Ambidextrous orientation and performance in corporate venture units: A multilevel analysis of CV units in emerging market multinationals

: The aim of this article is to examine absorptive capacity and innovation-oriented corporate culture as moderators of the relationship between the ambidexterity of a CV unit and its performance. Accordingly, we propose that the absorptive capacity of a CV unit and an innovation-oriented, corporate culture positively moderate the effect of a CV unit's ambidexterity on its performance. We test our hypotheses with data from 180 CV units and their Taiwanese parent firms. Our paper contributes to a more fine-grained understanding of the role of ambidexterity in the CV units of emerging-market MNCs.


Introduction
Ambidexterity is critical for the success and survival of firms (Junni et al., 2013;Simsek et al., 2009).For corporate venture (CV) units, ambidexterity is important because it allows CV units to aim at developing novel resources and capabilities (exploration) while ensuring that their work builds on and is integrated with current resources and capabilities (exploitation) (Hill and Birkinshaw, 2014).
Although prior research has established that CV-unit ambidexterity can improve CV-unit performance (Hill and Birkinshaw, 2014), such performance benefits may be difficult to achieve because of challenges related to ambidextrous designs, such as managing different mindsets (Gibson and Birkinshaw, 2004;Smith and Tushman, 2005) and implementing different or even "paradoxical" leadership styles (Lewis et al., 2014).Thus, researchers have called for examinations of the conditions and mechanisms under which ambidexterity enhances performance (Cao et al., 2009;Junni et al., 2013;O'Reilly and Tushman, 2013).
Accordingly, the aim of this article is to examine CV-unit absorptive capacity and an innovationoriented corporate culture as key moderators of the relationship between a CV unit's ambidexterity and its performance.These factors represent important organisational aspects that support value creation from innovation processes (Crossan and Apaydin, 2010).We test our model empirically in the dynamic context of 180 Taiwanese CV units.
This article contributes to a more fine-grained understanding of the role of ambidexterity in the CV units of emerging-market MNCs and thereby addresses calls to examine organisational aspects under which ambidexterity enhances performance (Cao et al., 2009;Junni et al., 2013;O'Reilly and Tushman, 2013).Specifically, we point to CV-unit absorptive capacity (Zahra and George, 2002;Jansen et al., 2005) as a key factor in boosting the positive performance effect of CV-unit ambidexterity.We also focus attention on the importance of culture in enhancing the benefits of ambidexterity (O'Reilly and Tushman, 2013) by showing that an innovation-oriented corporate culture can amplify the benefits of CV unit ambidexterity.Furthermore, CV-unit absorptive capacity and an innovation-oriented corporate culture interact to enhance the performance benefits of CV-unit ambidexterity, thereby testifying to the combined effect of these supportive organisational aspects.

Moderators of ambidexterity in CV units
Researchers have conceptualised ambidexterity in a variety of ways (see Birkinshaw and Gupta, 2013;Junni et al., 2013, for reviews).Building on March (1991), some scholars have regarded exploration and exploitation as mutually exclusive activities and argued that ambidexterity aims at a balance between exploration and exploitation (balanced ambidexterity) (Cao et al., 2009, Keller andWeibler, 2015;Fu et al., 2016).In turn, others have defined ambidexterity as an orthogonal construct where exploration and exploitation are pursued simultaneously (simultaneous ambidexterity) (Jansen, Tempelaar, Van den Bosch and Volberda, 2009).
In this study, we follow the theoretical approach to orthogonal ambidexterity and apply it in the context of the CV unit (Gibson and Birkinshaw, 2004).CV units are distinct business units that focus on development of opportunities for future growth (Block and MacMillan, 1993;Hill and Birkinshaw, 2014;Sharma and Chrisman, 1999).Accordingly, CV-unit ambidexterity refers to focusing on building new capabilities (exploration) while simultaneously focusing on exploiting existing capabilities (exploitation) (Hill and Birkinshaw, 2014).Expanding on prior research that shows that CV-unit ambidexterity relates positively to performance (Hill and Birkinshaw, 2014), we will propose in the following that absorptive capacity and an innovation-oriented corporate culture are key organisational moderators that further enhance the positive performance effect of ambidexterity.
