IB&M Master Thesis
Emerging market multinationals acquire abroad to innovate
at home: The moderating role of business group affiliation
Final version - 21.06.2017 Word count: 11.959
First Supervisor: Dr. Sathyajit Gubbi
Co-assessor: Melih Astarlioglu
Name: Annika Zerweck
Student number: S3168999
II
ABSTRACT
III
TABLE OF CONTENTS
1 Introduction ... 1
2 Theory and hypotheses ... 4
2.1 Theoretical background ... 4
2.1.1 Strategic asset-seeking cross-border acquisitions ... 4
2.1.2 New domestic product launches ... 5
2.2 Literature review and hypotheses development ... 6
2.2.1 The impact of asset-seeking CBAs on new domestic product launches .... 6
2.2.2 The moderating role of business group affiliation ... 9
2.2.3 The differential effect of specific BG characteristics... 11
3 Methodology ... 15
3.1 Empirical setting ... 15
3.2 Sample and data ... 15
3.3 Variable measurement ... 17
3.3.1 Dependent variable ... 17
3.3.2 Independent variable ... 18
3.3.3 Moderators and controls ... 19
3.4 Model and estimation ... 20
4 Results ... 22
4.1 Pre-analysis ... 22
4.2 Full model results ... 24
V
TABLES
Table 1: Means, S.D., Min, Max and Pearson’s correlations – Full model ... 23
Table 2: Means, S.D., Min, Max and Pearson’s correlations – BG model ... 23
Table 3: Zero-inflated negative binomial regression - Full model ... 25
Table 4: Zero-inflated negative binomial regression - BG model ... 27
Table 5: Coding examples for new product launches ... G Table 6: Coding examples for acquisition types ... G Table 7: Overdispersion of dependent variable - Full model ... I Table 8: Overdispersion of dependent variable - BG model ... J Table 9: Multicollinearity check - Full model version 1 ... K Table 10: Multicollinearity check - BG model version 1 ... K Table 11: Pearson’s correlations – Full model version 1 ... L Table 12: Pearson’s correlations – BG model version 1... L Table 13: Multicollinearity check - Full model version 2 ... M Table 14: Multicollinearity check - BG model version 2 ... M Table 15: Robustness test for one year time lag ... N Table 16: Robustness test with alternative time lags and dependent variables ... O FIGURES Figure 1: Conceptual model for hypotheses 1-2 – Full model... 11
VI
LIST OF ABBREVIATIONS
AMNC Advanced market multinational corporation
BG Business group
BSE Bombay Stock Exchange
CBA Cross-border acquisition
EMNC Emerging market multinational corporation
FDI Foreign direct investment
NB Negative binomial
R&D Research & development
T Tolerance
1
1 INTRODUCTION
In the recent decade, emerging economies shifted from being merely recipients of for-eign direct investment (FDI) to becoming major contributors. Their outward FDI has more than tripled since 2005 and accounted for more than 25% of the global amount in 2015. Thereof, roughly one third was spent on cross-border mergers & acquisitions, increasingly also in de-veloped economies (UNCTAD, 2016). While traditional internationalization theory fails to ex-plain this expansion pattern, more recent frameworks argue that multinational corporations from emerging markets (EMNCs) acquire strategic assets abroad as a springboard to overcome their latecomer disadvantage (Cuervo-Cazurra, 2012; Luo & Tung, 2007; Mathews, 2006). Re-lated to that, scholars made the interesting notion that EMNCs acquire these assets mainly for exploitation in their domestic markets (Hennart, 2009; Ramamurti, 2012). Recent waves of liberalizations in emerging economies led to an increased influx of foreign competition (Cas-tellacci, 2015), strengthening the need for EMNCs to engage in innovation catch-up (Awate, Larsen, & Mudambi, 2012) to stay competitive in their attractively growing domestic markets (World Bank Group, 2017).
2 faces intensified foreign competition not only due to market liberalization, but also because of the implementation of industry-specific reforms (Agrawal & Saibaba, 2001; Ahluwalia, 2002).
Moreover, only little is known about the conditions that enhance an EMNC’s ability to successfully master the hurdles on the way from acquiring abroad to introducing a new product at home (Hitt, Li, & Xu, 2016; Nair, Demirbag, & Mellahi, 2015). Challenges range from iden-tifying promising acquisition targets, transferring relevant assets from the distant subsidiary to the local headquarters and integrating them with the own knowledge base and R&D activities (Awate, Larsen, & Mudambi, 2015; Cloodt, Hagedoorn, & Van Kranenburg, 2006). These is-sues are especially pronounced for EMNCs that tend to suffer from inferior resource endow-ment, insufficient managerial experience and a weak institutional environment (Hennart, 2012; Khanna & Palepu, 2006).
3 In line with recent findings, this paper argues that asset-seeking CBAs will in general enhance new domestic product launches (Anderson et al., 2015; Li et al., 2016), but the rela-tionship is expected to be stronger for BG affiliates. BGs can increase their affiliate’s absorp-tive capacity, foster collaboration with the acquired unit and finance R&D projects (Mahmood et al., 2011). Furthermore, as not all BGs are alike and some BG characteristics are expected to be of more value than others, two additional BG-level moderators are assessed in a separate model, focusing solely on BG affiliated firms. First, BG acquisition experience is expected to be of positive impact, since affiliates get access to learnings about knowledge integration rou-tines across corporate cultures (Vermeulen & Barkema, 2001). Second, BG internationalization helps affiliates to identify relevant strategic assets abroad, and to bridge cultural and institu-tional distances in the revere knowledge transfer process (Lamin, 2013).
