Foreign minds, local legitimacy: the innovation dilemma of returnee entrepreneurs

Returnee entrepreneurship is recognized as a critical phenomenon in today's globalized economy. In China, returnee entrepreneurs (REs) play a pivotal role in reshaping innovation landscapes by combining international expertise with local market knowledge (Ahlstrom et al., 2008; Dai and Liu, 2009; Liu et al., 2010; Qin and Estrin, 2015; Qin et al., 2017; Lin et al., 2019; Liu, 2020; Tran and Truong, 2022). Our research question is stated upfront: How do REs in China leverage dual knowledge (local and overseas) such that local stakeholder engagement translates into enhanced innovation performance via legitimacy, and how do these knowledge sources condition that process? This question matters in China because legitimacy is socially embedded (Granovetter, 1985), while Guanxi, meaning trusted relational ties and reciprocal obligations, and network positions shape whether local stakeholders see a venture as acceptable (Ahlstrom et al., 2008; Burt and Burzynska, 2017). Prior work documents distinctive capabilities and resilience among returnees relative to local entrepreneurs (Liu, 2020), yet their ability to convert such assets into locally recognized legitimacy remains a critical challenge (Ahlstrom et al., 2008; Alexiou and Wiggins, 2018). This integration is foundational not only for gaining stakeholder trust but also for enhancing innovation performance (Salancik, 1977; Delmar and Shane, 2004; Fisher et al., 2016).

Although the literature underscores the importance of legitimacy and the competitive advantage conferred by international experience, a research gap remains in understanding how REs balance overseas and local knowledge to achieve legitimacy. Prior research has examined overseas knowledge, learning and recontextualization as separate advantages (Delmar and Shane, 2004; Liu et al., 2015; Fisher et al., 2016; Ren and Liu, 2019; Tran and Truong, 2022; Haddoud et al., 2025). Yet less attention has been paid to why local knowledge may help REs gain stakeholder acceptance, whereas overseas knowledge may create greater innovation value only after such acceptance has been secured (Alexiou and Wiggins, 2018).

Addressing this research question is vital because it explains how REs in rapidly evolving emerging markets convert foreign and domestic knowledge into local recognition. Legitimacy matters here because it opens access to resources, endorsement, and collaboration. We therefore contribute a focused account of knowledge duality under conditions of social embeddedness (Granovetter, 1985) and Guanxi salience in China (Ahlstrom et al., 2008; Burt and Burzynska, 2017). The Chinese context makes this issue sharper, since institutional and cultural differences can turn overseas knowledge into either a source of credibility or a source of local doubt (Liu, 2020).

One of the primary theoretical contributions of this research lies in its extension of established frameworks such as the Knowledge-Based View (KBV) and legitimacy theory. We clarify constructs and mechanisms early, distinguishing external and perceived legitimacy (Alexiou and Wiggins, 2018) and linking legitimation activities to commitment and resource mobilization (Salancik, 1977; Delmar and Shane, 2004). By integrating and refining these theories, the study demonstrates that while international knowledge can be a valuable asset, it may also be a liability if misaligned with local norms. This dual-role perspective not only enriches the current theoretical discourse but also broadens the applicability of these theories to diverse institutional settings, thereby enhancing their explanatory power (Fisher et al., 2016).

Empirically, the study advances current knowledge by providing robust evidence on how REs' dual knowledge influences their ability to gain legitimacy and, in turn, drive innovation performance. The findings clarify previous ambiguities regarding the roles of local and overseas knowledge and refine conceptual frameworks in the field of returnee entrepreneurship. Contributions are presented at multiple levels: from the specific phenomenon of REs' innovation performance to broader implications for the application of KBV and stakeholder theory, ultimately enriching the academic dialog on cross-border knowledge transfer.

The insights from this study offer significant implications beyond academia. For practitioners and policymakers, understanding the nuanced interplay between foreign and local knowledge can inform targeted support mechanisms for RE integration and innovation. We also outline future inquiry on perceived legitimacy measurement (Alexiou and Wiggins, 2018) and on network-embedded integration strategies grounded in social embeddedness (Granovetter, 1985). These broader implications underscore the study's value as a steppingstone for subsequent research on international knowledge integration and legitimacy-building.

Following this introduction, Section 2 presents a detailed review of the theoretical background and the development of the hypothesis. Section 3 outlines the methodology, including data collection and measurement scales. Section 4 discusses the results of the empirical analysis, while Section 5 highlights theoretical and practical implications. Finally, Section 6 concludes with suggestions for future research.

We consolidate the literature into a single stream and define the core constructs: local stakeholder engagement, legitimacy, innovation performance, local knowledge, and overseas knowledge. Legitimacy is a social judgment of acceptance, appropriateness, and desirability, with a measurable perceptual component (Zimmerman and Zeitz, 2002; Alexiou and Wiggins, 2018). Granovetter's embeddedness argument is useful here because economic exchange often relies on known reputation rather than formal safeguards alone. In China, this logic is visible in Guanxi, meaning trusted relational ties and reciprocal obligations, through which stakeholders assess whether a venture is credible and acceptable. Thus, legitimacy formation is socially embedded (Granovetter, 1985), while Guanxi and network position shape local recognition (Ahlstrom et al., 2008; Burt and Burzynska, 2017). Grounded in the Knowledge-Based View, we treat local and overseas knowledge as moderators of the paths from local stakeholder engagement to legitimacy and from legitimacy to innovation performance (Salancik, 1977; Delmar and Shane, 2004; Fisher et al., 2016).