The moderating role of absorptive capacity Cohen and Levinthal (1990) define absorptive capacity as the firm's ability to recognise, assimilate and apply new knowledge from the environment to the firm's benefit.Later conceptualisations (Zahra and George, 2002) and operationalisations (Jansen et al., 2005) have expanded absorptive capacity to include knowledge combination.In line with this, we understand absorptive capacity as an overall construct (Engelen et al., 2014) that involves interrelated elements concerning the recognition, assimilation, combination and application of new knowledge from the environment to benefit the firm (Cohen and Levinthal, 1990;Jansen et al., 2005;Zahra and George, 2002).Absorptive capacity combines outward-and inward-looking elements because it allows not only identifying and understanding external knowledge, but also understanding how it can be combined with existing firm-internal knowledge to benefit the firm (Cohen and Levinthal, 1990;Engelen et al., 2014).Hence absorptive capacity involves the CV unit's efforts to understand both the external (market) environment and the internal (firm) environment, and how they can be combined.
CV-unit ambidexterity can improve CV-unit performance because it allows the CV unit to aim not only at building new capabilities (exploration) but also at exploiting existing capabilities (exploitation) to better address current and future market opportunities (Hill and Birkinshaw, 2014).We suggest that CV-unit absorptive capacity further strengthens the positive performance effects of CV-unit ambidexterity.First, absorptive capacity involves continuous efforts by a CV unit to recognise new external market opportunities and to obtain knowledge from external stakeholders such as customers, trade partners and consultants (Jansen et al., 2005;Zahra and George, 2002).This increases the CV unit's awareness of state-of-the art technological and market developments so that exploration efforts can be better targeted to focus on the most value creating capabilities.In addition, absorptive capacity facilitates the frequent identification and acquisition of key internal MNC knowledge and resources.This is necessary for understanding the current internal resources and needs of the "internal corporate market"; more valuable opportunities can be identified and created for exploiting current resources and current and future market opportunities can be addressed more effectively.In contrast, when a CV unit has limited absorptive capacity, it risks focusing its efforts on activities that have a weaker fit with the needs of the external market and the firm's capabilities, making such efforts less likely to enhance CVunit performance.We therefore propose the following: Hypothesis 1.A CV unit's absorptive capacity positively moderates the relationship between its ambidexterity and performance.

The moderating role of an innovation-oriented corporate culture
We suggest that CV units supported by an innovation-oriented corporate culture benefit more from ambidexterity.An innovation-oriented corporate culture is a corporation-wide culture oriented toward innovation and initiated by top management; it supports openness, new ideas and risk-taking (Hurley and Hult, 1998;Lin et al., 2013) within the entire MNC.Because corporate culture embodies the deeply embedded assumptions, values, and norms of the corporation that guide behaviour, attention, and actions (Schein, 2004), corporate culture influences work activities, problem-solving, and relationship behaviour within the MNC (Chatman et al., 1998;O'Reilly and Chatman, 1996).
Hence, corporate executives at MNC headquarters who have instituted a strong innovationoriented corporate culture are likely to better understand, appreciate and support the use of existing capabilities (exploitation) and development of new capabilities (exploration) by their CV units (Raisch and Tushman, 2016).This may induce MNC headquarter executives to pay more attention to their CV units, give them more priority in the MNC strategic agenda and provide them with more resources, thereby enabling more effective use of current capabilities (exploitation) and creation of new capabilities that are more promising and valuable (exploration).Thus, the strategic importance and support gained from the greater strategic priority (Thornhill and Amit, 2001) stemming from an innovation-oriented corporate culture is likely to enhance the positive effect on performance of CV-unit ambidexterity.