4
2 THEORY AND HYPOTHESES 2.1 Theoretical background
2.1.1 Strategic asset-seeking cross-border acquisitions
Emerging economies have long been mainly recipients of FDI, but this is subject to change. Their outward FDI has more than tripled since 2005 and accounted for more than 25% of the global outward FDI in 2015. Thereof, roughly one third was spent on CBAs, increasingly also in advanced economies (UNCTAD, 2016). Scholars debate whether traditional interna-tionalization theory is sufficient to explain these FDI decisions of EMNCs (Cuervo-Cazurra, 2012; Hennart, 2012; Luo & Wang, 2012; Narula, 2006).
Dunning’s widely acknowledged eclectic paradigm assumes that an MNC controls su-perior resources which can be exploited in foreign markets (Dunning, 1988). Critiques argue that EMNCs lack such firm-specific advantages due to their inferior resource endowment and latecomer disadvantage (Mathews, 2006; Rugman, 2009). Accordingly, Mathews (2002) intro-duced the notion of the latecomer firm that uses resource-targeting strategies to catch-up with incumbent firms. Similarly, Luo and Tung’s (2007) springboard perspective identifies the ag-gressive, high-risk buy-in strategy of EMNCs as a means of acquiring strategic assets from mature MNCs. Such strategic asset-seeking CBAs are distinct from other types of FDI, such as resource-seeking, market-seeking or efficiency-seeking FDI (Dunning, 2009).
5 2.1.2 New domestic product launches
Some authors argue that EMNCs acquire those assets mainly for exploitation in their home markets (Hennart, 2012; Ramamurti, 2012). There a several supporting arguments for this. First, many emerging economies have recently undergone institutional reforms that came with market liberalizations (Elango & Pattnaik, 2007). This attracted foreign investments and increased the competitive pressure (Castellacci, 2015). Hence, EMNCs are urged to engage in innovation catch-up in order to defend their domestic market position (Awate et al., 2012). Second, the high growth rates of emerging markets make them an attractive business arena. As main drivers of the global economy, they are expected to continuously grow with a rate above the 4%-mark, while advanced economies fight stagnation with forecasted rates well below 2% (World Bank Group, 2017). Third, in their domestic markets, EMNCs avoid issues related to liability of foreignness. They even enjoy certain advantages due to their familiarity with the environment. For instance, they have superior insights into local customer preferences (Rama-murti, 2012) and exclusive access to complementary local resources not available to foreigners that allow EMNCs to create powerful resource bundles for the domestic market (Hennart, 2009).
6
2.2 Literature review and hypotheses development
2.2.1 The impact of asset-seeking CBAs on new domestic product launches
The mere intent to translate acquired assets into new domestic products does not nec-essarily mean that EMNCs have the ability to do so (Anderson et al., 2015). It is crucial that only relevant and strategic assets are purchased, which then need to undergo the process of integration and assimilation (Barkema & Schijven, 2008; Cloodt et al., 2006). Additional chal-lenges arise from the need to transfer the assets across country borders from the newly acquired target firm to the EMNC’s domestic headquarters (Hitt et al., 2016). Conventional forward knowledge transfers have been extensively studied, especially in the context of AMNCs’ in-vestments in emerging markets (Ambos, Ambos, & Schlegelmilch, 2006). However, the re-search body about reverse knowledge transfers is less comprehensive, and only few studies address the topic from an EMNC perspective (Nair et al., 2015).
In the traditional AMNC context, Ambos et al. (2006) find that a high competitive strength of the host country enhances the benefits a company can gain from reverse knowledge transfers. Moreover, subsidiaries in the strategic role of contributors or global innovators1,
which are expected to proactively enhance the MNC’s knowledge base (Gupta & Govindara-jan, 1991), are positively associated with reverse knowledge flows and related benefits (Ambos et al., 2006; Rabbiosi, 2011). Other findings suggest that the choice of an appropriate mix of subsidiary autonomy and coordination can enhance a headquarters’ learning from the periphery (Rabbiosi, 2011). Furthermore, a high absorptive capacity of the headquarters, defined as “abil-ity to recognize the value of new information, assimilate it, and apply it to commercial ends” (Cohen & Levinthal, 1990, p.128), plays a crucial role. A considerable overlap of the
1 Different definitions of subsidiary roles (Ambos et al., 2006; Gupta & Govindarajan, 1991; Nair et al.,
2015; Rabbiosi, 2011) are harmonized as follows: (1) Local implementer (low outflows; high inflows): Hold mostly local market knowledge and adapt MNC´s firm-specific advantages to local market conditions. (2)
Con-tributor (high inflows, high outflows): Knowledge brokers with considerable expertise that is shared with other
7 knowledge of target and acquiring firm allows the EMNC to better absorb and integrate new assets (Ambos et al., 2006; Yang, Mudambi, & Meyer, 2008). Similarly, absolute and relative size of knowledge bases (Ahuja & Katila, 2001; Cloodt et al., 2006) and the newness of the acquired technology (Van de Vrande, Vanhaverbeke, & Duysters, 2011) have been identified as determinants of an acquisition’s impact on innovation output.
Only few, more recent studies assess the concept of post-acquisition reverse knowledge transfer and subsequent innovation output in the particular context of EMNCs (Anderson et al., 2015; Awate et al., 2015; Li et al., 2016; Nair et al., 2015). Nair et al. (2015) could partly replicate earlier findings in their study of Indian EMNEs. First, contributor subsidiaries and acquisition targets in highly competitive countries contribute relatively more to a reverse knowledge flow than local implementers or subsidiaries in countries with a competitiveness similar to India’s. Second, close collaboration has a positive impact, especially in knowledge-intense industries. Two additional studies placed in the Chinese context could show that stra-tegic asset-seeking outward FDI is positively related to domestic innovation performance (An-derson et al., 2015; Li et al., 2016). When addressing the conditions under which this effect is enhanced, Li et al. (2016) find a positive moderating role for absorptive capacity and inward FDI. Since their study is based on the regional level of analysis, though, the role of firm-level factors remains unattended. Anderson et al. (2015) hypothesize that state-ownership will lead to superior post-acquisition innovation performance, but fail to find support for this notion.