Returnee entrepreneurs are individuals who return to their home country and establish a new venture after several years of professional experience or study abroad (Filatotchev et al., 2009; Gabay-Mariani et al., 2024). Returnees are more than 50% more likely to become entrepreneurs than locals (Uhlbach et al., 2022). This difference is not only a demographic fact. It reflects the entrepreneurial value of international mobility, which can expose returnees to advanced technologies, new market routines, foreign networks, and wider opportunity comparisons before they reenter the home market (Dai and Liu, 2009; Liu et al., 2010; Wang et al., 2011a, b; Qin and Estrin, 2015; Tran and Truong, 2022).

Over the last decade, returnee entrepreneurship has attracted growing attention across three connected strands. One strand examines whether returnee-led ventures perform and innovate differently from local ventures in high technology sectors (Dai and Liu, 2009; Liu et al., 2010; Li et al., 2012). Another study examines how overseas knowledge, transnational ties, and home-country embeddedness are transferred, adapted, and sometimes lost upon return (Qin and Estrin, 2015; Lin et al., 2019; Tran and Truong, 2022). A third focuses on institutional uncertainty, political and business relationships, and returnee liability in emerging economies (Bai et al., 2021; Mreji and Barnard, 2021). We view this attention as important because REs can bring uncommon knowledge to the home market, yet that knowledge improves innovation only when local stakeholders treat the venture as credible, useful, and appropriate.

That tension explains why legitimacy is central to REs rather than peripheral. After time abroad, REs often need to reengage local actors to rebuild recognition and trust (Wang et al., 2011a, b). Legitimacy is a social judgment of acceptance, appropriateness and desirability (Zimmerman and Zeitz, 2002). It helps young ventures survive, grow, and access resources because stakeholders are more willing to provide support once the venture appears reliable (Elfring and Hulsink, 2003; Kawai et al., 2020; Trunina et al., 2020; Zhang et al., 2021). The challenge is sharper for returnees because overseas knowledge may be attractive, but it can also look unfamiliar or misaligned with local norms. REs therefore face a double task: they must use foreign knowledge while also making that knowledge locally understandable through stakeholder engagement, commitment, and reputation-building (Ahlstrom et al., 2008; Lin et al., 2019; Liu, 2020).

Knowledge duality is therefore a defining feature of REs (Liu et al., 2015; Ren and Liu, 2019; Tran and Truong, 2022). Overseas knowledge includes technological, market, financial and network knowledge acquired in the host country, which differentiates REs from local entrepreneurs (Dai and Liu, 2009; Liu et al., 2010; Kenney et al., 2013; Qin and Estrin, 2015). Local knowledge refers to updated understanding of the home market, local institutions, and stakeholder expectations after return (Lumpkin and Lichtenstein, 2005). Prior research shows that overseas knowledge is useful only when recontextualized for the home market (Tran and Truong, 2022). In this China-based study, we therefore treat local and overseas knowledge as boundary conditions on the paths from local stakeholder engagement to legitimacy and from legitimacy to innovation performance (Delmar and Shane, 2004; Fisher et al., 2016).

Despite studies on the impact of knowledge (Wiklund and Shepherd, 2003) and legitimacy (Zimmerman and Zeitz, 2002) on firm survival and performance, our understanding remains limited of how knowledge shapes the acquisition of legitimacy in returnee entrepreneurship. Thus, our research question is: how does knowledge duality, overseas versus local, moderate the processes of legitimization and subsequently enhance REs' performance in innovation? We frame this through the lens of social embeddedness (Granovetter, 1985) and China-specific Guanxi dynamics (Ahlstrom et al., 2008; Burt and Burzynska, 2017). This study refers to REs' performance in innovation as noncontinuous innovation and the adoption of disruptive technologies (Chandy and Tellis, 1998; Christensen, 2013; Aránega and Cañero Serrano, 2024).

Previous studies have examined the returnee entrepreneurship phenomenon in various regions and countries, such as India (Pruthi, 2014), Mexico (Hagan and Wassink, 2016), Vietnam (Tran and Truong, 2022), Central and Eastern Europe (Martin and Radu, 2012; Piracha and Vadean, 2010), Africa (Mreji, 2020; Mreji and Barnard, 2021); however, there is a lack of research in China.

China is one of the countries with the most accelerated returnee entrepreneurship (Miao and Wang, 2017; Qin et al., 2017). Favorable government policies have encouraged Chinese professionals and graduates to return from developed economies and establish new ventures (Liu et al., 2018; Muldoon et al., 2024). Yet policy support does not automatically produce local acceptance. It remains unclear how REs in China balance and integrate overseas and local knowledge and how this balance affects legitimacy and innovation outcomes. This study therefore explores knowledge duality in Chinese returnee technology ventures and examines how legitimacy mediates the relationship between local stakeholder engagement and innovation performance. We clarify that legitimacy is the mediator, while local and overseas knowledge operate as moderators.

This study investigates the research question of RE-owned technology-based SMEs who have recently come from abroad to Zhongguancun, one of the largest concentrations of Chinese RE and China's first national high-tech zone in Beijing. Zhongguancun has a solid entrepreneurial atmosphere and innovative culture, targeting international frontiers and comprehensively serving innovation and entrepreneurship around the country (Dong et al., 2019; Liu et al., 2018; Ferreira et al., 2024).

The analysis of 178 respondents from the respondents demonstrates different types of knowledge (overseas versus local) effects on the enhancement or reduction of the legitimacy and performance of RE-owned technology-based SMEs.

