as inclusiveness, openness to diversity and gender equality, shape the contribution of women to firm management and performance. For instance, Hoobler et al. (2016) find that female leaders are more likely to positively affect their organization's performance in more inclusive and gender egalitarian contexts. Similarly, Post and Byron (2015) find that the presence of women in top management is more positively related to firm performance in contexts with greater gender parity. Krishnan and Park (2005) further suggest that women are more likely to be perceived as leaders in environments characterized by high levels of social interaction.

Conditions of greater inclusiveness and gender parity are more likely to occur in the university environment than in other contexts. In academia, women tend to enjoy more equal access to education, employment and political empowerment (Hausmann et al., 2012). Moreover, although academia is still regulated by masculine norms, universities are often characterized by work structures and cultural norms that support more inclusive relations between men and women (Bilimoria et al., 2008). These features suggest that organizations like ASOs may enable women to more fully leverage their managerial capabilities and thus positively contribute to ASOs' performance.

However, empirical evidence on the performance effects of female presence in ASOs and innovative ventures remains mixed. Civera and Meoli (2023) show that women-led firms underperform in terms of growth but exhibit higher survival rates, while academic female entrepreneurs outperform nonacademic ones and exhibit more similar performance to men-led firms. In contrast, Rodríguez-Gulas et al. (2018) show that firm growth is not affected by gender differences among Spanish ASOs. Other studies point to persistent biases: drawing from signaling theory, Lauto et al. (2022) argue that female ownership may signal lower venture viability to investors; Liao et al. (2023) underscore that such bias is amplified in more innovative ventures. Civera et al. (2022) argue that organizational and contextual factors may alter the performance effect of gender diversity in the entrepreneurial team. In general, review studies and meta-analyses on gender and entrepreneurship suggest that the underperformance of women-led firms is driven by barriers, such as limited access to finance, weaker networks and cultural constraints, rather than by differences in competence (Klapper and Parker, 2011; Jennings and Brush, 2013; Elam et al., 2019; Muscio and Vallanti, 2024). This view is also consistent with research stressing that women's entrepreneurial outcomes are shaped by structural and institutional conditions rather than by individual capabilities alone.

We contribute to this line of inquiry by focusing on the top management team rather than comparing women-led and male-led firms. Consistent with the upper echelons theory, we argue that the team is the relevant unit of analysis when examining the performance effects of management traits. Since the influence of female managers depends on reaching a critical mass (Torchia et al., 2011), the proportion of women in the team matters: a greater presence can increase collegial support (Foss et al., 2013), enabling women to contribute more effectively to strategic decisions.

The performance effect of a greater female presence in top management is generally explained in terms of the benefits of gender diversity (Zhang, 2020), as women may enrich strategic decision-making through a broader range of backgrounds, experiences and perspectives. However, we interpret the pattern expected in ASOs through the lens of “constrained diversity”: in academic spin-offs, male and female (academic) managers typically share similar scientific backgrounds, rooted in the university context. As a result, gender diversity may not substantially broaden managerial or market-relevant competences, limiting the degree of cognitive complementarity within the team and preventing the expected diversity benefits from materializing. The university environment may therefore “flatten” gender differences (Lauto et al., 2022), making academic teams relatively homogenous, particularly from the perspective of investors.

A further mechanism that may explain the negative performance effects of female involvement in academic spin-offs is what we interpret as a “double-layered” balance faced by academic women entrepreneurs. Beyond the well-documented challenge of reconciling professional and family responsibilities, academic women may simultaneously need to manage the dual roles of scholar and entrepreneur. While these role demands may also affect male counterparts, they may be particularly salient for women due to persistent gendered expectations and the unequal distribution of family responsibilities in many contexts (Hochschild and Machung, 2012; Jennings and Brush, 2013). This layered pressure may constrain the time and energy devoted to the spin-off, limiting their ability to drive growth and reducing the potential benefits of their participation in top management.

