Introduction
When women lead firms and hold top CEO positions, firms tend to try fewer mergers and acquisitions (hereafter, acquisitions) (e.g., Chen, Crossland, & Huang, Reference Chen, Crossland and Huang2016; Huang & Kisgen, Reference Huang and Kisgen2013), to produce more conservative financial reporting (Ho, Li, Tam, & Zhang, Reference Ho, Li, Tam and Zhang2015) and more conservative balance sheets (Palvia, Vähämaa, & Vähämaa, Reference Palvia, Vähämaa and Vähämaa2015). Indeed, a meta-analysis concluded that higher female upper-echelon representation is negatively related to strategic risk-taking, as measured by financial leverage, capital expenditures, and stock return volatility (Jeong & Harrison, Reference Jeong and Harrison2017). The role incongruence perspective (Eagly & Karau, Reference Eagly and Karau2002) largely explains that female leaders avoid pursuing risky actions because society expects them to follow communal gender roles, which are incongruent with agentic leadership roles, causing women to fear intense criticism if risky strategies fail (Gupta, Han, Mortal, Silveri, & Turban, Reference Gupta, Han, Mortal, Silveri and Turban2018; Post, Lokshin, & Boone, Reference Post, Lokshin and Boone2022).
In contrast, some argue that female executives have higher risk tolerance levels. For example, a survey-based study demonstrated that in Sweden, female executives are more risk-taking than their male counterparts (Adams & Funk, Reference Adams and Funk2012). A study using a sample of Chinese firms found that female-led rather than male-led firms demonstrated higher levels of corporate risk-taking (Farag & Mallin, Reference Farag and Mallin2018). Those mixed findings triggered calls to specify research contexts that may influence gendered behavioral outcomes (Jeong & Harrison, Reference Jeong and Harrison2017; Mah, Kolev, McNamara, Pan, & Devers, Reference Mah, Kolev, McNamara, Pan and Devers2023). For instance, with the findings that scrutiny influences leader cognition and moderates the relationship between CEO gender and corporate acquisitions, Gamache, Devers, Klein, and Hannigan (Reference Gamache, Devers, Klein and Hannigan2023: 23) encouraged ‘considering additional contexts that may differentially shape male and female CEOs' information processing and their decisions and outcomes’ (Gamache et al., Reference Gamache, Devers, Klein and Hannigan2023; Jeong & Harrison, Reference Jeong and Harrison2017).
Thus, role incongruence theory predominantly asserts that females avoid risky activities because of gender bias against women. However, the theory fails to explain the mixed findings about female leadership. As an alternative theoretical mechanism, we introduce the expectancy violation theory (Jussim, Coleman, & Lerch, Reference Jussim, Coleman and Lerch1987) to consider a novel theoretical context. This theory posits that in certain circumstances, females may be rewarded rather than punished for violating social gender role expectations, forming anticipated female agentic advantages. A female agentic advantage refers to the gender premium whereby female leaders are rewarded favorable evaluations for displaying male-typed agentic traits (Ma, Rosette, & Koval, Reference Ma, Rosette and Koval2022). While a growing body of research acknowledges female agentic advantage elicited by expectancy violation mechanism (e.g., Lanaj & Hollenbeck, Reference Lanaj and Hollenbeck2015; Schaumberg & Flynn, Reference Schaumberg and Flynn2017), few have explored how agentic advantage affects firm activities, prompting the research question: What circumstances might encourage female leaders to engage in the risk-taking actions that are typically attributed to males?
We address the question by examining corporate acquisitions, a risk-taking activity that is mostly decided by firm chairs, in the context of Chinese state-owned enterprises (SOEs), where the government is the major stakeholder and evaluator of top firm executives. According to expectancy violation theory, we argue that acquisitions are viewed as typical male-typed risk-taking activities (Eagly & Wood, Reference Eagly and Wood1999), but female chairs might be encouraged to engage in such masculine activity when they receive positive governmental evaluations that elicit female agentic advantages.
To understand the boundary conditions of expectation violation predictions, we introduce female executive representation and political appointments held by female chairs as moderating the relationship. The psychology literature acknowledges that when self-integrity is threatened, individuals seek to protect or restore their self-regard or self-worth (Steele, Reference Steele and Berkowitz1988). We propose that self-affirmation depends on both external situational cues and internal personal attributes. Lower female representation in executives is recognized as an external circumstance that creates higher stereotype threats toward female chairs; the higher an individual encounters threats, the stronger a desire to react against them (Steele, Reference Steele and Berkowitz1988). Political appointments, often viewed as an official recognition or reward for executives in Chinese companies, bolster female chairs' self-efficacy to tackle the negative gender stereotype (Bandura, Reference Bandura1986); the higher an individual's self-efficacy, the stronger a desire to take action to advance one's affirmation (Steele, Reference Steele and Berkowitz1988). We argue that female chairs react to the two circumstances with strong motivations to disconfirm gender stereotypes, thus pursue acquisition activities. We find support for our hypotheses in a sample of listed Chinese SOEs from 2008 to 2020. The results remain robust after using the propensity score matching (PSM) approach and difference-in-differences (DID) analysis on a subsample of firms experiencing exogenous chair successions.