The study provides theoretical implications for returnee entrepreneurship and the Knowledge-Based View (Kaplan et al., 2001; Barney, 2002; Balogun and Jenkins, 2003; Dai and Liu, 2009; Liu et al., 2010; Qin and Estrin, 2015; Qin et al., 2017; Lin et al., 2019; Tran and Truong, 2022). It also links organizing activities and commitment to legitimation (Salancik, 1977; Delmar and Shane, 2004) and connects venture identity with legitimacy across venture evolution (Fisher et al., 2016). First, it extends returnee entrepreneurship research by showing why knowledge is not automatically valuable: REs must first make their venture credible in the local market. Second, it enriches the Knowledge-Based View of innovation by applying it to returnee entrepreneurship, where knowledge moves across both national and sectoral boundaries. Policy and managerial implications will be discussed in the concluding section.

Knowledge is a critical strategic resource of a company (Kaplan et al., 2001; De Carolis, 2002) and a basis for explaining differences in firm performance and competitive advantage. The Knowledge-Based View originates from the Resource-Based Theory and treats knowledge as a resource that can generate sustained advantage (Barney, 1991; Kogut and Zander, 1992; Ariely, 2003; Makhija, 2003; Balogun and Jenkins, 2003). Knowledge is intangible, socially embedded within a company, difficult to imitate, and useful for differentiation (Wiklund and Shepherd, 2003). In this study, local knowledge refers to context-specific know-how gained after return, while overseas knowledge refers to technological and market know-how acquired abroad. We theorize both as moderators that condition the relationship from local stakeholder engagement to legitimacy and from legitimacy to innovation performance. This theorizing fits a setting in which knowledge is socially embedded (Granovetter, 1985) and Guanxi shapes the local interpretation of entrepreneurial action (Ahlstrom et al., 2008; Burt and Burzynska, 2017).

Two forms of knowledge affect REs' performance in innovation – knowledge received abroad (overseas knowledge) and knowledge obtained after returning (local knowledge) (Liu et al., 2015; Tran and Truong, 2022). REs' overseas knowledge enhances the likelihood of starting the business by 20% (Dai and Liu, 2009). Such knowledge is respected by the government entities in China (Ren and Liu, 2019). Some scholars estimate the influence of overseas knowledge of REs on sales growth, profitability in the local and international markets, level of internationalization, and export propensity (Bai et al., 2017; Filatotchev et al., 2009; Liu et al., 2015). Tran and Truong (2022) explain that Vietnamese overseas knowledge must be recontextualized to be impactful for businesses in the home country. It is an essential process since, without it, opportunities raised from overseas knowledge might be rejected or demolished by the local market (Värlander et al., 2016).

Local knowledge is acquired through observation of the environment, other firms' business experience, best practices, and advanced technology (Lumpkin and Lichtenstein, 2005). Vicarious learning is a mechanism for making sense of overseas knowledge (Tran and Truong, 2022) through interaction with local professionals, following home-market news, and attending formal and informal networking sessions. A lack of necessary local industry experience in the specific environment will hinder REs from developing the entrepreneurial attitudes, beliefs, and abilities specific to that environment. Local knowledge is deeply spatially embedded. It may take time for REs to learn and/or update their local knowledge. These definitions ensure the knowledge constructs are central rather than secondary in our framework.

In this section, we first explain the mediation effect of legitimacy on REs' performance in innovation. Then, we discuss how the strategic resource of REs' local and overseas knowledge moderate this mediation relationship.

A business is surrounded by a set of stakeholders including customers, suppliers, competitors, financing organizations, intermediary institutions, and universities/research organizations (Cantù et al., 2015; Doloreux, 2004; Nieto and Santamaría, 2007); and their interactions create value for the business. In a knowledge-based economy, the interactions and relations amongst various stakeholders can affect or are affected by an organization's objectives and achievements (Freeman, 1984) and are important for a firm's performance (Parmar et al., 2010). Studies advocate that the higher stakeholders' inclusion in business results, the higher the company's performance (Partanen et al., 2008; Trunina et al., 2020).

Stakeholder engagement improves shared expectations between the firm owner and interested stakeholders (Navis and Glynn, 2010), which drives legitimacy. Stakeholders' perceptions of an organization's ability to create value relative to competitors provide desirability, acceptance, and appropriateness (Rindova et al., 2005). We explicitly acknowledge perceived legitimacy as a measurable construct (Alexiou and Wiggins, 2018). Legitimacy is especially important for new ventures that face liability of newness, lack venture track records, and operate with limited resources (Zimmerman and Zeitz, 2002; Nahata, 2008; Partanen and Goel, 2017). To mitigate these limitations, new and small firms engage in low-cost legitimating activities, such as endorsements from relevant stakeholders (Zimmerman and Zeitz, 2002). In China, the same process is socially embedded (Granovetter, 1985): stakeholders do not rely only on formal claims, but also on known reputation and repeated relational behavior. Guanxi is therefore a local legitimacy channel through which loyalty, dedication, reciprocity and trust make a venture more acceptable to others (Ahlstrom et al., 2008; Burt and Burzynska, 2017; Alexiou and Wiggins, 2018).

Legitimacy is a key to unlocking access to other resources that are essential to the firm (Andrews, 1996). Entrepreneurs build legitimacy to access the resources for their business growth (Kawai et al., 2020). Although the legitimacy-performance association is complex, it contributes to a venture's financial performance (Roberts and Dowling, 2002) and promotes venture capital investment (Trunina et al., 2020).