Based on the above arguments, we hence formulate the following:

The technological environment in which a firm starts up is crucial for boosting or limiting its potential (Malerba and Orsenigo, 1997). In particular, technological resources are the main drivers of success in spin-offs, and the way technological knowledge is transferred to spin-offs is at the core of different abilities to achieve market performance. A relatively recent literature has investigated how different characteristics of the technological knowledge base at startups influence spin-off performance, by comparing corporate spin-offs with university spin-offs (Clarysse et al., 2010). For example, Wennberg et al. (2011) found that corporate spin-offs, thanks to the entrepreneurs' prior industry experience, perform better than university spin-offs, where entrepreneurs come from a research-focused environment and therefore typically lack a commercial orientation (Mustar et al., 2006).

Building on the recognition of the impact of context on entrepreneurship (Welter, 2011), we contribute to this research stream by investigating the performance of ASOs relative to other innovative firms. ASOs are generally considered a subset of new innovative firms, as they share a focus on the exploitation of technological innovation and face similar difficulties in achieving successful growth and performance (Mustar et al., 2006). ASOs and innovative startups tend to operate in highly dynamic environments and exhibit similar characteristics in terms of both market and financial uncertainty, due to the novelty of the technology and their strong reliance on hard-to-value intangible assets. However, a fundamental difference between the two types of firms lies in their institutional environment: ASOs are created by university personnel, and knowledge originates in a university context (Colombo et al., 2010; Rasmussen and Borch, 2010; Francois and Belarouci, 2022), whereas innovative startups do not benefit from the support of university structures and services. This contrast is also in line with work suggesting that entrepreneurial contexts differ in their opportunity structures and institutional conditions (Aldrich and Fiol, 1994; Shane, 2000; Terjesen et al., 2009).

We argue that the embeddedness in an academic environment leads to substantial differences between ASOs and other innovative ventures, which may explain the varying effect of gender diversity – namely a greater presence of female managers in the management team – on firm growth (Zhang, 2020). In innovative start-ups, the presence of women in management often introduces heterogeneity of backgrounds and experiences, contributing to richer strategic perspectives and a stronger market orientation. In contrast, in academic spin-offs, male and female (academic) managers typically share a much greater homogeneity of background and experience as a result of their common university affiliation. Thus, as noted above, such a “constrained diversity” may negatively affect firm performance. Furthermore, in the specific context of ASOs, the contribution of female executives in top management may be hampered by tensions related to the “double-layered balance”, that is, women's need to reconcile not only professional and family roles but also academic and entrepreneurial roles.

Therefore, the mechanisms through which gender diversity, measured as an increase in the proportion of women among top executives, may hinder firm performance are likely to operate more strongly in ASOs than in innovative startups. Accordingly, we expect the negative performance effect of female management to be greater in ASOs. Stated formally:

Italy is the research setting of our study. In this country, ASOs emerged in the early 2000s and increased steadily with time (Source: Link to the website). As a first step for building our dataset, names and VAT numbers of all the Italian ASOs were retrieved from the website of Netval (Link to the website). Netval is a nonprofit association, having public and private Italian universities and research institutes as partners, whose main target is to strengthen the promotion and exploitation of academic research, enhance technology transfer and new venture creation for the commercialization of research results. In 2022, the Netval website reported a total of 2,043 ASOs in Italy. We matched this list of ASOs with the Orbis (Bureau Van Dijk, BvD) database, using the company VAT numbers, to check the availability of accounting and company information. Full financial data and information on management and ownership structure were available for 1,581 ASOs, covering around 77% of the Italian ASO population, which were analyzed over the period 2015–2023. On average, these firms were observed for about 6.5 years, corresponding to 10,386 firm-year observations.