In addition, CV units often depend on the resources of regular business units and may compete for resources with them (Raisch and Tushman, 2016).If the corporate culture does not support innovation and knowledge-sharing, social conflict is likely to prevent successful collaboration (Badguerahanian and Abetti, 1995;Hill and Birkinshaw, 2014).Under an innovation-oriented corporate culture, it is more likely that innovation is understood as a corporate-wide endeavour across the CV units and regular business units.Accordingly, regular business units are less likely to fear that exploration and exploitation by the CV unit will cannibalise their existing resources and mandates (Raisch and Tushman, 2016).In fact, under an innovation-oriented corporate culture, sharing knowledge and updating old practices instead of protecting them is encouraged (Hurley and Hult, 1998).Thus, when guided by a strong innovation-oriented corporate culture, regular units are less likely to resist and more likely to support the CV-unit by sharing more high-quality assets, technologies and intellectual capital.This in turn improves the quality of CV-unit exploitation and also promotes CV-unit exploration of novel technological opportunities on the market.Without an innovation-oriented corporate culture, regular business units are more likely to engage in political games that hinder CV-unit value creation, such as sharing lower quality resources and knowledge with the CV unit to sabotage the value of CV unit exploitation or highlighting the dangers and risks of new technologies to undermine exploratory efforts.Furthermore, when a corporate culture is innovation-oriented, failure is less likely to be penalised (Hurley and Hult, 1998).Hence, regular units also perceive collaboration with CV units as less risky and are more likely to want to share existing capabilities (exploitation) and build new capabilities (exploration) with the CV unit (Raisch and Tushman, 2016).Taken together, these arguments lead us to propose the following: Hypothesis 2. An innovation-oriented corporate culture positively moderates the relationship between the ambidexterity of a CV unit and its performance.
The moderating roles of absorptive capacity and an innovation-oriented corporate culture Finally, we argue that the combination of a high level of CV-unit absorptive capacity and an innovation-oriented corporate culture will further enhance the positive relationship between CVunit ambidexterity and performance.As argued in the first hypothesis, a high level of absorptive capacity enhances the CV unit's ability to use existing capabilities (exploitation) and build new capabilities (exploration) to address current and future business opportunities more effectively.As stated in the second hypothesis, an innovation-oriented corporate culture can enhance the ability and motivation of the MNC headquarters and regular business units to share high quality existing capabilities (exploitation) with the CV unit and to support its efforts to develop new capabilities (exploration).We propose that the value added by interaction between an innovation-oriented corporate culture and CV-unit absorptive capacity is based on knowledge sharing.An innovation-oriented corporate culture encourages developing and sharing diverse ideas, interests and mental maps (Huber, 1991;Lee et al., 2017).As a result, when the MNC is under the umbrella of an innovation-oriented corporate culture, the MNC headquarters and regular business units will be better able and more motivated to advise the CV unit regarding the potential value of external opportunities.Accordingly, when supported by an innovation-oriented corporate culture, CV units with a higher level of absorptive capacity will find it easier to absorb knowledge and ideas from the MNC headquarters and from regular business units to guide their exploration.Because such guidance from the MNC headquarters and other units can significantly increase the usefulness and relevance of the external knowledge absorbed by the CV unit (Song et al., 2018), an innovation-oriented corporate culture can further amplify the moderating effect of absorptive capacity on the ambidexterity-performance relationship.
Furthermore, under an innovation-oriented corporate culture, risk tolerance and the desire to invest in new ideas are corporate-wide characteristics (Hurley and Hult, 1998).As a result, the MNC headquarters and regular business units are more motivated to provide CV units with their best resources for exploitation, thereby increasing the potential for creating value.A CV unit with a higher level of absorptive capacity will be better able to put these resources to good use because it will have a better understanding of how the rest of the MNC and its resources work.Also, the aligning effect of an innovation-oriented corporate culture (Huber, 1991) makes it less likely that the resource contributions (e.g.knowledge, technology and human resources) shared by the MNC headquarters and the regular business units will be "distant".If the CV unit has a high level of absorptive capacity it will be easier to combine these resources with those of the CV unit to enhance the effectiveness of CV-unit exploitation (Zahra and George, 2002).Taken together, an innovation-oriented, corporate culture therefore makes knowledge and resources in the firm's internal environment more useful and accessible for the CV unit, which is likely to boost the positive effect of a CV unit's absorptive capacity on the relationship between its ambidexterity and performance.We therefore hypothesise the following: Hypothesis 3.An innovation-oriented corporate culture increases the moderating effect of a CV unit's absorptive capacity on the relationship between its ambidexterity and performance.