Our study contributes to the female leadership literature by joining increasing challenges to the role incongruence perspective (Gamache et al., Reference Gamache, Devers, Klein and Hannigan2023; Jeong & Harrison, Reference Jeong and Harrison2017) explaining the negative relationship. We introduce the expectancy violation theory as an alternative lens to explain the positive effects of female leadership on firm risk-taking activities. By contextualizing Chinese SOEs, our study complements an opposite view of gender differences in risk-taking demonstrated in previous studies (Gamache et al., Reference Gamache, Devers, Klein and Hannigan2023; Jeong & Harrison, Reference Jeong and Harrison2017). Additionally, we contribute to the literature on SOEs (Bruton, Peng, Ahlstrom, Stan, & Xu, Reference Bruton, Peng, Ahlstrom, Stan and Xu2015), which needs delicate theoretical and empirical research because SOEs are often compared to non-SOEs without considering that SOEs are subtly heterogeneous (e.g., Ghorbani, Xie, Jin, & Wang, Reference Ghorbani, Xie, Jin and Wang2023; Li, Li & Wang, Reference Li, Li and Wang2019). We enhance the empirical research and offer a more nuanced understanding of a unique yet increasingly important context by examining how gender affects acquisitions. Finally, we contribute to the acquisition literature (Haleblian, Devers, McNamara, Carpenter, & Davison, Reference Haleblian, Devers, McNamara, Carpenter and Davison2009) by revealing that female executives use acquisitions as a means for confronting gender stereotypes.
Theoretical Background
Leader Gender and Firm Risk-Taking
Aligned with the pervasive role incongruence perspective, female leaders have been associated with low levels of firm risk-taking (Eagly & Karau, Reference Eagly and Karau2002) because they are more harshly evaluated and scrutinized for acting incongruently with gender role expectations. That is, males are expected to display agentic traits (e.g., assertiveness, dominance, and confidence), while females are expected to display communal traits (e.g., warmth, likeability, and selflessness). Female CEOs tend to draw investor skepticism upon their appointments (Lee & James, Reference Lee and James2007), earn less than male CEOs (Wang, Markóczy, Sun, & Peng, Reference Wang, Markóczy, Sun and Peng2019), and tend to be blamed for poor performance and strategic failures (Gupta et al., Reference Gupta, Han, Mortal, Silveri and Turban2018; Park & Westphal, Reference Park and Westphal2013). Aware of these gender biases, female leaders tend to take more conservative strategic actions to avoid potential negative career outcomes associated with underperformance (Post et al., Reference Post, Lokshin and Boone2022). For instance, they may avoid strategic change (Sidhu, Feng, Volberda, & Van Den Bosch, Reference Sidhu, Feng, Volberda and Van Den Bosch2021) and acquisitions (Post et al., Reference Post, Lokshin and Boone2022).
The role incongruence perspective is important for explaining gender differences in firm risk-taking activities, but it fails to fully explain emerging mixed findings. Expectancy violation theory (Jussim et al., Reference Jussim, Coleman and Lerch1987) offers an alternative lens, suggesting that female leaders may have gender advantages for risk-taking behaviors in certain circumstances. That is, when females show agentic behaviors that are incongruent with stereotyped gender-based expectations, they may elicit more favorable evaluations because they exceed expectations (Jussim et al., Reference Jussim, Coleman and Lerch1987). In Chinese public companies, for example, female CEOs have mitigated the compensation gap by taking risks (Wang et al., Reference Wang, Markóczy, Sun and Peng2019). Women leaders who possess female agentic advantage may show risk behaviors to counteract gender-based bias and garner positive evaluations. Indeed, some top-level women selectively display overly assertiveness and aggressiveness to be considered successful leaders (Trzebiatowski, McCluney, & Hernandez, Reference Trzebiatowski, McCluney and Hernandez2023; Vial, Napier, & Brescoll, Reference Vial, Napier and Brescoll2016).
In summary, expectancy violation theory diverges from role incongruence theory in predicting female agentic disadvantages and advantages, leading to contrasting predictions about leader gender influences on firm risk-taking. Which theory serves as an explanatory framework for the behavioral strategies of female leaders hinges on the evaluators' relationship with the firm's leadership (Joshi, Oh, & DesJardine, Reference Joshi, Oh and DesJardine2024) that shapes their motivation to observe and evaluate female leaders. In a context where evaluators lack the desire to notice female executives, they often lean toward gender-biased negative stereotypes (Neuberg & Fiske, Reference Neuberg and Fiske1987), consistent with the widely accepted role incongruence mechanism.
In contrast, when evaluators have the desire to notice female executives, but are at a remote distance and close observation is unfeasible, evaluators use availability heuristics to process vivid information in making decisions (Tversky & Kahneman, Reference Tversky and Kahneman1973). They might use behaviors that violate gender stereotypes (Macrae, Bodenhausen, & Milne, Reference Macrae, Bodenhausen and Milne1995) as a type of vivid information for assessing the capabilities of female leaders. Such contexts are more likely to trigger the expectancy violation mechanism, generate positive and favorable evaluations of counter-stereotypical leadership behaviors (Jussim et al., Reference Jussim, Coleman and Lerch1987), and convey female agentic advantage.
Acquisitions as Firm Risk-Taking Activities in Chinese SOEs
In a context characterized by female agentic advantage, remote external stakeholders are likely to favorably evaluate and reinforce risk-taking behaviors in female leaders. Female-led Chinese SOEs provide such a context in which women gain agentic advantage through counter-stereotypical behaviors. The Chinese government is aware that women are more underrepresented in the upper echelons of SOEs than in non-SOEs (Zhang, Reference Zhang2012). In response, the government is promoting gender diversity according to the ‘Outline for Women's Development in China (2011–2020)'. For instance, all governmental levels and branches must include at least one female leader. Driven by the rarity of female leaders and by goals for gender diversity, the government pays special attention to incumbent female leaders.