Entrepreneurial enterprises, especially those established by REs, often face low reliability and credibility, which create thresholds for their products, services, and market entry. To address these limitations, REs can build legitimacy that helps them gain access to local resources in the home environment, which are important but difficult for returnees to obtain (Armanios et al., 2017).

Stakeholder engagement and culture are connected because stakeholder networks carry social influence (Carrington et al., 2005). Geography may also influence entrepreneurial performance because potential stakeholders are often geographically concentrated (Johannisson and Huse, 2000). China's institutional collectivism means that network ties carry a different meaning for Chinese entrepreneurs, with greater emphasis on personal-level relationships (Zheng et al., 2014). Guanxi, trusted personal ties and reciprocal obligations; renqing, informal social obligations; ganqing, emotional attachment between business partners; and mianzi, recognition of social status, are important social features of Chinese business life (Lin and Lin, 2016). Deals often grow out of stakeholder relationships in which actors exchange favors and support over time (Cai et al., 2010; Wiegel and Bamford, 2014).

Knowledge is not only an intellectual resource (Karadag et al., 2023) but also a cultural resource (Ren and Liu, 2019). That means local knowledge is associated with local culture, allowing returnees to be more understandable to local stakeholders. By gaining local knowledge, they become insiders (Liu et al., 2019) and thus more trustworthy. Trust is essential for high-level cultural environments (such as China) because it allows the enterprise to reliably get and utilize the resources (Lefebvre et al., 2016; Theodoraki et al., 2018), receive a higher growth in profit, sales, ROI, ROA, and international revenue (Chung and Tan, 2017; Wang et al., 2011a, b). On the way of RE to receive legitimacy from the local environment, we propose that local knowledge can amplify this process. Thus:

The knowledge acquired abroad by REs can be a valuable asset for recognizing opportunities and acquiring resources (Bai et al., 2017; Tran and Truong, 2022). However, such knowledge may also pose a challenge for REs seeking legitimacy through stakeholder engagement in the home-country entrepreneurial context, due to the institutional distance and fit between the host and home countries. (Kostova, 1999) suggests that the degree of similarity or difference between home- and host-country institutions affects the transferability and applicability of knowledge and practices across contexts. Consequently, overseas knowledge dissimilarity may not be perceived as beneficial to stakeholders' self-interest or societal welfare (Suchman, 1995). Besides, the type, quality, and relevance of knowledge can also negatively influence the impact of stakeholders' engagement and legitimacy. REs' overseas knowledge, obtained from developed environments, may require specific infrastructure and/or developed institutions that may not be suitable for local entrepreneurial conditions characterized by less developed institutions (Ren and Liu, 2019). Alternatively, REs, by being away from their home country (especially an emerging economy), may miss the various changes that have occurred in the society (Zhou and Li, 2010). This may affect stakeholders' engagement and their acceptance of whether the REs' venturing activities are the “right thing to do” for the society (Suchman, 1995). Therefore, REs overseas knowledge may weaken the positive impact of stakeholders’ engagement on legitimacy. We hypothesize:

When legitimacy from local stakeholders is obtained, it should impact REs' performance in innovation. Performance in innovation. We argue that REs overseas knowledge strengthens the impact of legitimacy on REs' performance in innovation. REs have a unique resource, i.e. overseas knowledge, differentiating them from local entrepreneurs (Wang et al., 2011a, b). Overseas knowledge may include both tacit and explicit knowledge from the host market, such as advanced technological knowledge, new business models, or best practices that REs can bring to the home country. Such knowledge allows for the support and greater legitimacy of REs for two reasons. First, gaining such knowledge is time-consuming and costly for local entrepreneurs. Local stakeholders were not in the context in which the knowledge was created and developed, which makes it difficult to obtain it (Levin and Barnard, 2013). Also, if they want to acquire such knowledge, they have to spend money and energy, which is costly. So, the perception among local stakeholders like governments, suppliers, and customers, aware of the REs' competitive advantage, may be enhanced to encourage their innovative venturing in the home country. Governments may provide specific conditions and rewards for REs to pursue innovation activities in the home country. In the case of China, favorable government policies have led to an increasing number of returnees from developed countries establishing new ventures in China (Li et al., 2012; Liu et al., 2018). Such support signals legitimacy for REs to do innovative activities in the eyes of different interest groups (Levie and Autio, 2011). Second, emerging economies with ambitions for economic development harness existing potential to exploit opportunities for further development. For instance, the perception of brain drain has changed to brain gain (Saxenian, 2005), thereby influencing China's innovative capability. Small and medium technology-based enterprises that returnees have created are now part of the new Chinese economy (Wang and Bao, 2015). Thus, stakeholders perceive that REs have unique overseas knowledge that enables them to create value for the economy. As a result, higher legitimation will be granted to REs innovative activities. Thus:

A study by Ma et al. (2019) of 226 REs who have started businesses in China finds that domestic (local) knowledge does not positively affect their venture performance. Local knowledge is important for Res, but their stakeholders also possess it. Therefore, utilizing local knowledge will not enhance these stakeholders' perception of an RE organization's ability to create value.

The research framework of this study is represented in Figure 1.