8 economies so that they can launch more products in their home market. Moreover, this study addresses the need to identify contingencies that enhance their ability to do so.
Several arguments support the notion that EMNCs are indeed likely to benefit from strategic asset-seeking CBAs through an increase in new domestic product launches. First, this type of acquisition intent naturally implies that the acquisition target takes on a role of contrib-utor or innovator subsidiary and proactively enhances the knowledge stock of the EMNC. Sec-ond, EMNCs acquire strategic assets preferably in developed countries, which tend to be more competitive than their home markets (Ambos et al., 2006; Nair et al., 2015). Third, through their superior knowledge of local customer preferences (Ramamurti, 2012) and access to com-plementary local resources (Hennart, 2012) EMNCs can successfully adjust and recombine the new assets in the domestic market. The acquisition may “simply provide the missing link in the quest to develop a new product.” (Ambos et al., 2006, p.296). Last, especially in the phar-maceutical industry, it is a time-consuming process to develop and commercially launch a new product (Nerkar & Roberts, 2004). The acquisition of strategic assets is a means to considerably speed up this process (Nair et al., 2015) and to avoid time compression diseconomies (Dierickx & Cool, 1989) related to a purely organic development. Anecdotal evidence can be given with the acquisition of the Italian Etna Biotech by the Indian EMNC Cadila Healthcare Ltd. in 2008. The deal was commented by the Managing Director, Mr. Pankaj R. Patel as follows:
“As an integrated player in the field of healthcare, we have always been exploring opportunities that can take us to the next level of excellence. With this acquisition, we will be at the forefront of innovation for vaccine research and development."
One year later, the EMNC was able to launch the swine flu vaccine VaxiFlu-S in the Indian market. According to the given arguments, the following is proposed:
Hypothesis 1: Strategic asset-seeking CBAs in advanced economies have a positive influence
9 2.2.2 The moderating role of business group affiliation
While in general assuming that EMNCs are able to increase their product launches in the post-acquisition period, some of them may be more capable of doing so than others. As already outlined, the transfer and exploitation of foreign assets comes with a variety of chal-lenges and so far, only little is known about the conditions that favour this process in the par-ticular case of EMNCs (Hitt et al., 2016; Nair et al., 2015). A phenomenon that is widespread in emerging markets (Khanna & Yafeh, 2007) and has proven to improve firm performance (Fisman & Khanna, 2004; Iona, Leonida, & Navarra, 2013; Khanna & Rivkin, 2001) and in-novation performance of emerging market firms (Castellacci, 2015; Chang et al., 2006; Choi et al., 2011; Kim & Lui, 2015; Mahmood & Mitchell, 2004; Mahmood et al., 2011; Wang, Yi, Kafouros, & Yan, 2015) is BG affiliation. However, to date, no study assesses the role that BGs can play in facilitating reverse knowledge transfers so that affiliated firms can successfully expatriate acquired strategic assets at home. This paper seeks to fill this gap, as there are ex-tensive reasons to believe that BG affiliates will be able to outperform their independent com-petitors in this regard.
10 and internal transactions. Hence, BGs as paragons of the local economy can compensate for institutional and external market deficiencies (Chang et al., 2006; Khanna & Yafeh, 2007).
Besides, BGs provide an innovation infrastructure comprised out of several building blocks: knowledge sourcing, vertical intermediation, scientific labour markets and capital (Mahmood & Mitchell, 2004). These put affiliates at an advantage over independent firms in their reverse transfer, integration and commercialization of strategic assets. First, BGs can con-siderably enhance their affiliates’ absorptive capacity as member firms can tap into the pooled knowledge stock of the whole BG. BGs usually have ties along the whole supply chain, so that technological knowledge and information about its commercial potential residing with suppli-ers or buysuppli-ers is made available in the group network (Mahmood et al., 2011). Moreover, the internal labour market allows to develop scientific talent in group-owned academies or research institutes. It is common practice to transfer these experts, so that they support the most prom-ising projects. Personal interaction allows to exchange even tacit knowledge, which is hard to accomplish via external markets (Chang et al., 2006).
Second, BG affiliation can help enhance the collaboration between the acquired sub-sidiary and the affiliate’s headquarters. BGs usually enjoy well-established reputations, so that its affiliates are seen as trustworthy partners (Mahmood & Mitchell, 2004). Often, affiliates operate under the same umbrella brand (Fisman & Khanna, 2004), which signals their commit-ment to the BG’s culture and quality standards. With that, the perceived attractiveness of the acquiring firm increases for the target firm, which positively influences the acculturation be-tween the two units (Nahavandi & Malekzadeh, 1988). This again strengthens the mutual un-derstanding and collaboration between the units. Hence, it is more likely that knowledge em-bedded in the human resources of the acquisition target is openly shared.
11 uncertain R&D projects (Khanna & Yafeh, 2007). This mechanism of mutual insurance gives affiliates access to more patient capital for better conditions than available on the market (Belenzon & Berkovitz, 2010). As a result, BG members are less constrained in the quantity of new product developments they pursue, and might even engage in more ambitious R&D pro-jects that require the implementation of causally ambiguous strategic assets (Gubbi & Elango, 2016; Mahmood & Mitchell, 2004). Taken together, this paper posits:
Hypothesis 2: BG affiliation moderates the relationship between strategic asset-seeking
CBAs in advanced economies and new domestic product launches of the acquiring firm from the emerging economy such that BG affiliates can launch more new domestic products than independent firms.
The conceptual model underlying the first two hypotheses is illustrated in Figure 1.