Although the Chinese government or its State-owned Assets Supervision and Administrative Commission (SASAC) is the primary stakeholder and evaluator of SOEs, they lack access to detailed information about SOEs and their executives. To counter the remoteness and inability to closely observe firm executives, they may rely on heuristically vivid information to judge executive behaviors (Tversky & Kahneman, Reference Tversky and Kahneman1973). Corporate acquisitions, a risk-taking and counter-stereotypical behavior for female firm leaders, are such a type of information. When acquisitions are considered favorable activities, governmental authorities will positively evaluate female leaders. The expectancy violation mechanism then elicits agentic advantage.
In Chinese SOEs, female leaders are more likely to be rewarded rather than penalized for risk-taking acquisitions. First, the Chinese government desires, encourages, expects, and positively evaluates acquisitions as a primary means for expanding and reforming SOEs. Since the SASAC was established in 2003, the Chinese government has actively encouraged market-oriented practices, most notably mergers and acquisitions, to foster globally competitive SOEs (Hassard, Morris, Sheehan, & Yuxin, Reference Hassard, Morris, Sheehan and Yuxin2010).
Second, acquisitions are a risky strategic activity that aligns with leadership agency and leadership role expectations (Chng & Wang, Reference Chng and Wang2016). While acquisitions may render the firm leader at high risk of being punished for potential economic loss (Haleblian et al., Reference Haleblian, Devers, McNamara, Carpenter and Davison2009), SOEs are regulated to pursue social and political goals, such as social stability (Wang & Luo, Reference Wang and Luo2019) rather than to purely focus on achieving market success through acquisitions (Ghorbani et al., Reference Ghorbani, Xie, Jin and Wang2023). For example, some SOEs initiate acquisitions to acquire strategic assets and provide employment, which leads to governmental praise (Buckley, Yu, Liu, Munjal, & Tao, Reference Buckley, Yu, Liu, Munjal and Tao2016; Wang & Luo, Reference Wang and Luo2019).
In summary, governmental agencies tend to favor and positively evaluate female leaders who undertake acquisition activities. Thus, counter-stereotypical acquisition behaviors may elicit positive agentic advantages rather than unfavorable role incongruence evaluations (Jussim et al., Reference Jussim, Coleman and Lerch1987).
Hypotheses Development
Leader Gender and Corporate Acquisitions in Chinese SOEs
In Chinese SOEs, female leaders may seek positive governmental evaluations by leveraging agentic advantage. First, through SASAC, the government is the ultimate principal of all Chinese SOEs, and thus a predominant stakeholder and evaluator of SOE CEOs, chairs, or executives. The agents can increase career prospects (Fan, Lau, & Young, Reference Fan, Lau and Young2007) through promotions to parent firms, larger-sized or better SOEs, or through political promotions to full-time or part-time government sector positions (Cao, Lemmon, Pan, Qian, & Tian, Reference Cao, Lemmon, Pan, Qian and Tian2019). In general, Chinese SOE executive career advancements depend on governmental appointments that will be more lucrative than the inadequate monetary incentives derived from the national compensation policy, which has long been regulated and scrutinized to suppress pay inequality (Raynard, Lu, & Jing, Reference Raynard, Lu and Jing2020). Lacking pay-for-performance (Conyon & He, Reference Conyon and He2011; Ke, Rui, &Yu, Reference Ke, Rui and Yu2012) and receiving lower-than-market monetary compensation, SOE executives seek governmental recognition or political promotion that brings advanced social status, prestige, and power. Thus, the government is the most critical evaluator for SOE executives.
Given that the governmental agency has such a pivotal evaluative role, SOE executives will actively anticipate and comply with their evaluation preferences (Wang & Luo, Reference Wang and Luo2019). However, compared with male executives, females encounter more gender-based obstacles. They are less represented in SOEs than they are in non-SOEs (Zhang, Reference Zhang2012), and face higher challenges in breaking the glass ceiling (Kuhn & Shen, Reference Kuhn and Shen2013; Leung, Reference Leung2002). Male dominance in the bureaucratic system further exacerbates sex discrimination (Leung, Reference Leung2002). To counter disadvantages, female executives in SOEs utilize agentic advantage through government-desired counter-stereotypical acquisitions.
By demonstrating counter-stereotypical traits and violating social role expectations, female leaders convey heuristic clues indicating competence and reliability beyond initial governmental expectations. In contrast, male leaders fail to acquire credit for equivalent acquisition activities because risky acquisitions provide information congruent with male stereotypes. Consequently, remote and external government stakeholders are more likely to reinforce agentic behaviors of female rather than male leaders through favorable evaluations. Therefore, female leaders are likely to engage in acquisitions, leading to our first hypothesis:
Hypothesis 1 (H1): In SOEs, female leaders are more likely than male leaders to undertake acquisition activities.
Contingency Conditions of Expectancy Violation Theory
We have argued that anticipated governmental rewards motivate SOE female leaders to demonstrate counter-stereotypical traits that violate social expectations for women, mainly by engaging in acquisitions. Although female leaders generally face gender stereotype threats, they vary in the intensity of their motivations to tackle threats. Self-affirmation theory (Steele, Reference Steele and Berkowitz1988) explains that individuals are more strongly motivated to disprove stereotypes when situational cues imply a higher threat toward their self-regard or self-worth (Steele, Reference Steele and Berkowitz1988) and when they have higher levels of efficacy (Yeager et al., Reference Yeager, Purdie-Vaughns, Garcia, Apfel, Brzustoski, Master, Hessert, Williams and Cohen2014).
In this section, we thus identify two sources of self-affirmation, one external situational cue and one internal self-efficacy attribute, and explore how they may influence how leader gender relates to firm acquisitions in Chinese SOEs. Specifically, we examine the moderating effects of female executive representation regarded as gender stereotype threats and political appointments served as self-efficacy-enhancing governmental recognition.