Returnee entrepreneurship is a major phenomenon observed in Beijing's Zhongguancun area. This talent flow contributes to technological and entrepreneurial gaps and supports China's economic development (Dai and Liu, 2009; Li et al., 2012). Compared with REs from India, whose companies are mainly focused on expanding exports of services and products (Pruthi, 2014), or Africa, where weak institutions affect technology-enabled returnee firms (Mreji and Barnard, 2021), Chinese REs are supported by several programs, including the Hundred Talents Program, Chunhui Scholar Program, and Project of Thousand Youth Talents. In 2016, ten immigration policies launched by the Ministry of Public Security were piloted to attract more overseas talents. By the end of 2016, more than 30,000 returnees were working in Zhongguancun (Liu et al., 2018). Until 2015, 21 science parks had been launched in Zhongguancun, specifically targeting REs (Wang and Bao, 2015). Given this context, our design considers social embeddedness when defining the population and interpreting responses (Granovetter, 1985), while measure adaptation also accounts for Guanxi norms in China (Ahlstrom et al., 2008; Burt and Burzynska, 2017).

To test the above hypotheses, we launched a questionnaire. To identify the respondents, we considered the definition of SME for the Chinese context accepted by the Ministry of Finance and China National Bureau for the technology enterprises (i.e. fewer than 2000 employees, or an annual turnover of less than or equal to RMB mln 300, or total assets of less than or equal to RMB mln 400).

We collected data in various ways: first, the questionnaire was sent directly by email to companies located in Zhongguancun Software Park, Changping Biomedical Base, Daxing Biomedical Base, Haidian Pioneer Park, Wangjing Science and Technology Park, and Haidian Venture Park. Second, the study questionnaires were distributed at local conferences related to returnees' entrepreneurship (e.g. the Zhongguancun Overseas Chinese Innovation and Entrepreneurship Conference). Lastly, personal connections were used to distribute the questionnaire. In total, 516 questionnaires were distributed, and 201 (38%) responses were obtained. In this study, we included 178 responses because all respondents provided their companies' names, which allowed us to manually verify their websites and select only firms working with cutting-edge technologies that apply complex technical solutions in IT, biomedicine, energy, and advanced manufacturing.

The survey items were designed using a 5-point Likert scale. Higher scores indicate stronger agreement. The language of the questionnaire was Chinese. In adapting items, we used committee translation and cultural review to ensure fit with social embeddedness (Granovetter, 1985) and Guanxi norms in China (Ahlstrom et al., 2008; Burt and Burzynska, 2017).

The validity of the questionnaire was tested by a group of five experts from related fields, including an executive at a technology enterprise, a consultant, scholars with expertise in related fields, and a government officer. Filled in the test version of the survey and provided feedback related to the quality of questions, their timing, and several recommendations for improvements.

Respondents are top-level managers (CEO, VP of business development, COO, partner, director, etc.) who were fully familiar with the development of the firm's technology and its financial dynamics. All respondents had spent at least 2 years abroad for study or work and had arrived from the USA (44%), Canada (6%), the EU (36%), and other Asian countries outside China (14%). In terms of ownership, 105 (94.6%) companies are Private Enterprises (PEs), 5 (4.5%) Foreign-Invested Enterprises (FIEs), and 1 (<1%) Collectively Run Enterprises (CREs). 46% of the RE-founded companies have at least one patent.

Table 1 presents the statistics for the measurement scales. The Cronbach alphas ranged from 0.825 to 0.925, proving that the scales of the constructs are reliable DeVellis (2003). Kaiser-Meyer-Olkin (KMO) statistic ranged from 0.747 to 0.909, indicating acceptable sampling adequacy, with a significant p-value for Bartlett's test. Additionally, we established discriminant and convergent validity for the questionnaire (extracted variances are greater than 0.6, with significant correlation coefficients). Reliability/validity for the added perceived legitimacy subscale are reported alongside other constructs in Table 1 (Alexiou and Wiggins, 2018).

We undertook several procedures to estimate the magnitude of common method bias. First, because data were gathered from a single respondent, common method variance was examined using Harman's single-factor test (MacKenzie and Podsakoff, 2012). A single factor extracted 29.4% of the total variance, which is less than 50%. Thus, there is no risk of common method bias. Second, we separated the dependent, independent, moderator, and control variables. This procedure helped to mitigate the risk of rationalizing the answers of the respondents (Podsakoff et al., 2003). In addition, we collected objective and subjective measures of a firm's number of submitted patent applications relative to the previous year. Such measures were used to estimate common method bias. The subjective and objective measures were significantly correlated (r = 0.227, p < 0.05) and yielded similar findings. We also emphasized anonymity and reduced evaluation apprehension at the outset; these procedural remedies align with the cultural salience of face/Guanxi (Ahlstrom et al., 2008; Burt and Burzynska, 2017).

We follow prior studies in measuring constructs (Table 1). For all multi-item constructs, we used 5-point Likert scales unless noted.

This paper draws on research on non-continuous innovation and disruptive technologies, including Chandy and Tellis (1998) and Christensen (2013). Five items were designed to measure the firm's innovation performance, characterizing performance in terms of products, technologies, and services relative to other local companies in their field.

Respondents were asked to indicate the collaboration of their firms with external stakeholders that brings values for their business – customers, suppliers, competitors, financing organizations, intermediary institutions, and universities/research organizations (Cantù et al., 2015; Doloreux, 2004; Nieto and Santamaría, 2007).