Sample
We collected our data in Taiwan in three industry sectors: high technology (38.0%), manufacturing1 (36.7%) and electrical engineering (25.3%).Emerging-market MNCs in these sectors are likely to be active in corporate venturing because they invest heavily in a technology portfolio (Zahra et al., 2000).Data from the Taiwan Economic Journal (TEJ) database show that there were 90 diversified MNCs in these sectors in 2013.One of the authors contacted the CEO of each MNC.The CEOs provided a list of the CV units and their top management team members.This resulted in a list of 237 CV units with 1440 CV-unit directors.The CV units were profit centres with their own management team and located in different geographic regions.
We conducted two surveys for temporary separation of the key variables of the study.Furthermore, to separate the sources of key variables, we used archival data to measure the dependent variable (CV-unit performance).In the first round of data collection in early 2014, we contacted the CV-unit directors and asked them to rate their CV-unit's absorptive capacity.After four weeks and three reminders, we had collected responses from 1160 directors of CV units (an 80.6% response rate) in 198 CV units.To be able to test for interrater reliability, we excluded 18 CV units from which we received complete responses from only one director.As a result, our final sample in the first data collection round included responses from 1142 directors of CV units (a response rate of 73.9%) in 180 CV units.
In the second survey, which took place 6 months later, we contacted the same 1142 CV-unit directors and asked them to rate CV-unit ambidexterity.To collect corporate level information regarding the innovation-oriented corporate culture, we also contacted corporate-level executives at the MNC headquarters.We received responses from 180 corporate-level executives in 79 MNCs.Our final sample consisted of responses from 79 MNCs to 180 CV units.
We reduced common method variance ex ante by collecting the independent and dependent variables from separate sources.In addition, to reduce common method variance ex post, we used complex statistical methods that included multilevel analyses and interaction terms to test our hypotheses (Chang et al., 2012).Furthermore, we conducted Harman's one-factor test, which showed that no single factor explained the majority of the variance.We also examined the potential effects of common method variance by identifying a suitable marker variable (socially desirable response) and applying a partial correlation procedure (Podsakoff et al., 2003).The socially desirable response variable did not correlate with the constructs in the model.

Measures 2
Independent variable: CV-unit ambidexterity.Based on Hill and Birkinshaw (2014), we measured CV-unit ambidexterity by asking the directors of the CV units to rate the importance of exploitative and explorative objectives for their CV units.The exploitative CV-unit objectives related to more effective utilisation of corporate human assets 3 (items 1 and 4), more effective utilisation of corporate assets in general (item 2) and monetisation of existing corporate technologies and intellectual capital (item 3).Composite reliability (CR) (0.80) and average variance extracted (AVE) (0.58) were acceptable.The explorative CV-unit objectives included creation of breakthrough technology for the corporation, investment in disruptive technologies that potentially cannibalise existing technologies and provision of a window on emerging technologies (CR = 0.89, AVE = 0.67).
Moderating variable: CV-unit absorptive capacity.Building on the absorptive capacity measure of Jansen et al. (2005), the directors of CV units answered 21 questions that measure CV-unit absorptive capacity (see Appendix 1 for a list of individual items).For the absorptive capacity measure, we aggregated the scores of the 21 questions for each CV unit (CR = 0.97, AVE = 0.58).