Moderating effect of female executive representation
When women are in the minority, their presence tends to be more salient and subject to gender stereotypes (Kanter, Reference Kanter1977). Being in the minority but having a salient role is a situational cue that poses stereotype threats to identity (Richman, VanDellen, & Wood, Reference Richman, Vandellen and Wood2011) and a desire to restore or protect the individual's self-regard or self-worth (Steele, Reference Steele and Berkowitz1988). A supportive empirical study showed that when top management teams (TMTs) have fewer females, the workplace is more male-dominated and the gender-based pay gap is increased (Castellaneta, Conti, & Kacperczyk, Reference Castellaneta, Conti and Kacperczyk2020). Female leaders realize that they must work harder to counter the negative gender stereotypes and secure their careers. In other words, female executives who are acutely aware of gender stereotype threats adopt reactive career strategies to deal with widespread gender bias (Post et al., Reference Post, Lokshin and Boone2022).
Chinese SOEs are a typical microcosm of Chinese society featuring guanxi culture (Chen, Chen, & Huang, Reference Chen, Chen and Huang2013)Footnote 1 in which career advancement depends on job-related competence and guanxi strategies (Leung, Reference Leung2002). Women usually encounter obstacles in building male-dominated social networks (Abraham, Reference Abraham2020), even more so in Chinese SOEs where men dominate upper management and official government evaluator positions. Therefore, female executives in SOEs confront inherent gender-related obstacles in seeking a guanxi approach and fostering strong networks. Thus, most tend to demonstrate their competence by working hard and achieving tangible results (Liu, Reference Liu2013), particularly by using acquisition for achieving recognition (Haleblian et al., Reference Haleblian, Devers, McNamara, Carpenter and Davison2009).
Therefore, we argue that a smaller proportion of female incumbents in upper echelons activates the gender stereotype threat, highlighting the threat to female leaders' gender identity and causing great insecurity and suffering. As responsive career strategies, female leaders will pursue counter-stereotypical behaviors as a type of self- affirmation actions to tackle the gender biases, unsafety, and negative feelings, thereby protecting their self-worth (Steele, Reference Steele and Berkowitz1988). The more salient the gender stereotype threat, the more likely that female leaders will undertake acquisition activities to generate female agentic advantage through favorable government evaluations. Consequently, when an executive team has low female representation, the gender effect on SOE acquisitions proposed in Hypothesis 1 will be stronger.
Hypothesis 2 (H2): Female leaders will be more strongly and positively associated with acquisition activities when an executive team has low female representation.
Moderating effect of female leaders' political appointment
Major governmental stakeholders convey favorable evaluations to SOE executives by awarding them part-time roles in prestigious political councils or state organs (Li & Liang, Reference Li and Liang2015), as part-time delegates to the People's Congress (PC; China's legislative body) or to the Chinese People's Political Consultative Conference (CPPCC; China's political consultative body) (Li & Liang, Reference Li and Liang2015; Zhang, Marquis, & Qiao, Reference Zhang, Marquis and Qiao2016). Corporate leaders use the political appointments to communicate with and advise the government on important policies (Li & Liang, Reference Li and Liang2015). When female leaders earn political appointments, they are likely to be more motivated to disconfirm stereotypes because the positive governmental feedback boosts their self-efficacy and confidence about meeting challenges associated with gender stereotypes (Liu, Liu, Wang, & Zhang, Reference Liu, Liu, Wang and Zhang2021).
The government grants the business sector a strict quota for determining political appointments, which corporate leaders must earn through substantial effort and achievements. Therefore, political appointments indicate positive governmental evaluations of firm leadership (Cao et al., Reference Cao, Lemmon, Pan, Qian and Tian2019). Such positive feedback strongly motivates female leaders to disconfirm negative stereotypes and perform well (e.g., Yeager et al., Reference Yeager, Purdie-Vaughns, Garcia, Apfel, Brzustoski, Master, Hessert, Williams and Cohen2014). Their newfound self-efficacy generates cognitive dissonance (Festinger, Reference Festinger1957) that drives them to challenge negative stereotypes (Schmader, Johns, & Forbes, Reference Schmader, Johns and Forbes2008), and to feel more confident about proactively displaying counter-stereotypical behaviors, to restore their self-regards. We thus expect female leaders who have political appointments to show a stronger gender effect on SOE acquisitions.
Hypothesis 3 (H3): Female leaders who have political appointments will be more likely to undertake acquisition activities.
Methods
Sample and Data
Our sample includes all SOEs listed on Shanghai and Shenzhen stock exchanges in China from 2008 through 2020. We collected data from the China Stock Market and Accounting Research (CSMAR) database, which has been widely used in research on Chinese publicly listed companies. It provides credible information about companies' financial and background statistics, such as chair characteristics (e.g., Li & Lu, Reference Li and Lu2020; Zhang et al., Reference Zhang, Marquis and Qiao2016). After excluding (1) 490 firm-year observations in the finance industry (Chen, Luo, Tang, & Tong, Reference Chen, Luo, Tang and Tong2023) and (2) 2028 firm-year observations with missing data, our final sample included 11,112 firm-year observations with 1,265 SOEs. Two-sample t-tests revealed that the final sample had no significant differences from the excluded sample in critical attributes, including firm size and performance. We winsorized all the financial data at the 1% and 99% levels to minimize potential outlier influences.
Given that chairs are the most potent decision-makers and ultimate leaders in Chinese firms (Li & Liang, Reference Li and Liang2015; Zhang et al., Reference Zhang, Marquis and Qiao2016), we addressed the board chair in the SOEs. Data on the chair's political appointment were manually coded from chairs' biographical descriptions provided by CSMAR and company websites. We also crosschecked the names on the official websites of the two legislative organizations – PC and PPCC. Firm acquisition information was obtained from CSMAR's China Listed Firms' Merger & Acquisition, Asset Restructuring Research Database (Schweizer, Walker, & Zhang, Reference Schweizer, Walker and Zhang2019).