Since legitimacy has no material form and exists as a symbolic representation of stakeholders' collective evaluation, its value can only be assessed indirectly. To estimate company legitimacy, we used four items utilized before (Elsbach, 1994; Partanen and Goel, 2017; Trunina et al., 2020). The four items assess external legitimacy by examining how stakeholders evaluate the firm's capabilities and expertise. To validate individual subjective measures, we collected data from company publications and websites. We counted the number of mentions receiSMEs obtained from suppliers, customers, financial organizations, and professional associations regarding their achievements. The secondary data were collected for RE companies that provided their names in the questionnaire. The number of citations was correlated with survey based legitimacy (r = 0.51, n = 42, p < 0.05), validating the study's measures. To capture perceived legitimacy, we also used a brief 4-item Likert subscale tapping loyalty, dedication, reciprocity, and trust adapted to the RE China context (Alexiou and Wiggins, 2018). Scores were averaged, and higher values indicate stronger perceived legitimacy.

The overseas knowledge variable is derived from Chandler and Lyon (2009) and Tran and Truong (2022) and is operationalized with three items. The local knowledge variable is derived from Lumpkin and Lichtenstein (2005), Chandler and Lyon (2009), and Tran and Truong (2022) and is operationalized with four items. Both knowledge constructs are theorized as moderators in our framework. Their inclusion reflects the social embeddedness of knowledge use (Granovetter, 1985) and the local role of Guanxi in China (Ahlstrom et al., 2008; Burt and Burzynska, 2017).

We control for firm size and firm age variables because younger and smaller firms can suffer more from the liabilities of newness and smallness (Stinchcombe, 2000). Also, the number of employees is recognized as a determinant of financial performance. Firm size was the natural logarithm of the total number of full-time employees. Firm age was the natural logarithm of the number of years since the firm's establishment. At the firm level, we included a dummy variable Industry. For this research, we involve the IT firms because it is a core industry sector for Zhongguancun hub (Dong et al., 2019). We employ dummy variables to indicate whether the firm received any form of financial support from the Chinese government (e.g. R&D subsidies, funding from the Innofund program) and whether the company has a business abroad. The period during which RE lived abroad was also included and calculated as the natural logarithm of the number of years spent in the host country.

Table 2 provides descriptive statistics and correlations among the variables used in the regression analyses. There is a positive and significant correlation between the RE firm's innovation performance and its age, number of years spent abroad, local network, legitimacy, and both types of knowledge. The effects of the firm's size and government support of RE firms were found to be statistically insignificant.

Primary data was analyzed using SPSS 22.0. To demonstrate the findings clearly, we provide a conceptual results-map in Figure 3 that summarizes the supported paths and interactions. Table 3 represents the results of the hierarchical OLS regression analysis. Variables were entered iteratively, starting with the control variables, estimating the effects of the company's size, age, industry type, financial support from the government, time spent abroad, and business abroad on Res' performance in innovation (Table 3, Model 1). All reported coefficients are standardized (β). Models 2 and 3 include the effects of the moderating variables. Overseas and local knowledge are positively related to REs' innovation performance (Model 2: β = 0.423; Model 3: β = 0.374, p < 0.01), which confirms the results of previous studies (e.g. Liu et al., 2019). Models 4, 5, 6, and 9 are intended to assess the mediation effect using Baron and Kenny's criteria. According to this approach, mediation is established if: (1) the independent variable (stakeholder engagement) is significantly related to the dependent and the mediating variables; (2) the mediator (legitimacy) is significantly related to the dependent variable, and (3) the influence of the independent variable on the dependent variable is attenuated when the mediating variable is included in the regression equation (Baron and Kenny, 1986). Therefore, Model 4 and Model 9 check the main effect of local stakeholder engagement on Res' performance in innovation and legitimacy, which showed that these effects are positive and statistically significant (β = 0.365, β = 0.302, p < 0.01). Thus, the results satisfy the first condition for mediation. Model 5 presents the effect of the mediator on the dependent variable. Results show that the influence of legitimacy on Res' innovation performance (β = 0.565, p < 0.01) is positive and statistically significant. The second criterion is also fulfilled. When the mediator is introduced into the regression equation (Model 6), the effect of stakeholder engagement is no longer significant (β = 0.168; p > 0.1), confirming the third requirement for mediation. The results map in Figure 3 reflects this pattern (from local stakeholder engagement to legitimacy to innovation performance; the direct path is attenuated when legitimacy enters). Therefore, our results suggest that collaboration with local stakeholders affects Res' innovation performance, and that this effect is mediated by the firm's legitimacy. These findings confirm Hypothesis H1.

Hypothesis H2 predicts that the higher the level of local knowledge, the more it amplifies the association between local network and legitimacy. In Model 11, the interaction effect is positive (β = 0.411) and statistically significant (p < 0.01), suggesting that the moderating effect of local knowledge on the relationship between local network and legitimacy increases as the level of local knowledge increases. Therefore, Hypothesis H2 is supported. In Model 10, the interaction effect between overseas knowledge and engagement with the local stakeholders is negative (β = −0.223) and statistically significant (p < 0.01). Therefore, Hypothesis H3 is supported. Model 7 tests the moderation effect proposed in Hypothesis H4. The results show that the coefficient of the interaction between the firm's legitimacy and RE's overseas knowledge is positive and statistically significant, related to the firm's innovation capability (β = 0.282, p < 0.05). Thus, Hypothesis H4 is supported. Additionally, following (Aiken et al., 1991), we performed a slope test showing that the effect of legitimacy on Res' performance in innovation is positive. Figure 2 shows the relationship between legitimacy and performance in innovation is stronger for firms with RE utilizing more overseas knowledge. In Model 8, the interaction effect between local knowledge and legitimacy is negative (β = −0.229) and statistically significant (p < 0.01). Therefore, Hypothesis H5 is supported. These four interactions (local stakeholder engagement × local knowledge; local stakeholder engagement × overseas knowledge; legitimacy × overseas knowledge; legitimacy × local knowledge) are also highlighted in Figure 3 for interpretability.