Figure 1: Conceptual model for hypotheses 1-2 – Full model
2.2.3 The differential effect of specific BG characteristics
13 Hence, it is proposed:
Hypothesis 3a: For BG affiliates from an emerging economy, BG acquisition experience
moderates the relationship between strategic asset-seeking CBAs in advanced economies and new domestic product launches such that, the higher the BG acquisition experience, the more new products can be launched domestically.
While acquisitions are in general a complex endeavour, managing them across borders is even more challenging. The difficulties begin already with the search for and selection of an appropriate target. Moreover, after acquiring abroad, differences between national cultures and institutional environments need to be overcome (Shimizu et al., 2004). Additionally, EMNCs tend to not follow gradual internationalization steps. In order to acquire sophisticated assets, they enter distant cultural blocs directly in a high commitment mode (Luo & Tung, 2007). Therefore, drawing on extensive international experience is crucial for CBA success (Barkema & Drogendijk, 2007). A high level of internationalization, defined as “the degree to which firm’s sales revenue or operations are conducted outside its home country” (Elango & Pattnaik, 2007, p.542), is a crucial source of intra-organizational competitive advantage that fosters learning (Iona et al., 2013). When firms lack international experience, they can learn capabili-ties for operating abroad within their domestic BG network (Elango & Pattnaik, 2007).
14 In addition, BG internationalization can help bridge cultural and institutional differ-ences, and with that fosters the reverse knowledge transfer and integration process. Building on experiential learning theory, Perkins (2014) finds that MNCs benefit from a great breadth and depths of institutional experiences in subsequent foreign investments. Being familiar with the rules of the game reduces uncertainty related to the investment. For example, affiliates may be aware of regulations about the cross-border transfer of intellectual property in the post-ac-quisition period. Besides, the importance of culture intelligence for successful cross-cultural interpersonal knowledge exchange gained broad acknowledgement (Earley & Mosakowski, 2004). Affiliates can request the transfer of managers with global experience or a deep under-standing of a particular host country culture within the BG network, which in turn supports the asset transfer and integration process. Taken together, this paper posits:
Hypothesis 3b: For BG affiliates from an emerging economy, BG internationalization
moder-ates the relationship between strategic asset-seeking CBAs in advanced economies and new domestic product launches such that, the higher the BG internationalization, the more new products can be launched domestically.
The following illustration depicts the underlying conceptual model of the second part of the proposed research, which focuses solely on business group affiliates.
15
3 METHODOLOGY 3.1 Empirical setting
The knowledge-intense pharmaceutical industry in India is a suitable empirical setting for the proposed research. First, Indian firms have compelling reason to participate in the surg-ing CBA activity of the global pharmaceutical industry that rose by $ 61 billion alone in 2015 (UNCTAD, 2016). They are forced to catch-up with an intensified foreign competition at home due to market liberalizations in the 1990s (Ahluwalia, 2002) and because of two industry-spe-cific reforms. In 1995, the Drug Order disrupted industry prices in a way appealing to foreign investors (Gubbi, Aulakh, & Ray, 2015) and by 2005, the government implemented product patents and higher intellectual property protection for foreign investors (Agrawal & Saibaba, 2001). Second, defending domestic market share in India is highly attractive. The local econ-omy is expected to grow with more than 7% within the next three years and estimates for the pharmaceutical industry are even higher (World Bank Group, 2017). Third, the pharmaceutical business landscape in India is composed of a mix of BG affiliated and standalone firms (Chit-toor et al., 2009).
3.2 Sample and data
16 with Orbis for company data and Zephyr for acquisition data. Both firms engaging in acquisi-tions and those relying solely on organic development during the observation period are in-cluded in the sample. Only with that approach, differences in product launch behaviour can be traced back to the impact of acquisitions (Ahuja & Katila, 2001; Fowler & Schmidt, 1988).
Due to the cross-sectional and time-series nature of the data, a panel design was main-tained. A dynamic of new listings and delisting activities in the selected period inevitably leads to an unbalanced panel structure. Under the assumption that the gaps in the firm-year observa-tions are caused randomly, the used estimation is robust to that (Cameron & Trivedi, 2010). Since the effect of strategic asset-seeking CBAs won’t influence the product launch behaviour of the acquiring firm immediately, a time lag is required in the study design (Ahuja & Katila, 2001). While the time lags employed by related studies range from one to five years (Choi et al., 2011; Cloodt et al., 2006; Kim & Lui, 2015; Van de Vrande et al., 2011), a one-year time window is regarded as sufficient for the given case. As outlined in chapter 2.2.1, speed to mar-ket is a driving motive behind strategic asset-seeking CBAs. Moreover, those deals often in-clude product marketing rights, which allow a relatively quick commercialization. The proce-dure of lagging all explanatory variables by one year reduces the panel frame to the eleven years from 2001-2011 with 224 firms and 1997 firm-year observations.
17 Furthermore, 24 firms had to be excluded due to missing data, eliminating 180 firm-year ob-servations. Lastly, all firms with less than 2 firm-year observations over the panel period were dropped, so that inferences about causality are possible (Lamin, 2013). This led to a final sam-ple of 172 firms with 1531 firm-year observations. Data was available for all eleven years for more than 50% of firms. The final panel period 2001-2011 is sufficient, because it covers the critical time after major institutional changes in the Indian pharmaceutical industry as outlined previously. 126 of the firms in the sample are independent and 46 BG affiliated firms. For testing hypotheses 1 and 2, the full sample is used. The test of hypothesis 3a and 3b requires to build a BG subsample comprising only data on the 46 BG affiliated firms.