Dependent Variables
Acquisition activities
Following prior research (Shi, Zhang, & Hoskisson, Reference Shi, Zhang and Hoskisson2017), we measured acquisition activities as the annual number of acquisitions, operationalized as the number of acquisitions announced during a year, and coded 0 for no acquisition announcements. In the robustness check, to ensure that the sample included only meaningful acquisitions influencing political evaluations of chairs, we limited the sample to significant acquisitions with transaction values over 1 million yuan (Seo, Gamache, Devers, & Carpenter, Reference Seo, Gamache, Devers and Carpenter2015). The results remain unchanged.
Independent and Moderating Variables
Chair gender
Chair gender, the independent variable, was dichotomous, coded 1 for females and 0 for males.
Female executive representation
We measured female executive representation as the ratio of the number of female members (excluding board chair) to the total number of firm executives (including board directors, top management team, and supervisors) in a year. From female chairs' perspectives, lower numbers would indicate increased situational stereotype threat.
Chair political appointment
Following the political connection literature (e.g., Zhang et al., Reference Zhang, Marquis and Qiao2016), we measured political appointments by whether chairs served as part-time delegates to the National People's Congress (PC) or the Chinese People's Political Consultative Conference (PPCC), the two important political councils in China (Cao et al., Reference Cao, Lemmon, Pan, Qian and Tian2019; Li & Liang, Reference Li and Liang2015; Zhang et al., Reference Zhang, Marquis and Qiao2016), which have five hierarchical levels, from township to national. Higher hierarchy is associated with higher political status, so we coded political appointment as a continuous variable by considering hierarchical levels. Thus, the evaluation ranged from 0 to 5 for each chair, where 0 indicates no political appointment.
Control Variables
We controlled for several firm and chair characteristics. We included a comprehensive list of control variables that may affect corporate acquisitions as documented by prior studies. At the firm level, we controlled for firm size (log of total revenue), leverage (ratio of total debt to total assets), board size (log of total number of board directors), and board independence (the proportion of independent directors). Firm performance was controlled by including two variables: accounting performance (return on assets, ROA, calculated as the ratio of net income to total assets) and market performance (Tobin's Q, calculated as the ratio of total market capitalization to total assets). Administrative hierarchical distance from the central government influences whether chairs are willing to climb the political ladder (Hu & Xu, Reference Hu and Xu2022). Therefore, we controlled for administrative hierarchical distance by state ownership (the percentage of shares owned by the largest shareholder) and monitoring government (a dummy variable coded 1 when the central government is the largest shareholder and 0 for local government) (Wang, Wijen, & Heugens, Reference Wang, Wijen and Heugens2018).
Regarding chair-level characteristics, we controlled for chairs' bargaining power by including chair ownership (the proportion of chair-owned shares), chair tenure, and chair duality (a dummy variable coded 1 when a chair is also CEO). Compensation has been shown to motivate leaders to acquire (Seo et al., Reference Seo, Gamache, Devers and Carpenter2015), so we include chair compensation (log of cash compensation). The practice is consistent with all executive compensation studies in China because most Chinese firms lack long-term incentive plans (Wang et al., Reference Wang, Markóczy, Sun and Peng2019). In terms of education, SOE executives who have PhD degrees are more likely to be politically promoted (Yang, Wang, & Nie, Reference Yang, Wang and Nie2013). We therefore included a dummy variable for a PhD degree (coded 1 for a doctoral degree). We also controlled for chair age and average board age to account for potential decreases in risk-seeking with age. Finally, Chinese provinces vary in terms of monitoring capacity and governmental pressure, so we included fixed province effects. We also included fixed industry effects and fixed year effects in all models.
Results
We used negative binomial regression models for hypotheses testing because the dependent variable was a count with over-dispersion. Our results were robust to use Poisson regressions. In addition, we lagged our independent, moderating, and control variables by one year to minimize the possibility of reverse causality.
Table 1 presents the descriptive statistics and the Pearson correlations of the variables used in the regression models. Female chairs account for 3.4% of the sample. Chair gender is positively and significantly correlated with acquisition activities (corr. = 0.04, p < 0.001), implying that SOEs with female chairs announce more acquisitions on average. This correlation provides preliminary evidence. To detect multicollinearity concerns, we ran ordinary least squares (OLS) regressions to calculate each model's variance inflation factor (VIF). The highest VIF value is 2.89, which is under 10, but multicollinearity is still possible because chair gender has a 0.91 correlation and interaction with female executive representation (p < 0.001). Following executive gender studies (Gupta, Mortal, Chakrabarty, Guo, & Turban, Reference Gupta, Mortal, Chakrabarty, Guo and Turban2020), we dichotomized female executive representation such that the highest quintile is considered ‘high’ and the rest are considered ‘low'. Following Kalnins's (Reference Kalnins2008) recommendations about detecting and mitigating multicollinearity concerns, we reported interaction terms in the correlation table. As presented in Table 1, after the moderator was dichotomized, the highest correlation was reduced to 0.70. Most correlations are under 0.4, indicating unlikely multicollinearity.
Descriptive statistics and correlation
Notes: N = 11,112; Correlation coefficients are bold where p < 0.05.