To test hypothesis 2, based on the comparison between ASOs and innovative startups, we built a sample of “innovative startups”, i.e. new ventures operating in high-tech industries. In Italy, the expression “innovative startups” refers to a specific category of firms, introduced by Law n. 221/2012 with the aim of offering fiscal and financial incentives to firms that develop and produce innovative high-tech products or services. To benefit from these incentives, companies must meet multiple requirements. First, they should possess an innovative character, which is defined by the presence of at least one of the following conditions: (1) at least 15% of the company's expenses can be attributed to R&D activities; (2) at least 1/3 of the total workforce are PhD students, holders of a PhD or researchers; alternatively, 2/3 of the total workforce must hold a Master's degree; and (3) the enterprise is the holder, depository or licensee of a registered patent or the owner of an original registered computer program. Second, innovative startups are required to develop, produce and commercialize innovative goods or services of high technological value. Third, these firms must not result from a merger, split-up or divestiture of a company or branch. Fourth, firm age is limited to five years or less. Fifth, firms must be headquartered in Italy or in another EU country with at least a production site in Italy. Sixth, firm turnover should not exceed 5 million euros. Finally, firms must not have distributed profits. Data on firm names and VAT numbers of innovative startups are available from the Italian Ministry for Economic Development; however, for simplicity, we directly collected the data from AIDA (BvD), which provides the corresponding list of innovative startups.

To avoid overlap between the two firm groups in our analysis, ASOs that were also qualified as innovative startups were excluded from the latter group. After selecting only firms for which accounting data were available, we obtained a sample of 2,980 innovative startups from 2006 to 2018, covering 20% of the Italian population of innovative startups. These firms were analyzed on average for 6.4 years over the period 2015–2023, ranging from a minimum of 1 year to a maximum of 9 years, and corresponding to 19,272 firm-year observations.

Table 1 shows the list of variables used and their description. Growth performance is the dependent variable of the study. Growth and profitability represent two key performance indicators (Jang and Park, 2011). However, since for both ASOs and innovative ventures in their early stage, growth, rather than profitability, is likely to be the primary firm objective (Ensley and Hmieleski, 2005; Clarysse et al., 2010; Rodríguez-Gulas et al., 2018), firm growth can be considered a more appropriate proxy for performance. We therefore measure firm growth as an annual percentage change in firm sales (Firm growth).

As far as our independent variables are concerned and in line with extant literature (Civera and Meoli, 2023), female involvement in management is operationalized as the ratio between the number of female managers in the company and the total number of managers (Female management). Women in management are present in 22% of the ASOs and 12% of the innovative startups.

To ensure that our findings are robust to alternative ways of assessing the presence of women in management, we also used additional measures of female involvement in management, as shown in the section on robustness tests. To test hypothesis 2 on the varying effect of female management in different types of firms, we included a dummy variable (Spinoff) equal to 1 for ASO and 0 for innovative startups.

According to main literature (Ensley and Hmieleski, 2005; Rodríguez-Gulas et al., 2018; Lauto et al., 2022; Prencipe et al., 2023), we control for several factors that may explain growth performance. Firm age is measured by the number of years since the firm was incorporated. Cash holdings is measured as the ratio of cash and cash equivalents to total assets. We also control for the effect of profitability, which can be considered a further proxy of the financial resources available to the firm. Profitability is measured as the ratio between EBITDA (earnings before interest, tax, depreciation and amortization) and total assets (Profitability). Firm size is measured by the natural logarithm of total assets. Furthermore, we control for the effect of the financial structure by including short- and long-term financial debt (Debt). Specifically, this variable is operationalized as the natural logarithm of (1 + total financial debt), allowing us to account for skewness while retaining firms with zero debt. Two additional control variables are related to the ownership and governance structure of the firm: Foreign management is a binary variable equal to 1 if at least one manager has a nationality other than Italian and 0 otherwise; Venture capital is a binary variable taking the value of 1 if the firm has at least one venture capitalist in its capital, 0 otherwise.

The empirical model to test hypothesis 1 considers ASO performance as a function of female involvement in management (Female management), in addition to controls, as follows:

To test our hypothesis 1 on the effect of female involvement in management in ASOs, regression analysis was performed on the sample of ASOs only (N = 10,386). To test our hypothesis 2 on the difference in the effect size of female management in ASOs and innovative startups, we carry out the regression analysis on the pooled sample of both firm types (N = 29,658 from 10,386 ASOs and 19,272 innovative startups), using an interaction term for a dummy variable Spinoff, that is a binary variable equal to 1 for ASOs and 0 for innovative startups and female management (Female management x Spinoff). The following model is used in regression to test hypothesis 2:

All main model specifications include year and industry-fixed effects to control for unobserved heterogeneity across sectors and time. In terms of estimation method, we used both ordinary least square (OLS) models and models based on the instrumental variable (IV) technique to address endogeneity concerns. There are mainly two potential causes of endogeneity: (1) omitted variable bias and potential reverse causality, as firms with higher expected performance may systematically differ in their managerial composition. In addition, the share of female executives may both influence firm performance and be influenced by it, since better-performing firms may hire more employees, including women (Abdallah et al., 2015); (2) unobservable factors that may simultaneously affect both the independent and dependent variables of the models. Previous studies (Adams and Ferreira, 2009; Reguera-Alvaredo et al., 2015) suggest that endogeneity is a critical reason for mixed findings in studies on female presence in management and firm performance. Indeed, higher-performing firms are more likely to have the resources needed to promote greater gender diversity and therefore to have more women in top management positions than lower-performing firms.

From an empirical standpoint, we applied many tests to investigate endogeneity. First of all, we used the Durbin–Wu–Hausman test (augmented regression test), which suggested the lack of endogeneity issues. Afterward, we tested the presence of endogeneity with the “female involvement in management” variable. The IV strategy allows us to isolate the exogenous variation in female involvement in management. Identification relies on two key assumptions: relevance, meaning that the instrument is sufficiently correlated with the endogenous variable and exogeneity, implying that the instrument affects growth performance only through its impact on female managerial involvement (Bascle, 2008). We tested endogeneity using the Stata endog option in three models: (1) the model with ASOs, (2) the model considering both ASOs and innovative startup and (3) the model including the interaction term Female involvement in management × Spinoff. These tests failed to reject the null hypothesis of exogeneity for the variable Female management. Thus, the null hypothesis that female involvement is endogenous was rejected, providing evidence that female involvement in management is not an endogenous variable. Furthermore, in line with recent literature (Rieger et al., 2022; Roccapriore and Pollock, 2022; Cadenovic et al., 2023), we follow Busenbark et al. (2022) and apply the robustness of inference to replacement (RIR) method, based on the KONFOUND test (Kenneth and Xu, 2017) to assess the robustness of our estimates to potential endogeneity bias, accounting for multiple sources of bias beyond omitted variables (Frank et al., 2013). Such a test indicates the percentage of the estimates that should be biased because of endogeneity in order to overturn a statistically significant parameter (Busenbark et al., 2022). In our models (1), (2) and (3) above defined, the RIR tests showed a considerable variability (ranging from 33% to 62%) in the potential endogeneity bias. Although, according to several tests fulfilled, endogeneity can be excluded, we wanted to verify how our results could change, in terms of sign and magnitude, running a 2SLS model. In our 2SLS model, the variables to be instrumented are those concerning the female involvement in the management. It is worth noticing that finding a fully valid instrument is far from trivial as it implies both a correlation with the endogenous variable and an unrelatedness with the error term, i.e. variables that contribute to determining the presence of women in management but that are not related to performance (Bascle, 2008). We relied on prior studies to select an appropriate instrument. Following Garnero et al. (2014), we argue that the level of involvement of women is contingent upon industry characteristics, since in certain industries, such as the fashion and clothing industries, a greater involvement of women typically occurs. We thus assume that industry affiliation affects the likelihood of the firm hiring female managers. Specifically, our selected instrument is the industry-average presence of women in management. The theoretical rationale underpinning the use of this instrument is that the industry-average presence of female is very likely to explain a significant amount of firm's presence of women in a firm's management team but does not directly affect firm performance.

Overall, the IV approach yields results that are consistent with the OLS estimates in both sign and magnitude. The validity of the IV approach relies on the relevance and exogeneity of the instruments, which are supported by the standard diagnostic tests discussed above. Therefore, these results should be interpreted as providing causal evidence conditional on instrument validity rather than as definitive causal estimates.

Tables 2 and 3 show the descriptive statistics of the ASOs and innovative startups, respectively, and Table 4 reports the mean statistical differences in terms of T-test ^[1].