3.3 Variable measurement
3.3.1 Dependent variable
18 to reach additional patients or usage scenarios. Furthermore, each product launch was counted separately if an announcement referred to multiple introductions. In total, the coding process revealed 494 new product launches. Next, the launch market was identified. 233 new product launches took place in the domestic Indian market, while 261 new products were introduced abroad. Coding examples are presented in Appendix I.
3.3.2 Independent variable
CBA activities of the sample firms were collected in a similar manner, as content anal-ysis is especially suitable to infer the acquisition intent from the BSE announcements (Morris, 1994). Since its mere announcement does not guarantee the finalization of a transaction, only CBAs with confirmed completion were included. All deals resulting in an ownership of less than 10% of equity were excluded to ensure that the acquiring firm has substantial control over the target’s strategic assets (La Porta, Florencio Lopez-de-Silanes, & Shleifer, 1999). Moreo-ver, only acquisition targets in developed economies as classified by the United Nations Con-ference on Trade And Development (UNCTAD, 2016) are counted. Overall, those steps re-turned 60 completed CBAs in advanced economies.
19 simultaneously, with 43 acquisitions identified as asset-seeking and 39 classified as market-seeking. Five acquisitions had to be eliminated since no acquisition rationale was given.
3.3.3 Moderators and controls
BG affiliation is operationalized as a dummy variable, assigning “0” to independent and “1” to affiliated firms. The underlying BG operationalization is derived from the Prowess database.
The measure for BG acquisition experience counts all acquisitions made by the BG prior to the focal acquisition of the affiliate. This includes both domestic and cross-border acquisitions made in a variety of industries. since each acquisition is expected to create valuable learnings (Gubbi & Elango, 2016). The third moderator, BG internationalization, is a discrete count var-iable that measures the number of countries in which the BG has operating subsidiaries (Tall-man & Li, 1996; Yang et al., 2008). This measure captures the range of international opera-tions, which matters for organizational learning (Johanson & Vahlne, 1977).
A set of control variables is included in the respective models. First, the level of inno-vation input is likely to influence the level of outputs (Ahuja & Katila, 2001), hence, a measure of firm R&D expenditure is included. Second, the effect of firm age is investigated. Older firms may be stuck in legacy processes and structures that inhibit their ability to transform and deploy knowledge acquired from international markets into new products (Naldi & Davidsson, 2014). Thirdly, firm size as the logarithm of total assets owned is accounted for. It is a known predictor for innovation outcomes, with larger firms having a higher propensity to successfully innovate (Belenzon & Berkovitz, 2010). Fourth, acquisition experience and internationalization are not only likely to have a positive impact on BG-level. Hence, firm acquisition experience and firm
internationalization are controlled for, keeping the operationalization constant. Furthermore, it
20 on the firm’s historical innovation capability (Blundell & Griffith, 1995). For example, it is likely that a new product launched in tablet form will be introduced as injection in subsequent years. Hence, lagged domestic product launches is included as control variable. Similar control variables are included on BG-level, with BG size measured as the logarithm of total assets of the BG, BG age calculated as the years since the first affiliate became part of the group and a measure of BG R&D expenditure.
3.4 Model and estimation
The dependent variable is a discrete, nonnegative integer that counts the number of times the event of a new domestic product launch occurs. This data type comes with typical characteristics. Events usually occur only rarely (Cameron & Trivedi, 2013), which leads to a distribution heavily skewed towards zero event counts. The skewness value of 7.26 along with a kurtosis of 67.27 depicted in Table 1 illustrate this asymmetrical clustering of the data (Field, 2009). This indicates that the majority of firms launch only few or even no new products, while a small group of firms accounts for a considerable proportion of launches (Choi et al., 2011). The respective histograms for the full model and the BG model can be found in Appendix II.
21 the full model (p=0.0001) and the BG model (p=0.0274) revealed that this is indeed the case. Zero-inflation means that not all zeros are caused by the same process, but that the so called excess zeros are generated separately. In a mixed model approach, the NB model is comple-mented by a logit model that identifies which of the two processes a zero event is associated with (Long & Freese, 2006). Transferred to the reality of Indian pharmaceutical firms, one could argue that not all firms engage in announcing their product launches on BSE. As already mentioned in chapter 3.2, this may be related to firm size, which is why firm size was selected as inflator factor in the full model, and BG size in the BG model, respectively. Both inflator factors prove to be highly significant in the respective models.
22
4 RESULTS 4.1 Pre-analysis
First, multicollinearity was checked based on Pearson’s correlations, variance inflator factors and tolerance values2. Respective calculations can be found in Appendix III. Multicol-linearity was detected between internationalization and acquisition experience both on firm and BG-level. Many of the sample firms seem to acquire mainly abroad and in a wide range of countries, which then affects both measures in a similar manner. Therefore, a firm’s export intensity calculated as the ratio of exports to sales (Gubbi & Elango, 2016) was selected as alternative proxy for internationalization (in the following referred to as BG and firm
interna-tionalization 2). With that, the required thresholds are met.
Additionally, firm and BG R&D expenditure exceed acceptable collinearity values. Ap-parently, R&D expenditures are zero for a large part of the firm-year observations simultane-ously on firm and BG-level. Moreover, sometimes the same value was reported on both levels, which occurs when the sample affiliate was the only BG affiliate to engage in R&D. Since the BG measure includes the R&D expenditures of all affiliates, the firm-level measure can be excluded from the BG model without losing its impact, so that multicollinearity is avoided.
Moreover, the study design was critically reviewed regarding endogeneity. One could argue that firms with a superior ability to launch new products may have more financial re-sources to acquire assets. This issue is regarded as sufficiently mitigated due to the time-lagged approach (Chang et al., 2006). Besides, BGs could pick firms with an especially high innova-tion performance as affiliates (Belenzon & Berkovitz, 2010). However, affiliainnova-tion to a BG is less a rational decision due to economic reasons, but rather a historical status often based on informal relationships (Mahmood et al., 2011) and stable over time (Lamin, 2013).