Moderating variable: Innovation-oriented corporate culture.Building on Hurley and Hult (1998), we measured innovation-oriented corporate culture by asking the corporate executives in the MNC headquarters to evaluate the following aspects: acceptance of technical innovation, active seeking of innovative ideas, acceptance of innovation, penalisation for new ideas that do not work (reverse coded) and innovation resistance (reverse coded).(CR = 0.94, AVE = 0.74).
Dependent variable: CV-Unit Performance.We obtained CV-unit financial performance data from the Taiwan Economic Journal (TEJ) database, which is the most comprehensive financial database in Taiwan (Chu, 2004;Hwang et al., 2012).TEJ includes information about the profitability of each reportable unit.We used the CV unit's annual gross profit rate as the measure of CV-unit performance, which is calculated as follows: . We used performance data for 2017 to allow for a time lag between the dependent variable and the independent and moderating variables collected in the two surveys in 2014. 5ontrol variables.Following Hill and Birkinshaw (2014), we controlled for CV-unit size (number of full-time employees in the CV unit), the age of the CV unit (years), the number of investments undertaken by the CV unit, and the CV unit's internal investments (Sharma and Chrisman, 1999).We obtained this information from the CV-unit managers in connection with the second survey round.In addition, we controlled for CV-unit capital investments based on information from the TEJ database in 2014 (coded 1 where the CV unit had investments recorded in the TEJ database and 0 where it did not).Furthermore, we controlled for the MNC financing structure, the total number of CV units in the MNC and whether the MNC's main business was in the high-technology sector (Hill and Birkinshaw, 2014) with information collected from the corporate-level executives.Finally, following Hill and Birkinshaw (2014), we controlled for the autonomy of the CV unit, including vertical and horizontal autonomy, to assess the extent to which CV-unit managers had the authority to make investment decisions.

Results
In our research design, we had a hierarchical data structure with two levels of variables, namely variables at the level of the CV units (Level 1, N = 180), which were nested within emergingmarket MNCs (Level 2, N = 79) (Andersson et al., 2014). 6The CV units were distinct profit centres with their own management team and located in various regions of the emerging-market MNC.Based on qualitative interviews, we assured that these units indeed had mandates to develop new business opportunities for the MNC such as new products and services.We followed the diagnostics recommended for multilevel models (Aguinis and Culpepper, 2015;Aguinis et al., 2013).We examined our hypotheses using ML-SEM using Mplus (version 7.0) software (Muthén and Muthén, 2015). 7able 1 shows the descriptive statistics and a correlation table, including dependent, independent and control variables as well as interaction terms (Kalnins, 2018).We first built a measurement model to establish construct validity and reliability.The measurement model fit the data well (χ2 = 1607.09,p = 0.000; df = 403, RMSEA = 0.05, CFI = 0.98 GFI = 0.91, TLI = 0.98).The construct reliabilities and average variances for the multi-item constructs were acceptable as reported in the 'Measures' section above.Concerning discriminant validity, we verified that the average variance extracted exceeded all corresponding squared pairwise correlations (Shook et al., 2004).