On average, innovative startups show higher levels of sales growth relative to ASOs (3.00 versus 2.27), with higher variability, but a lower presence of female managers (0.15 versus 0.16). Specifically, female managers are present in 1,233 out of 10,386 ASOs (11.8%) and in 1,416 out of 19,272 innovative startups (7.3%) ^[2]. Tables 5 and 6 do not show particularly relevant correlations.

To ensure that multicollinearity was not an issue, we also computed variance inflation factors (VIFs) of all the variables. The highest VIF is 1.60, which is far below the generally employed cut-off of 10 (or, more cautiously, 5) for regression models, confirming that multicollinearity did not bias our results.

Table 7 shows the main results of the econometric analysis carried out to test hypotheses 1 and 2. Column 1 displays the regression results on the sample of ASOs, while Column 2 displays the sample of innovative startups. Column 3 shows the results of the pooled sample of ASOs and innovative startups. Specifically, Columns 1, 2 and 3 use OLS as an estimation method, whereas Column 4 presents the estimates based on the IV technique, as described above. Interestingly, some control variables, such as Tangibility, exhibit different signs and levels of significance across ASOs and innovative startups. This pattern further supports the appropriateness of treating the two groups as distinct samples, as it suggests the presence of different underlying dynamics.

Hypothesis 1 posits that female involvement in management negatively affects ASO performance. The coefficient for Female management, capturing women's involvement in management, is negative and statistically significant in Column 1 (p < 0.01), suggesting that a higher presence of women in the management team is associated with lower ASO performance. This result supports hypothesis 1, although its interpretation should remain cautious and context-specific. Instead, Column 2 of Table 7 reports the results for the sample of innovative startups, showing the lack of any statistical significance. Columns 3 and 4 of Table 7 display the results on the pooled sample of ASOs and innovative startups, incorporating the binary variable Spinoff (reflecting the difference for ASOs as compared to innovative startups) and the main effect of the variable Female management, using OLS and IV models, respectively. Columns 3 and 4 of Table 7 also display the interaction term Female management × Spinoff. The coefficient for Spinoff is negative and statistically significant, indicating that firm growth is lower in ASOs, compared to innovative startups. The moderating effect, measured by the interaction term Female management × Spinoff, is negative and statistically significant in Columns 3 and 4 (p < 0.05), suggesting the negative performance effect of female management is strengthened in ASOs, with a different role in ASOs in comparison to innovative startups. Thus, hypothesis 2 is confirmed: the negative relationship between women involved in management and growth performance is stronger in academic spin-offs compared to innovative startups.

Figure 1 graphs the interaction effect from the estimated regression coefficients in Column 3.

This figure shows that as female involvement in management increases, ASO performance tends to decrease, whereas the opposite relationship is observed for innovative startups, where performance increases with higher levels of female management. This plot confirms the varying role of female management in ASOs and innovative startups.

We conducted several additional tests to delve deeper into the effects of female management and assess the robustness of our findings. We employed alternative measures to capture the focal variable of our study, i.e. female involvement in management. Specifically, based on the idea that the effect of the presence of female managers can be significant, especially when women hold a high proportion of management positions, we replicated our analyses using alternative variables as independent ones. First, we use the variable Number of female managers, which accounts for the number of women involved in the management of the company.

Moreover, we consider the case of management teams dominated by the presence of women, which best reflects the achievement of a female critical mass (Torchia et al., 2011) in our sample, characterized by a small size of the management teams. Specifically, we consider the subset of the ASOs identified by the variable named Critical mass, that is a binary variable equal to 1 when women among managers are equal or greater than three. Such a relevant presence of women occurs in 1,816 cases (405 ASOs and 1,411 innovative startups).

Finally, as a further case of women-led ASOs, we measure the effect of female management through the binary variable Wholly female management, which takes the value of 1 when the management team is entirely formed by women. Such a relevant presence of women occurs in 1,688 cases (688 ASOs and 1,000 innovative startups).

Table 8 shows the regression results when these different proxies of female management – Number of female managers, Female critical mass and Wholly female leadership – are used. Consistent with the model specification used in Table 8, in the analyses carried out on the pooled sample of ASOs and innovative startups, the binary variable Spinoff is included to capture the difference in performance between spinoffs and innovative startups, using also an interaction term.