2 A Pearson’s correlation above 0.8 is an initial indication of multicollinearity. Eventually, VIF and T
23
Table 1: Means, S.D., Min, Max and Pearson’s correlations – Full model
Mean S.D. Min Max 1 2 3 4 5 6 7 8 9 Skewness Kurtosis
1 Domestic product launches 0.15 0.77 0.00 11.00 7.26 67.27
2 Asset-seeking acquisitionst-1 0.03 0.18 0.00 2.00 .11* 6.97 56.25
3 BG dummy 0.30 0.46 0.00 1.00 .20* .11* .89 1.79
4 Firm age t-1 23.25 17.40 1.00 109.00 .03 .07* .24* 2.34 9.16
5 Firm size t-1 2.94 1.86 -3.22 8.24 .23* .24* .43* .36* .35 2.35
6 Firm R&D expenditure t-1a 3.38 12.48 0.00 140.06 .36* .23* .28* .14* .50* 6.57 54.38
7 Firm performancet-1 -29.16 428.90 -13200 87.91 .02 .02 .05 .04 .07* .03 -22.65 624.70
8 Firm acquisition experience t-1 0.46 1.74 0.00 17.00 .24* .38* .20* .09* .46* .71* .02 5.42 37.58
9 Firm internationalization 2 t-1 0.23 0.26 0.00 1.07 .13* .15* .04 -.01 .41* .27* .06* .25* 1.03 2.96
10 Domestic product launches t-1 0.15 0.76 0.00 11.00 .41* .11* .20* .05 .26* .43* .02 .32* .14* 7.38 69.81
n = 1516 | * p < .05 | a figures in millions of US-Dollar | revised model after multicollinearity check
Table 2: Means, S.D., Min, Max and Pearson’s correlations – BG model
Mean S.D. Min Max 1 2 3 4 5 6 7 8 9 10 11 12 Skew-ness
Kurtosis
1 Domestic product launches 0.39 1.27 0.00 11.00 4.47 26.15
2 Asset-seeking acquisitionst-1 0.06 0.25 0.00 2.00 .08 4.60 25.44 3 BG acquisition experience t-1 0.38 1.22 0.00 11.00 .10* .42* 5.58 41.15 4 BG internationalization 2 t-1 0.24 0.19 0.00 0.83 .27* .16* .17* 4.94 28.94 5 Firm age t-1 29.61 19.19 1.00 103.00 -.06 .08 .07 .10* 1.59 5.60 6 Firm size t-1 4.18 1.71 0.67 8.24 .23* .29* .39* .42* .20* -0.15 2.23 7 Firm performancet-1 2.29 25.47 -219.04 70.21 .08 .12* .14* .24* .12* .23* -4.85 34.68
8 Firm acquisition experience t-1 0.99 2.40 0.00 14.00 .22* .41* .83* .36* .01 .56* .20* 2.92 11.32
9 Firm internationalization 2 t-1 0.24 0.21 0.00 0.77 .26* .14* .17* .71* .02 .48* .22* .34* 2.79 10.24
10 BG age t-1 42.56 26.95 10.00 139.00 -.11* .12* .24* -.09 .62* .07 .02 .05 -.12* 1.40 4.57
11 BG size t-1 5.45 1.45 2.22 10.20 .09 .19* .35* .19* -.01 .42* -.13* .42* .22* .20* 0.35 2.99
12 BG R&D expenditure t-1a 10.12 19.71 0.00 140.06 .32* .25* .48* .49* .07 .61* .22* .75* .44* -.02 .46* 3.45 17.62
13 Domestic product launches t-1 0.38 1.26 0.00 11.00 .40* .09 .15* .27* -.04 .25* .10* .28* .27* -.09* .12* .36*
4.56
24
4.2 Full model results
The regression results for the full model based on 172 firms and 1,516 firm-year obser-vations are given in Table 3. The improvement over the baseline model is quite small with the log likelihood3 increasing from -414.86 to -414.78 when adding the independent variable in model 2. This indicates that other than expected asset-seeking acquisitions may be only a weak predictor of domestic product launches. For model 3, the model fit achieves a value of -412.44. Model 1 shows the baseline model that includes only the controls. Firm size (ß=-0.6405; p=0.0002) has a significant and negative effect. A positive and strongly significant relationship was found for firm R&D expenditures (ß=0.0197; p=0.0001), the lagged dependent variable (ß=0.3518; p=0.0001) and the BG affiliation dummy (ß=1.1261; p=0.0048). Besides firm size, those controls relate to new domestic product launches as predicted. The other control variables all depict non-significant coefficients. Model 2 adds the dependent variable to the regression. While H1 predicts a positive effect of asset-seeking CBAs on new domestic product launches, the coefficient of asset-seeking acquisitions is indeed positive in this model, but quite far from significance (ß=0.1520; p=0.7802). Hence, model 2 does not support H1.
In model 3, the interaction effect between asset-seeking acquisitions and BG affiliation is tested. First, it needs to be noted that even though effect size and p-value improve, the posi-tive impact of asset-seeking acquisitions does still not become significant (ß=1.4634; p=0.2649). Hence, H1 needs to be rejected. Quite surprisingly, the interaction of BG affiliation and asset-seeking acquisitions depicts a negative sign (ß=-1.7884), but remains insignificant (p=0.2116). Consequently, no support for H2 predicting a positive moderating effect of BG affiliation was found. The coefficients of the control variables all keep their signs and signifi-cance, and no remarkable change in effect sizes can be observed.