Table 2 reports the negative binomial regression results for the effect of chair gender on corporate acquisitions. Model 1 includes high female executive representation, chair political appointment, and all control variables. The coefficient of high female executive representation was insignificant (β = −0.062, p = 0.155), but results for control variables were generally consistent with the literature (Chen et al., Reference Chen, Crossland and Huang2016). Model 2 of Table 2 includes the independent variable, chair gender, which was positively significant in predicting acquisition activities (β = 0.200, p = 0.030), suggesting that female rather than male SOE chairs engage in more acquisitions, supporting H1. In terms of economic magnitude, companies with female rather than male chairs announced 22% more acquisitions.Footnote 2
Negative binomial regressions predicting acquisition activities
Notes: robust standard errors in parentheses; †p < 0.10, *p < 0.05, **p < 0.01, ***p < 0.001; All tests are two-tailed.
Model 3 of Table 2 tests the moderating H2, which suggests that the gender effect in acquisition proposed in H1 is stronger under lower female executive representation. In Model 3, we included the interaction of chair gender and high female executive representation and found a negative and marginally significant coefficient of this interaction term (β = −0.335, p = 0.067). To evaluate the magnitude of the moderation effect, we calculated the marginal effects of chair gender using the STATA margins command, which is 0.131 (p = 0.123) under high female executive representation (1 SD above the mean). The marginal effect was 0.548 (p = 0.024) under low female executive representation (1 SD below the mean), indicating that female-led companies had 73% (=exp (0.548)−1) more acquisitions than male-led companies. Figure 1 plots the results of the two-way interaction (using the STATA marginsplot command). Figure 1(a) illustrates that under low levels of female executive representation, female chairs engaged in more acquisitions than male chairs did. However, the pattern reversed under high female executive representation. Moreover, the slope was −2.126 (p = 0.069) for female chairs and −0.013 (p = 0.930) for male chairs. The two slopes are significantly different (p = 0.073), suggesting a significant interaction. Taken together, the results support H2.
Interaction plots (a) Moderating effect of female executive representation (b) Moderating effect of chair political appointment
Model 4 of Table 2 tests H3 predicting that the gender effect in acquisition proposed in H1 is stronger among chairs with political appointment. Model 4 introduces the interaction between chair gender and chair political appointment. This interaction term had a positive and significant coefficient (β = 0.107, p = 0.017). Figure 1(b) shows that chair political appointments had a moderating effect. Specifically, under increased chair political appointment, female rather than male chairs showed a more pronounced pattern of acquisitions. When chair political appointment was 0, chair gender had a 0.022 marginal effect on the number of acquisitions (p = 0.808); however, when chair political appointment was 5, the marginal effect was 0.601 (p = 0.017) indicating SOEs with female chairs announced 82% (=exp (0.601)−1) more acquisitions than those with male chairs. Slope tests showed that the simple slope was 0.102 (p = 0.019) for female chairs and -0.000 (p = 0.993) for male chairs, a significant difference (p = 0.020). Taken together, the results support H3.
Additional Analyses and Robustness Check
We conducted several additional analyses to further evaluate the robustness of our findings.
Endogeneity concern: PSM analysis
Selection bias may determine how chair gender affects firm acquisitions (Francis, Hasan, Park, & Wu, Reference Francis, Hasan, Park and Wu2015). Specifically, firms deliberately hire female chairs, so some factors may correlate with firm acquisition activities (our dependent variable) and with hiring a female chair (our independent variable). To mitigate endogeneity concerns, we followed others (e.g., Gupta et al., Reference Gupta, Mortal, Chakrabarty, Guo and Turban2020) and used PSM analysis to create a matched sample of firms with female and male chairs respectively. By matching treated (firms with female chairs) and control groups (firms with male chairs) on the estimated probability of being treated, PSM helps reduce selection bias because female-led firms align with male-led firms on observable covariates. First, we conducted a logistic regression for each year to estimate the probability of having a female chair by including the two moderators and all control variables as covariates (industry and province fixed effects included). We matched 377 female-chair firm-year observations with 377 male-chair firm-year observations without replacement. Table 3 presents the results of balance tests of the treated and the control groups for the matched and unmatched samples. For the unmatched sample, the t-statistic showed a significant difference between the treated and the control group for several critical covariates, such as firm size and board size. However, a good balance was achieved through the matching process, as the t-statistic showed that the mean of all covariates between the treated and the control group exhibited insignificant differences in the matched sample.
Balance tests from propensity score matching approach
We then used the propensity-matched sample to examine how chair gender affects firm acquisition activities and the moderating effects. Table 4 presents the results of negative binomial regressions based on the propensity score-matched sample. Consistent with the results from the total sample, a positive and significant coefficient remains for chair gender in Model 2 (β = 0.308, p = 0.007). Specifically, female chairs engaged in 36% (=exp (0.308)−1) more acquisitions than male chairs, further supporting H1. We also found a negative and significant interaction term of chair gender and high female executive representation in Model 3 of Table 4 (β = −0.497, p = 0.041), indicating a stronger gender effect in firm acquisitions under low female executive representation, supporting H2. Model 4 of Table 4 shows a positive and significant interaction term of chair gender and chair political appointment (β = 0.148, p = 0.016), suggesting an amplified gender effect in firm acquisitions under higher levels of chair political appointment. Again, results based on propensity score-matched sample supported H3.
Negative binomial regressions predicting acquisition activities (propensity-matched sample)
Notes: robust standard errors in parentheses; †p < 0.10, *p < 0.05, **p < 0.01, ***p < 0.001; all tests are two-tailed.