All the results are in line with our main outcomes. The results of these analyses are similar to those presented in Table 7, suggesting the findings are robust to the use of different measures of female management. Regarding the difference between ASOs and innovative startups, the negative and statistically significant interaction terms number of female managers x Spinoff (column 3, p < 0.01), critical mass x Spinoff (column 6, p < 0.01), and Wholly female management x Spinoff (column 9, p < 0.01) indicate the effect of female management is negative in ASOs relative to innovative startups, further supporting our hypothesis 2.

In addition, to further support our approach, we conducted analyses using a binary indicator of female leadership (female CEO). While the results are broadly consistent, this specification is less informative, as it does not capture management team composition. Accordingly, we rely on team-level measures in line with Upper Echelons Theory; CEO-based results are available upon request.

Further tests are reported in Table 9 to examine our research question in greater depth. As an additional robustness check, we first assess the effect of female involvement in management by lagging the explanatory variables. We then present two additional specifications using industry-adjusted sales growth as the dependent variable to account for sector-specific growth dynamics. Moreover, we extend the analysis by examining the impact of female involvement on long-run growth. Finally, we investigate how female involvement in management affects the likelihood of entry into the product market, distinguishing between ASOs and innovative startups.

In Column 1, the results based on lagged explanatory variables confirm our main findings. Both the sign and significance of the key coefficients remain broadly unchanged, supporting the robustness of our baseline results.

In Columns 2 and 3 of Table 9, the results based on industry-adjusted sales growth further confirm our main findings. Female management retains a negative and statistically significant effect in ASOs, while the interaction term remains negative and statistically significant. The magnitude of the coefficients is even larger, suggesting that our results are not driven by sector-specific growth differences.

We further assess the robustness of our findings by adopting alternative measures of growth performance. Specifically, we operationalize firm growth as a three-year moving average of sales growth over t₀–t₂ in Column (4), while in Column (5), we consider sales growth after three years for ASOs and innovative startups. The results reported in these columns are consistent with our main findings in Table 7, confirming that the presence of women in management is associated with lower firm growth also in the longer run.

In the last Column of Table 9, we report an additional test concerning entry into the product market. Particularly in the context of research-based entrepreneurial initiatives, achieving accelerated growth is closely linked to full entry into the market. In most cases, an ASO gets revenues from nonoperational activities, meaning that the firm is still incubated or a “proto-company” (Fasano et al., 2023). Indeed, beyond revenues derived from sales, most ASOs also benefit of revenues related to grants, start-up competition awards and similar activities. ASOs can be defined as having fully entered the product market when their sales largely outnumber their nonoperational revenues. Thus, we explore the role of women in accelerating market entry by performing the analyses using the likelihood of firm entry into the product market as the dependent variable. Specifically, consistent with Fasano et al. (2023), we operationalized the entry into the product market with a binary variable, taking the value of 1 if firm sales outnumber three times or more nonoperating revenues (24,159 cases), and zero otherwise (5,499 cases). Given the binary nature of the dependent variable in these models, univariate binary choice models are appropriate. Specifically, we adopted a probit specification, which falls into this class of models.

The findings of these analyses are shown in Table 9. Column 2 displays the results of the probit estimates on the pooled sample of ASOs and innovative startups. It still shows the difference between ASOs and innovative startups, with the former showing a negative effect of female management, highlighting the role of women as an obstacle to entering the product market.

These findings provide further support for our core arguments on the effect of the presence of women in the management of ASOs. In Section 5, we will offer further interpretation of these findings, as well as provide a discussion of the implications of our research.