3 It needs to be noted that other than R2 in linear regressions, the log likelihood does not express how
25
Table 3: Zero-inflated negative binomial regression - Full model
Model 1 Model 2 Model 3
Firm aget-1 -0.0109 0.1623 -0.0110 0.1581 -0.0105 0.1834
(0.0078) (0.0078) (0.0079)
Firm sizet-1 -0.6405** 0.0002 -0.6475** 0.0001 -0.6874** 0.0000
(0.1733) (0.1603) (0.1417)
Firm R&D expendituret-1 0.0197** 0.0001 0.0199** 0.0001 0.0207** 0.0006
(0.0051) (0.0050) (0.0060)
Firm performancet-1 0.0037 0.6103 0.0036 0.6132 0.0037 0.6085
(0.0072) (0.0071) (0.0073)
Firm acquisition experience t-1 0.0474 0.4120 0.0463 0.4356 0.0582 0.2141
(0.0578) (0.0594) (0.0468)
Firm internationalization 2t-1 0.6414 0.3455 0.6278 0.3565 0.5444 0.4483
(0.6800) (0.6808) (0.7179)
Domestic product launchest-1 0.3518** 0.0001 0.3473** 0.0002 0.3709** 0.0003
(0.0917) (0.0936) (0.1032)
Asset-seeking acquisitionst-1 0.1520 0.7802 1.4634 0.2649
(0.5447) (1.3126)
BG affiliation dummy 1.1261** 0.0048 1.1332** 0.0025 1.3248** 0.0000
(0.3992) (0.3749) (0.3020)
BG affiliation dummy*asset-seeking acqui-sitionst-1 -1.7884 0.2116 (1.4317) Constant 1.1790† 0.0959 1.2059† 0.0822 1.2287† 0.0675 (0.7080) (0.6939) (0.6721) Number of firms 172 172 172
Observations (thereof zeros) 1,516 (1,420) 1,516 (1,420) 1,516 (1,420)
Wald chi2 73.74** 78.07** 103.81**
Log likelihood -414.86 -414.78 -412.44
26
4.3 BG model results
Table 4 depicts the regression results for the BG model based on 46 BG affiliates and 450 firm-year observations. The model fit improves marginally from -270.39 in the first model over -269.93 in model 2 to -269.14 in model 3. The best improvement can be found for model 4 (log likelihood=-266.99) after inclusion of the second interaction term.
The baseline model 1 with the controls depicts a negative and significant coefficient for
firm age (ß=-0.0323; p=0.0197). The lagged dependent variable (ß= 0.3634; p=0.0000), firm internationalization 2 (ß=1.9141; p=0.0613) and firm size (ß=0.4141; p=0.0269) relate
signif-icantly and positively to new domestic product launches, and with that meet expectations. For all other controls, the regression returned relatively small and insignificant coefficients. Model 2 adds the independent variable asset-seeking acquisitions. Interestingly, its coefficient is neg-ative (ß=-0.4671), but so far remains without significance (p=0.2001).
H3a expects BG acquisition experience to positively moderate the relationship between
asset-seeking acquisitions and new domestic product launches. This is tested in model 3. The
coefficient of BG acquisition experience as a standalone explanatory variable is indeed positive (ß=0.0246), but remains insignificant (p=0.9132). Surprisingly, the interaction effect of BG
acquisition experience and asset-seeking acquisitions returns a negative coefficient
(ß=-0.3362). Even though effect size and significance (p=0.0704) indicate that this negative influ-ence is rather marginal, H3a must be rejected. Despite a negligible switch of sign for the insig-nificant control variable firm acquisition experience, effect sizes, significance levels and signs remain consistent for the other variables in the model.
Model 4 tests the interaction effect between asset-seeking acquisitions and BG
27
Table 4: Zero-inflated negative binomial regression - BG model
Model 1 Model 2 Model 3 Model 4
Firm aget-1 -0.0323* 0.0197 -0.0314* 0.0231 -0.0323* 0.0344 -0.0317* 0.0251 (0.0138) (0.0138) (0.0153) (0.0142) Firm sizet-1 0.4141* 0.0269 0.4251* 0.0218 0.4190* 0.0287 0.4401* 0.0295 (0.1872) (0.1853) (0.1916) (0.2022) Firm performancet-1 -0.0014 0.8801 -0.0012 0.8917 -0.0009 0.9219 0.0002 0.9853 (0.0091) (0.0090) (0.0089) (0.0091)
Firm acquisition experience t-1 -0.0233 0.7623 -0.0110 0.8867 0.0175 0.8873 0.0312 0.7117
(0.0771) (0.0774) (0.1237) (0.0844) Firm internationalization 2t-1 1.9141† 0.0613 1.9198† 0.0603 1.9276† 0.0677 1.9006 0.2157 (1.0228) (1.0220) (1.0551) (1.5353) BG aget-1 0.0117 0.2973 0.0114 0.3052 0.0127 0.3225 0.0123 0.3004 (0.0112) (0.0111) (0.0128) (0.0118) BG size t-1 -0.4112 0.2332 -0.3967 0.2457 -0.4096 0.2348 -0.4247 0.2232 (0.3449) (0.3417) (0.3448) (0.3486) BG R&D expenditure t-1 0.0033 0.6752 0.0023 0.7769 0.0011 0.8959 -0.0002 0.9850 (0.0079) (0.0081) (0.0085) (0.0086)
Domestic product launches t-1 0.3634** 0.0000 0.3879** 0.0001 0.3819** 0.0001 0.3337** 0.0004
(0.0869) (0.0984) (0.0980) (0.0945) Asset-seeking acquisitions t-1 -0.4671 0.2001 -0.1118 0.7010 -7.3019** 0.0005 (0.3646) (0.2912) (2.1064) BG acquisition experiencet-1 0.0246 0.9132 (0.2259) BG internationalization 2 t-1 -0.1808 0.9226 (1.8598) Asset-seeking acquisitions t-1*BG acquisition experiencet-1 -0.3362† 0.0704
(0.1858)
Asset-seeking acquisitionst-1*BG internationalization 2t-1 12.6747** 0.0003
(3.4920)
Constant -1.1075 0.5151 -1.2427 0.4663 -1.1895 0.4855 -1.1025 0.5220
(1.7012) (1.7058) (1.7054) (1.7220)
Number of firms 46 46 46 46
Observations (thereof zeros) 450 (382) 450 (382) 450 (382) 450 (382)
Wald chi2 75.20** 78.08** 91.76** 153.5**
Log likelihood -270.39 -269.93 -269.14 -266.99
28 BG affiliates appear to rather be disadvantaged over standalone firms when it comes to asset exploitation in the post-acquisition phase. While BG internationalization 2 itself depicts a non-significant negative coefficient (ß=-0.1808; p=0.9226), in the interaction with asset-seeking
acquisitions, the sign flips to the positive (ß=12.6747) with a considerably strong effect size
and very high significance (p=0.0003). As a result, H3b is supported. Besides, firm
interna-tionalization 2 loses its significance in this model, and firm performance, firm acquisition ex-perience as well as BG R&D expenditure switch their signs. Since all three variables have very
small coefficients and are highly insignificant, this is regarded as unproblematic.