Endogeneity concern: DID analysis
To further alleviate endogeneity concerns, we used chair succession as an event shock and employed the DID approach to examine chair gender effects on corporate acquisitions. First, we identified all chair succession events in Chinese SOEs from 2008 to 2020. Second, using data from the CSMAR's ‘CSRC's Enforcement Actions Research Database’ (Jia, Ding, Li, & Wu, Reference Jia, Ding, Li and Wu2009), we identified chair succession events in which the company had committed misconducts within the two years prior to the succession. Our results remained robust when identifying chair successions based on the presence of misconduct within a one-year window. We precluded succession events that were planned and greatly ensured the exogeneity of these events. Third, we deleted chair succession events in which the outgoing chair or the successor had a tenure shorter than two years because the individual would likely be an interim chair reluctant to engage in long-term strategic actions (Ballinger & Marcel, Reference Ballinger and Marcel2010). Eventually, we had 365 chair succession events during the time frame. Nine of the successions involved a change from a male to a female chair (male_female); 8 involved a change from a female to a male chair (female_male). The control group included 348 successions with no change in chair gender.
We conducted a DID analysis to examine how changes in chair gender during succession impacted corporate acquisitions. We generated a dataset of 1,460 (=365 × 4) observations containing a four-year window for each succession event, with a two-year window for both pre-event and post-event periods. To gather comprehensive data on both dependent and independent variables, we merged the firm-year dataset with the entire sample used for primary hypotheses testing. We dropped 325 observations that had missing variables, leaving a final sample of 1,135 observations for DID analysis. We then constructed the variable post to denote the post-event years. In the DID analysis, our primary variables of interest were the interaction term of male_female and post and the interaction term of female_male and post. If our results were robust, acquisition activities should increase after the succession involving a change from male to female chair (compared to the control group with no gender change after succession), so the coefficient of male_female × post should be positive.
Meanwhile, the coefficient of female_male × post should be negative, indicating that acquisition activities should decrease after the succession when a firm chair changes from female to male, compared to the control group. Model 1 of Table 5 shows the results of the DID analysis. The coefficient of male_female × post was positive and significant (β = 0.538, p = 0.012), while the coefficient of female_male × post was insignificant (β = −0.274, p = 0.482). Thus, DID analysis moderately supported H1.
Robustness checks: DID analysis and alternative measure of government's reward
Notes: RSE, robust standard error; †p < 0.10, *p < 0.05, ***p < 0.001; all tests are two-tailed.
Alternative operationalization of political appointment indicating positive governmental evaluation
The Chinese government rewards positively evaluated SOE chairs with political appointments and compensation (Cao et al., Reference Cao, Lemmon, Pan, Qian and Tian2019), but SOE executive compensation may be an imprecise and insensitive indicator because the government scrutinizes and restricts the compensation to minimize executive-employee pay disparities (Raynard et al., Reference Raynard, Lu and Jing2020). Therefore, in the primary analysis, we used the chair's political appointment as an indicator. In the robustness check, we conducted an additional analysis using chair compensation (logged) as a proxy for the reward representing positive governmental evaluation, which triggers self-efficacy in firm leaders. The results are reported in Model 2 of Table 5. Chair compensation had a positive and significant coefficient of interaction with chair gender (β = 0.029, p = 0.044), additionally supporting H3.
Gender effects in non-SOEs
SOEs differ from non-SOEs in how executives are evaluated. Specifically, SOE executives rely on positive governmental evaluations. In non-SOEs, the main evaluators for executives are usually the major or largest shareholders who can closely observe executive actions. Thus, we theorize that SOE contexts encourage female leaders because the government and its agencies are remotely located and are usually positively affected when female chairs make vivid counter-stereotypical evaluation decisions. Consequently, female agentic advantage is elicited and female leadership is positively related to firm acquisitions. In non-SOE contexts, however, we anticipate that female chairs will underperform male chairs in terms of firm acquisitions. In the additional analyses, we ran the models that test our three hypotheses in the non-SOE sample and found nonsignificant correlations. The additional results further support our contention that Chinese SOEs are a unique context where women disprove stereotypes by exhibiting counter-stereotypical traits.
Discussion
The widely accepted role incongruence perspective states that female executives tend to avoid taking risks. We introduce an alternative perspective, however, and show that female leaders may be more inclined than male leaders to take risks in certain contexts. That is, we examine Chinese SOEs where governmental agents, as major stakeholders and evaluators, ascribe agentic advantage to female executives who pursue firm acquisitions. We find strong support for our predictions that female leaders are more likely to take risks in contexts where risky behaviors can bring female leaders favorable evaluations. The effect is accentuated under two contingencies: higher gender stereotype threats and stronger self-efficacy. Specifically, female chairs are more likely than their male counterparts to engage in acquisition activities in Chinese SOEs, especially in firms that have lower female executive representation or when female chairs hold political appointments.
Theoretical Contributions
Our study makes several important contributions to literatures on mergers and acquisitions, female executives, and SOEs. First, we demonstrate that an anticipated female agentic advantage enables leader gender to positively affect risk-taking. The literature presents contradictory predictions regarding how female and male leaders differ in risk-taking behaviors, but the predominant role incongruence perspective explains that female leaders face gender disadvantages (Eagly & Karau, Reference Eagly and Karau2002), that female-led firms will be more conservative because female leaders may receive more gender-based bias and criticism if risky strategies fail (Post et al., Reference Post, Lokshin and Boone2022; Wang et al., Reference Wang, Wijen and Heugens2018). In contrast, the expectancy violation theory proposes that women could earn female agentic advantage through risk-taking behaviors (e.g., Adams & Funk, Reference Adams and Funk2012; Farag & Mallin, Reference Farag and Mallin2018). Here, we offer new insight into why and under what circumstances female executives might benefit by seeking risky strategies and violating expectancies.