The role of women in business has increasingly attracted the interest of entrepreneurship and management scholars. Building on the upper echelons theory (Hambrick et al., 1996; Carpenter et al., 2004; Hambrick, 2007), we examine the relationship between the presence of women in the top management team and growth performance in ASOs and innovative startups. Consistent with our prediction, we find a negative association between female managers and ASO growth performance. In contrast, the relationship is positive in innovative startups, when statistically significant. Our findings are robust to the use of alternative measures of female involvement in management, including cases in which women reach “critical mass” (Rosener, 1995; Torchia et al., 2011) or dominate the management team. A similar positive effect of female management on performance in innovative startups has also been recently documented by La Rocca and Schifilliti (2025). In contrast, our results for ASOs differ from prior empirical evidence on Italian ASOs offered by Civera and Meoli (2023). However, our analysis covers a substantially larger sample (1,581 ASOs versus 420) over a recent period (2015–2023 versus 2006–2018). In line with the mixed evidence in the literature (e.g. Dezsö and Ross, 2012; Bauweraerts et al., 2022; Civera and Meoli, 2023), our results show that the relationship between gender diversity and performance is not uniform but varies significantly across institutional environments. This highlights the contextual nature of gender diversity effects and suggests that institutional, organizational and cultural factors shape the impact of women's involvement in management.

Our findings should be interpreted with caution. The observed negative association between female involvement in management and performance in academic spin-offs should not be understood as reflecting differences in managerial competence between men and women. Rather, our results point to contextual and institutional constraints that may limit the extent to which female managers can fully contribute to firm outcomes in specific environments. From this perspective, the academic context, which is characterized by strong institutional embeddedness, hierarchical structures and demanding career paths, may shape the way diversity operates within management teams. These contextual features may constrain the realization of diversity benefits that are more likely to emerge in more flexible and market-oriented environments such as innovative startups. Consistent with the upper echelons perspective (Hambrick and Mason, 1984; Hambrick, 2007), our findings point to the idea that team heterogeneity translates into richer strategic cognition only when all voices are effectively integrated into decision-making. In this sense, our study suggests that the performance effects of team diversity may depend not only on team composition but also on the conditions that enable diverse perspectives to shape strategic outcomes.

This paper makes contributions to different research streams. First, we contribute to the research on the role of women in the upper echelons by offering large-scale empirical evidence on the performance effect of female management in the context of academic entrepreneurship. Research has shown that the involvement of women in leading management or governance roles offers valuable contributions in terms of innovation, creativity, conflict resolution and networking capability (Wiersema and Bantel, 1992; Rosener, 1995; Brewer et al., 2002; Nielsen and Huse, 2010; Dezsö and Ross, 2012; Bauweraerts et al., 2022). We enrich this research line by investigating the specificities of academic ASOs. In particular, we complement prior research by shifting the focus from the individual to the team as a level of analysis, hence from the dichotomy male-managed versus female-managed firm to the degree of female involvement in the management team (Civera and Meoli, 2023).

Second, this paper contributes to the research on the role of context in shaping the gender diversity-performance link by exploring the varying sensitivities of ASOs and innovative startups to the effects of female involvement in management. Prior studies have shown that, although these two types of firms share common characteristics, in terms of orientation to innovation and the crucial role of technology, the difference in terms of embeddedness in a research environment and link with the universities has substantial implications (Mustar et al., 2006). We further extend this line of inquiry by showing how such a difference also has implications in terms of the performance effects of gender diversity in the upper echelons. By comparing academic spin-offs and innovative startups, our results indicate that the effects of female involvement in management are contingent upon institutional embeddedness (i.e. the degree to which ventures are anchored in the formal and informal rules, cultural norms and hierarchical structures of the university environment), thus suggesting that the mixed findings of prior research may be explained, at least in part, by differences in institutional and organizational contexts.

Our study contributes to the literature by offering a context-sensitive interpretation of the gender diversity–performance relationship. Particularly, we interpret our findings through mechanisms such as “constrained diversity” and double-layered balance. In academic spin-offs, gender diversity may operate as a form of constrained diversity, where similar backgrounds between women and men limit the emergence of complementary perspectives. In contrast, in innovative startups, gender diversity may more fully display its spillover effects. In addition, female managers need to strike a balance between the demands of academic careers and entrepreneurial ventures, on top of work–life challenges. This double-layered balance may dilute the potential for effective contribution of women to the firm's growth trajectory, making the diversity costs more visible than the benefits. While the mechanisms of constrained diversity and double-layered balance are consistent with prior literature, they are not directly tested in our empirical models and should therefore be interpreted as theoretically informed explanations.