The coefficients of determination (R-squared), Adjusted R-squared, and F-value in the models indicate acceptable goodness of fit. Variance inflation factor (VIF) values were assessed to assess multicollinearity among variables in the models (Table 3). Results showed no indication of multicollinearity as the highest mean VIF was well below the suggested threshold value of 10 (Hair et al., 2013).

We estimated the robustness of our research analysis by performing three additional analyses.

First, the Sobel test (Sobel, 1982) was performed (Table 4) to investigate the significance of the mediation effects, providing additional support for the hypothesized mediated relationships by assessing changes in the significance of the indirect effect. It shows that legitimacy mediates local stakeholders' engagement and performance in the innovation of the RE-founded firm (32.6% of the total effect of innovation ability is attributable to the mediation variables).

Second, given that the Sobel test can skew results due to the small sample size (Preacher and Hayes, 2004), we used a bootstrapping technique to confirm the mediation effect, thereby enhancing the robustness of the results (Table 5). This analysis supports the proposed mediating effect. Mainly, the results indicate that the indirect effect falls within the 95% confidence interval, supporting the presence of mediation (Preacher and Hayes, 2004).

Third, we established the robustness of our research model by employing structural equation modeling (SEM) in AMOS23 to retest our hypotheses. Results using the SEM method support the mediating role of legitimacy and the moderating effect of acquired local and transferred overseas knowledge. The core of verifying the theoretical model with SEM is the model fit (Byrne, 2010). Fit indices of the causal model are: w2/df = 2.21, RMSEA (root mean square error of approximation) = 0.078, NFI (normed fit index) = 0.91, NNFI (non-normed NFI) = 0.91, GFI (goodness-of-fit index) = 0.84, AGFI (adjusted goodness-of-fit index) = 0.81, and CFI (comparative fit index) = 0.93. All of them are below the thresholds. Thus, the empirical results from SEM are consistent with our initial findings.

The aim of this paper was to investigate how REs' overseas and local knowledge shapes the process through which local stakeholder engagement builds legitimacy and improves innovation performance. The findings demonstrate the positive mediating role of legitimacy in the relationship between local stakeholder engagement and innovation performance, consistent with prior work on reputation, venture capital and performance (Roberts and Dowling, 2002; Trunina et al., 2020). In China, this mediation is socially embedded (Granovetter, 1985) and channeled through Guanxi ties (Ahlstrom et al., 2008; Burt and Burzynska, 2017). The point is not simply that relationships matter. Rather, local stakeholders are more likely to support REs when overseas knowledge is attached to a known and accepted reputation. That is why stakeholder engagement carries heightened weight for acceptance in this setting. RE ventures need recognition from suppliers, customers, government officials, financial institutions, and scientific institutions because such recognition lends credibility and consistency to their innovation activities (Kawai et al., 2020; Trunina et al., 2020).

As a unique phenomenon, REs are influenced by dual environments (Wang et al., 2011a, b). Applying the Knowledge-Based View (Kogut and Zander, 1992; Kaplan et al., 2001), we explained that overseas and local knowledge affect RE firms differently. Our findings show an asymmetric pattern: local knowledge strengthens the stakeholder engagement-to-legitimacy link yet can dilute the legitimacy-to-innovation payoff through redundancy; overseas knowledge weakens the engagement-to-legitimacy link when fit is uncertain, but strengthens the legitimacy-to-innovation link once acceptance has been gained. Overseas knowledge is valuable for firm performance (Levin and Barnard, 2013; Bai et al., 2017; Tran and Truong, 2022), but it may not be immediately applicable to local entrepreneurial environments with different institutional contexts (Ren and Liu, 2019). These patterns fit embeddedness logic, where the value of economic action depends on the relational setting in which it is judged (Granovetter, 1985), and they fit Chinese evidence that unfamiliar practices require relational translation before they are accepted (Ahlstrom et al., 2008; Burt and Burzynska, 2017; Ren and Liu, 2019). Proper application of local and overseas knowledge is therefore critical for innovation performance. Their simultaneous use adds value when overseas know-how is locally embedded through Guanxi-mediated ties.

The proper application of local and overseas knowledge to gain intangible resources, such as legitimacy, can enhance innovation performance. Particularly, our results showed that although accumulated knowledge REs received abroad plays an important role in identifying opportunities and obtaining resources (Bai et al., 2017; Liu et al., 2015), in the home entrepreneurial context, such knowledge can easily lock REs in the trajectory of the past and decrease the engagement with local stakeholders, thus reducing the firm's legitimacy.

A lack of necessary local knowledge in China hinders REs from developing the entrepreneurial abilities needed for effective collaboration with external stakeholders in the home country. Local knowledge helps REs align with community norms, understand institutional and political changes, and reduce the suspicion attached to unfamiliar practices (Liu, 2020; Tran and Truong, 2022). Our findings are consistent with Tran and Truong (2022) and show that higher levels of local knowledge reinforce the effect of stakeholder engagement on legitimacy. The Chinese evidence adds a sharper point: stakeholder engagement matters more where legitimacy is socially embedded (Granovetter, 1985), and overseas knowledge yields innovation gains after a locally accepted reputation has been built (Liu, 2020).