We then proceeded to test the hypothesised structural model.Following the recommendations by Andersson et al. (2014), we examined the direction of cross-level moderation effects. 8The hypothesised structural model was compared with four other nested models using the Satorra-Bentler scaled chi-square difference test (Muthén and Muthén, 2015) (see Table 2).When compared with more parsimonious models, the hypothesised model had the best fit, as demonstrated by the positive and significant chi-square difference tests.Table 3 summarises the results.Although we did not hypothesise a direct effect between CV-unit ambidexterity and CV-unit performance, the results confirmed the direct relationship (estimate = 0.29, p = 0.000; 95% confidence interval = CI = [0.02,0.56], △R 2 = 16%9 ; effect size10 = f 2 = 0.19).This effect size value approximately signifies a medium effect size according to Cohen (1988) guidelines (f 2 As to the hypotheses, the results support Hypothesis 1 by showing that CV unit absorptive capacity is a significant positive moderator of the relationship between CV-unit ambidexterity and CV-unit performance (estimate = 1.49, p = 0.000; 95% CI = [0.39,2.59], △R 2 = 14%, effect size = f 2 = 0.18), which corresponds approximately to a medium effect size according to Cohen (1988) guidelines (f 2 Our findings also support Hypothesis 2 by indicating that an innovation-oriented corporate culture positively moderates the relationship between CV-unit ambidexterity and CV-unit performance (estimate = 0.85, p = 0.000; 95% CI = [0.18,1.52], △R 2 = 10%, effect size = f 2 = 0.16).This effect size value can be interpreted as approximately medium according to Cohen (1988) guidelines (f 2 Furthermore, the results support Hypothesis 3 as demonstrated by significant three-way interaction between innovation-oriented corporate culture, CV-unit absorptive capacity and CV-unit performance (estimate = 0.31, p = 0.000; 95% CI = [0.06,0.56], △R 2 = 8%, effect size = f 2 = 0.15).This effect size value approximately signifies a medium effect size according to Cohen (1988) guidelines (f 2 For further testing of Hypothesis 3, we plotted CV unit absorptive capacity and innovation-oriented corporate culture following Cohen et al. (2003) and Dawson and Richter (2006) (see Fig. 2).In addition, we conducted simple slope difference tests. 11Finally, we split the sample into "high" and "low" ambidexterity subsamples using the mean as the splitting criterion and examined the effect of the interaction between absorptive capacity and innovationoriented corporate culture, which was more significant in the "high ambidexterity" sample.These additional tests were in line with Hypothesis 3. The control variables were not significant.

Discussion
Our aim in this article was to examine CV-unit absorptive capacity and innovation-oriented corporate culture as key moderators of the relationship between CV-unit ambidexterity and CVunit performance.Our results showed that a CV unit's absorptive capacity had a positive moderating effect on the relationship between CV-unit ambidexterity and CV-unit performance.This finding suggests that ambidextrous CV units are likely to create more value when they possess learning-oriented capabilities such as absorptive capacity (Zahra and George, 2002;Jansen et al., 2005).In addition, our results showed that an innovation-oriented corporate culture positively moderated the relationship between CV-unit ambidexterity and CV-unit performance.This extends prior research on the role of corporate culture as an important contextual enabler of innovation processes (O'Reilly and Tushman, 2008;Tushman and O'Reilly, 1996) that allows the CV unit to create more value from ambidexterity (Badguerahanian and Abetti, 1995;Narayanan et al., 2009).Finally, we found that an innovation-oriented corporate culture increased the moderating effect of CV-unit absorptive capacity.This result illustrates the role of corporate culture as an underlying corporate-level facilitator (O'Reilly et al., 2009) in MNCs that enhances the effects of CV-unit absorptive capacity.This is in line with research suggesting that culture and absorptive capacity are important organisational aspects (Crossan and Apaydin, 2010) that function as key facilitators and amplifiers of innovation processes.
By illustrating the interaction mechanisms between CV-unit ambidexterity, CV-unit absorptive capacity and an innovation-oriented corporate culture, we contribute to a more nuanced understanding of the role of ambidexterity in the CV units of emerging-market MNCs.While CV-unit ambidexterity can be performance-enhancing on its own (Hill and Birkinshaw, 2014), we identify absorptive capacity, an innovation-oriented corporate culture and their combined effect (interaction) as important factors that help further boost the performance of CV-unit ambidexterity.This contributes by addressing calls for examining when ambidexterity is most performance-enhancing (e.g., Junni et al., 2013;O'Reilly and Tushman, 2013).By enabling a better understanding of external technological and internal corporate developments, CV-unit absorptive capacity can help the CV unit to focus on developing the novel capabilities (exploration) that enhance performance most and to build effectively on current internal resources in order to address market opportunities (exploitation).This enhances the positive effect of CV-unit ambidexterity.In addition, we establish that an innovation-oriented corporate culture is a critical aspect for the CV unit because it facilitates additional support and more engaged participation for exploitation and exploration by other parts of the MNC.Furthermore, an innovation-oriented corporate culture interacts with absorptive capacity to enhance the performance benefits of CV-unit absorptive capacity because absorptive capacity increases the CV unit's ability to achieve optimal allocation of the high quality resources and support it receives from other parts of the MNC as a result of the corporation-wide, innovation-oriented culture.These findings provide empirical support for early studies on ambidexterity that theoretically emphasised the importance of a supporting corporate culture (e.g., O'Reilly and Tushman, 2013;Tushman and O'Reilly, 1996).This suggests that the benefits of CV-unit ambidexterity are increased if supported by an organisation-wide culture (Crossan and Apaydin, 2010;O'Reilly et al., 2009) that enhances the effects of CV-unit ambidexterity and absorptive capacity.