The practical implications of our results are threefold. First, for managers of ASOs, our results suggest that increasing female representation in top management may not be sufficient and could be less effective in the absence of supportive organizational conditions. In this respect, initiatives such as structured mentoring programs, more inclusive decision-making processes and greater attention to workload and role design may help create an environment in which women can contribute more effectively. Managers may also consider policies that support the integration of academic and entrepreneurial responsibilities, for instance through flexible role design and clearer allocation of managerial tasks. Without these supporting mechanisms, initiatives aimed at increasing gender diversity risk being less effective or falling short of their intended goals. Second, for universities, incubators and technology transfer offices, our findings suggest that gender diversity should not be treated as a purely numerical objective. Instead, these actors could complement diversity targets with concrete support mechanisms, such as targeted mentoring for female academics, leadership development initiatives and networking platforms that facilitate access to entrepreneurial ecosystems. Similarly, entrepreneurial support programs, venture incubators, and investors could promote female-led innovation teams not only by providing funding, but also by integrating capacity-building activities (e.g. mentoring, coaching, and governance support) aimed at strengthening inclusive leadership and decision-making practices. Third, from a policy standpoint, our evidence suggests that policy makers should avoid one-size-fits-all gender policies and instead recognize that the barriers and enabling conditions faced by women may differ across entrepreneurial contexts. This implies designing differentiated support schemes for ASOs and innovative startups, by combining, for example, financial incentives with programs that address context-specific structural and cultural barriers. Overall, our results highlight the importance of policies that take into account the interaction between institutional environments and gender dynamics in shaping entrepreneurial outcomes.

Several limitations of our study should be acknowledged. Although our dataset is unusually rich, the analysis is limited to the Italian context. Moreover, despite the inclusion of industry fixed effects and the use of industry-adjusted performance measures, we acknowledge that residual sectoral heterogeneity may still influence our results. Different industries are characterized by distinct technological regimes, growth opportunities and competitive dynamics, which may shape both the composition of management teams and firm performance. Although our empirical strategy mitigates these concerns, future research could further explore the interaction between gender diversity and sector-specific characteristics by adopting more fine-grained industry classifications or focusing on single-industry settings. In addition, while we discuss a set of plausible mechanisms underlying the observed relationships, our empirical models do not directly test them. Future research could therefore more directly examine the channels through which institutional embeddedness shapes the gender diversity–performance relationship, for example, by developing empirical proxies for constrained diversity, access to decision-making power or the role conflicts experienced by academic women entrepreneurs.

More broadly, our study points to several promising avenues for future research on the interplay among women in management, performance and entrepreneurial context. Future research could extend this study through a more in-depth examination of the causality nexus through research designs better suited to causal identification. In particular, future research could complement our analysis by employing propensity score matching techniques to create more balanced and comparable subsamples of academic spin-offs and innovative start-ups, thereby strengthening causal inference and mitigating issues that may arise from differences in sample size and observable characteristics. In a similar vein, the role of moderators and the boundary conditions of the female management-performance relationship deserve further investigation. Scholars could examine how institutional settings (e.g. national gender norms, university policies), industry characteristics (e.g. high-tech vs low-tech), funding structure (public vs private) and organizational culture moderate the gender–performance link. Moreover, to understand the role of the institutional context, both cross-country and cross-institutional comparisons are needed; accordingly, extending this study beyond the Italian setting would help understand how generalizable these effects are and the extent to which the effects of women's involvement are contingent upon institutional and country contextual factors, especially how it varies in countries with different gender equality institutions or academic entrepreneurial ecosystems. Such research efforts would help clarify the boundary conditions under which gender diversity in management contributes to growth in innovation-driven ventures.

To sum up, our study reinforces the notion that context matters when assessing the effects of female leadership on firm performance. For both scholars and practitioners, then, gender diversity in management should not be seen as a universal panacea, but as a strategic dimension whose value is contingent on organizational design, institutional support and contextual conditions.