The research has three theoretical contributions. First, it clarifies the core constructs, namely local stakeholder engagement, legitimacy, innovation performance, local knowledge, and overseas knowledge. Legitimacy is the mediator, while the knowledge variables are moderators (Salancik, 1977; Delmar and Shane, 2004; Fisher et al., 2016; Alexiou and Wiggins, 2018). This research contributes to the field of returnee entrepreneurship by explaining why REs' knowledge advantages do not automatically translate into innovation. They become effective when stakeholders interpret the venture as credible and appropriate, a process tied to perceived legitimacy and known reputation (Alexiou and Wiggins, 2018).

Second, this research contributes to the Knowledge-Based View by contextualizing knowledge in returnee entrepreneurship. Knowledge is a double-edged resource: local and overseas knowledge can support venturing, but their effects depend on where and how they are applied. In the venturing process, REs must decide which knowledge to foreground at each stage. This aligns with organizing activities and commitment in legitimation (Salancik, 1977; Delmar and Shane, 2004) and with identity legitimacy dynamics across venture life cycles (Fisher et al., 2016). Third, we explain when and how overseas knowledge must be recontextualized through socially embedded ties (Granovetter, 1985) and China-specific Guanxi mechanisms (Ahlstrom et al., 2008; Burt and Burzynska, 2017).

Returnee entrepreneurs in China must navigate the trade-offs and synergies between overseas acquired knowledge and local market realities. Their international experience allows them to introduce innovative products, services and processes that may be valued by investors, customers and regulators, but success depends on integration with local business networks. To gain legitimacy and optimize performance, REs must actively adapt foreign expertise to local knowledge. Engagement with suppliers, business partners and mentors is critical for market insight, regulatory updates and industry resources. Without this adaptation, REs may struggle to build trust and long-term collaboration. Because REs can be perceived as a threat to established norms and protocols, managers should build a clear legitimacy playbook: secure local sponsor endorsements, publish visible commitments and stage organizing activities that signal alignment with local expectations (Salancik, 1977; Delmar and Shane, 2004; Ahlstrom et al., 2008). They should also monitor perceived legitimacy through loyalty, dedication, reciprocity and trust to detect soft resistance early (Alexiou and Wiggins, 2018).

Policymakers should move beyond general support mechanisms and implement targeted interventions that directly facilitate REs' market integration. This includes developing specialized intermediary organizations that provide structured access to domestic resources (Armanios et al., 2017), offering adaptation and reacculturation training (Liu and Almor, 2016; Ma et al., 2019), and establishing platforms that bridge REs with key local stakeholders. To be effective, these platforms should include government-backed incubators, industry roundtables, and digital knowledge-sharing networks that enable REs to apply overseas expertise within China's evolving business landscape. Embedding Guanxi brokers and local anchor partners into these platforms can reduce perceived threat and accelerate acceptance (Ahlstrom et al., 2008; Burt and Burzynska, 2017). Institutionalizing these mechanisms can enhance REs' access to tangible resources and critical legitimacy, ultimately strengthening China's innovation ecosystem.

First, while local stakeholder engagement positively improves firm legitimacy, reverse causality may occur. Higher legitimacy may attract more stakeholder engagement because association with a credible venture can bring reputation or other interests. The same applies to innovation performance: successful innovation can strengthen legitimacy. Future studies should test reverse causality to make the model more robust. Second, the paper analyzes only RE-owned technology-based SMEs located in Zhongguancun, Beijing, whose founders recently returned from abroad. Given the importance of REs in emerging economies such as India and Mexico, future research should test this model comparatively across countries experiencing growth in returnees (Pruthi, 2014; Hagan and Wassink, 2016). Including social embeddedness measures (Granovetter, 1985) and Guanxi measures (Ahlstrom et al., 2008; Burt and Burzynska, 2017) would allow stronger cross-context tests of culturally contingent mechanisms.

Last constraint in this research lies in its inability to distinguish among distinct categories of stakeholders. Throughout the developmental stages of enterprises, they engage with diverse stakeholders, including suppliers, shareholders, employees, and customers. It is crucial to acknowledge that the interests of various stakeholders will have different influences on a company's innovation performance. Therefore, it is recommended that future studies conduct a more nuanced analysis by differentiating among these stakeholder categories, thereby providing a more comprehensive understanding of their respective impacts on innovation in the corporate context. Finally, to address measurement scope, future work should develop and validate a short perceived legitimacy Likert scale (e.g. loyalty, dedication, reciprocity, trust) adapted to RE–China, and test its convergence with external legitimacy indicators (Alexiou and Wiggins, 2018).

This study demonstrates that legitimacy is the critical mechanism through which returnee entrepreneurs translate local stakeholder engagement into measurable innovation performance. The analysis reveals an asymmetric relationship in which local and overseas knowledge exert distinct effects. Local knowledge enhances legitimacy formation by aligning entrepreneurial practices with local stakeholders' expectations and by helping REs build an accepted reputation. Overseas knowledge may initially weaken the effect of engagement on legitimacy when domestic stakeholders view foreign practices as unfamiliar, yet it reinforces innovation performance once those practices are attached to legitimized operations.

The findings therefore clarify a precise dual pathway: local knowledge intensifies stakeholder-driven legitimacy, and overseas knowledge amplifies the innovation value of legitimacy once acceptance has been secured. Collectively, these results show that innovation in returnee-led technology ventures depends not only on the stock of knowledge REs bring home, but also on whether that knowledge is relationally embedded and locally recognized. For Chinese REs, global expertise is most powerful when translated into a known and trusted reputation.

The authors thank the reviewers for their thoughtful and constructive comments, which improved the quality and clarity of the manuscript.