Regarding limitations, our study focuses on Taiwanese MNCs and their CV units, which may have unique characteristics such as a strong emphasis on corporate venturing in emerging market economies in general (Zahra et al., 2000) and in Taiwan in particular (Ministry of Economic Affairs, 2017).Future studies could examine the extent to which the results may vary across various emerging economies and to which extent they apply to MNCs and their CV units in more mature developed countries.Also, our sample included internal CV units within corporations whereas the dynamics of external venturing are likely to be different and require incorporating the role of external parties such as alliance partners.Furthermore, the CV-unit context does not allow for direct generalisation of the impact of ambidexterity on regular business units.
Concerning future research directions, by focusing on absorptive capacity and an innovationoriented culture, we demonstrate the importance of these key organisational moderators of ambidexterity in a CV-unit context.We hope that researchers will build on our study to identify and test the effects of other possible moderating factors, such as coordination (Jansen et al., 2005;Van den Bosch et al., 1999) and organisational configuration (Zimmermann et al., 2018).Furthermore, we demonstrate the importance of multilevel considerations, which could be taken further to identify microfoundational moderators of CV-unit ambidexterity such as individuallevel capabilities (Chang et al., 2009;Chandrasekaran et al., 2012).This could also be taken to an even higher level to compare the role of capabilities across MNCs from different countries.
For instance, emerging economies may benefit from government initiatives around innovation (Ramamurti and Hillemann, 2018).It would be interesting to examine the role of macro-level factors such as moderators of organisational ambidexterity and how they may interact with capabilities at the lower levels of analysis to influence performance.Our point here is that analysing the role of key moderators at different levels of analysis would help broaden our understanding of how they interact with ambidexterity across different levels to influence performance.
Our study provides key takeaways that managers can implement at the CV-unit and corporate levels to enhance the performance benefits of CV-unit ambidexterity.Managers of ambidextrous CV units need to focus on building strong absorptive capacity while executives at MNC headquarters can support ambidextrous CV units by creating and maintaining an innovationoriented corporate culture.
To conclude, our multilevel study provides novel insights by showing how CV-unit and corporate-level factors interact in determining the performance effects of a CV unit's ambidexterity in emerging-market MNCs.We hope that our study will serve as an inspiration for further multi-level studies on the moderators of CV-unit ambidexterity.

c 1 =
all financing decisions on venture investments made by MNC headquarters, 0 = not all financing decisions on venture investments made by MNC headquarters.a 1 = MNC in a high-technology sector; 0 = MNC not in a high-technology sector.b 1 = CV unit had investments recorded in the TEJ database in 2014, 0 = CV unit did not have investments recorded in the TEJ database in 2014.

Hypothesis 3
CV-unit ambidexterity × CV-unit absorptive capacity × Innovation-oriented corporate culture → CVp = 0.000, one-tailed tests.n = 180 at the CV-unit level (level 1); n = 79 at the MNC level (level 2).Innovation-oriented corporate culture measured at the MNC level; other variables measured at the CV-unit level.Unstandardised estimates.LLCI = lower level of the 95% confidence interval; UCLI = upper level of the 95% confidence interval.

Table 2 .
Comparison of structural models.Based on the Satorra-Bentler scaled chi-square difference test. a