Moreover, our study advances female leadership theory by contextualizing high risk-taking preferences among female leaders. Innate risk tolerance (Adams & Funk, Reference Adams and Funk2012) has been used to explain why some female executives seek more risks than male executives (Farag & Mallin, Reference Farag and Mallin2018). Instead, we acknowledge that context has a strong influence, aligned with strategy scholars who have turned increasing attention to contextualizing the influence of female leadership (Gamache et al., Reference Gamache, Devers, Klein and Hannigan2023). More broadly, two primary contextualizing perspectives include the demand-side regarding how stakeholders evaluate female executives and the supply-side regarding how female executives impact firm strategies (Mah et al., Reference Mah, Kolev, McNamara, Pan and Devers2023). To explain how executive gender relates to strategic outcomes, we introduce anticipated major stakeholder evaluation to show that the demand side interacts interdependently with the supply side in affecting female risk-taking leadership (Fernandez-Mateo & Kaplan, Reference Fernandez-Mateo and Kaplan2018). Future research would benefit by introducing the demand and supply link for considering responsive strategies that female executives use to tackle stakeholder-held gender stereotypes.
Second, our study contributes to SOE research that has long examined how SOEs distinctly differ from non-SOEs (Li, Xia, Long, & Tan, Reference Li, Xia, Long and Tan2012). For example, both entities are often driven by different strategic objectives (Ghorbani et al., Reference Ghorbani, Xie, Jin and Wang2023). Executives in SOEs are more concerned about political incentives and tend to prioritize alignment with government goals, such as unemployment reduction; while non-SOEs are more focused on performance (Wang & Luo, Reference Wang and Luo2019). Comparative studies, however, find that SOE executive heterogeneity should be better studied (Bruton et al., Reference Bruton, Peng, Ahlstrom, Stan and Xu2015). In response, we examine internal heterogeneity and demonstrate that female-led rather than male-led SOEs engage in more acquisitions. Moreover, women are profoundly underrepresented in SOEs (Zhang, Reference Zhang2012), so we need more understandings about this gender-related issue. Our work challenges the prevailing assumption that female leaders are more risk-averse (e.g., Chen et al., Reference Chen, Crossland and Huang2016; Huang & Kisgen, Reference Huang and Kisgen2013), and may inspire further investigations on gender differences within the context of SOEs.
Third, our study contributes to the M&A literature. Leaders often engage in corporate acquisitions primarily to reap personal gains, such as increased compensation and social status (Seo et al., Reference Seo, Gamache, Devers and Carpenter2015; Shi et al., Reference Shi, Zhang and Hoskisson2017). However, we lack research regarding how leaders of different genders differ in leveraging acquisition functions. By framing acquisitions in Chinese SOEs as a counter-stereotypical behavior that can trigger female agentic advantage, our study demonstrates that female leaders confront gender stereotypes when they pursue acquisition activities.
Practical Implications
By revealing gender effects on corporate acquisitions and the mechanism of female agentic advantage in the SOE context, our study has important practical implications for female leaders, firm stakeholders, and evaluators. We suggest that the major stakeholders and evaluators responsible for assessing corporate executives may need to consider whether their assessments involve gender stereotypes, because such evaluations may cause female executives to form responsive career strategies that substantially influence subsequent firm-level outcomes. We further suggest that the government should avoid the symbolic approach of merely mandating a quota for females in leadership positions. Instead, to tackle the pressing issue of female underrepresentation in SOEs, female leaders should be encouraged to actively challenge gender stereotypes and female agentic behaviors should be favorably evaluated. Female executives who aspire to shatter the career glass ceiling may benefit from learning how to leverage female agentic advantage and challenge entrenched gender biases. Furthermore, female executives tend to respond radically to stereotype threats in the workplace (Post et al., Reference Post, Lokshin and Boone2022). Thus, firms having limited female representation in upper echelons or led by female executives who receive positive personal feedback should remain vigilant regarding how female executives will perceive and react to gender stereotype threats.
Limitations and Future Research Directions
Our study contains limitations that may provide directions for future fruitful research. First, although we theorize that agentic female advantage is associated with acquisition engagement and the corresponding governmental positive evaluations, we should recognize that women have varying innate risk propensities. Women in the upper echelons appear to be a paradox: they have lower risk-taking propensity than male colleagues but want to demonstrate competence by breaking gender stereotypes (Post et al., Reference Post, Lokshin and Boone2022). Although we find that female chairs are eager to demonstrate their competence through risk-taking in the Chinese SOE context, our data prohibit excluding innate risk-taking propensities. Future work may utilize field studies to directly explore these effects. Another challenge is how to measure stereotype threats. Directly measuring female chairs' perceptions is challenging, especially over long periods. We used a proxy of low female executive representation to indicate whether the female leader perceived gender stereotype threats. Future research may collect large-scale data of firm leaders' activities, speech, and body language, and use machine-learning techniques to code psychological constructs (Harrison, Josefy, Kalm, & Krause, Reference Harrison, Josefy, Kalm and Krause2023).
Data availability statement
The data supporting the findings of this study has been published in Open Science Framework at https://osf.io/bnf29/
Acknowledgments
We thank the editors and three anonymous reviewers for their valuable and constructive feedback on this article. We acknowledge the financial support from the National Natural Science Foundation of China (#71972048) and the Sci-Tech Innovation Foundation of Fudan University School of Management (# 20210106), and the financial support for open access of the article from Fudan University Library.
Danyang Zhu (dyzhu17@fudan.edu.cn) is a postdoctoral researcher at the School of Management, Fudan University. She earned her PhD from Fudan University. Her research interests focus on strategic leadership, female executives, and organizational resilience, with recent interests in firm digitalization.
Xu-Hong Li (xhli@fudan.edu.cn) is a professor of strategic leadership and organizational behavior at the School of Management, Fudan University. Her current research focuses on the micro-foundation of strategic leadership and corporate governance, as well as entrepreneurs in Chinese technology-intensive companies and state-owned companies. Her work has appeared in journals such as Academy of Management Journal and International Journal of Stress Management, among others.
Open access