4.4 Robustness tests
A variety of robustness tests were done. On the one hand, several modifications have been applied to the analysis with the one-year time lag. On the other hand, alternative time windows of two and three years were implemented. Key figures are outlined in Appendix IV.
First, it was found that the results hold true for an alternative measure of firm and BG
internationalization 2 by repeating the regression with a 3-year average of export to sales ratio
prior to the acquisition. Second, calendar year dummies were included in the model to account for any macroeconomic and regulatory influences over the panel time frame (Lamin, 2013). With that, the negative interaction between asset-seeking acquisitions and BG acquisition
ex-perience loses its significance. However, the p-value was already quite weak in the main model.
Third, a similar effect was found for a re-run of the analysis without the lagged dependent variable. Besides insignificant but still negative results for H3a, results are mainly the same for the static model. Fourth, alternative inflator factors were selected in order to account for the possibility that other factors influence whether key events are announced. A re-run of the full model with firm internationalization 2 replacing firm size as inflator factor remained without considerable changes. The same holds true for the BG model, where instead of BG size, firm
pre-29 analysis, all predictors were standardized before the regression to refute any remaining concern (Aiken, West, & Reno, 1991). While this naturally affects coefficient sizes, signs and signifi-cance levels are robust.
The comparison of results across different time lags confirms the robustnuss for the full model, but yields some notable variability in the BG model. In the full model, the positive effect of asset-seeking acquisitions becomes slightly significant for the 3-year time window (ß=0.7814; p=0.093). At the same time the negative interaction with BG affiliation also crosses the threshold to significance (ß=-1.0153; p=0.059). This further supports the unexpected notion of a constraining effect of BG affiliation. In the BG model, BG acquisition experience loses its significantly negative impact on the relationship between asset-seeking acquisitions and
do-mestic product launches in both additional time lags. This shows that the negative moderating
impact of BG acquisition experiences seems to be strongest in the first year after the acquisi-tion, and deteriorates after longer periods of time. Second, there is remarkable change in results with regards to H3b. Even though coefficient signs are stable, effect sizes drop quite drastically in the interaction with BG internationalization 2 and significance is lost. This can partly be explained with the smaller sample as larger time lags inevitably come with less observations. Nevertheless, it needs to be noted that BG internationalization 2 seems to considerably lose its relevance in later stages in the post-acquisition period.
4.5 Supplementary analysis
30 the newly launched products are actually introduced in foreign markets, with the U.S. market playing a leading role. This goes hand in hand with the finding that many acquisition rationales showed indications for both an asset-seeking and market-seeking intent. Even though the do-mestic market is attractive in terms of high growth rates (World Bank Group, 2017), superior access to complementary resources (Hennart, 2009) and familiarity with the local environment (Ramamurti, 2012), the Indian pharmaceutical firms appear to simultaneously seek interna-tional expansion.
Therefore, the regression was repeated with foreign product launches as dependent var-iable. However, results were not considerably different as applicable in Table 16. For the full model, H1 remains unsupported, depicting a negative, insignificant coefficient. While the in-teraction of BG affiliation switches to the positive, it is still insignificant, further rejecting H2. In the BG model, results for H3a remain unchanged, confirming the negative moderating im-pact of BG acquisition experience. The interaction with BG internationalization 2 is still posi-tive and highly significant, further supporting H3b. Looking at the joint effects for all product launches, results are similar. Hence, the initial speculation can be largely refuted.
31
5 DISCUSSION
This paper expected that CBAs with an asset-seeking intent enhance new domestic product launches of the acquiring firm. It turns out that this notion finds no support. The find-ings indicate that Indian pharmaceutical firms’ domestic product launch behaviour remains largely unaffected by their strategic assets-seeking CBAs in advanced economies. There are several possible explanations for this. First, the data collection procedure unveiled that acqui-sitions by Indian pharmaceutical firms in advanced economies are still rather rare with only 43 cross-validated strategic asset-seeking cases over a 12-year period. This is surprising, as the pre-analysis of the empirical setting suggested a substantially different picture. It seems that the pharmaceutical industry only marginally accounts for the rising outward FDI of India that accumulated to substantial $ 7.5 billion in 2015 (UNCTAD, 2016), at least in the form of asset-seeking acquisitions. This may be due to a lack of affordable acquisitions targets, which Awate et al. (2015) mention as common issue for mature industries such as pharmaceuticals.
32 et al. (2015) illustrate the case of an acquired German subsidiary, which engages in opportun-istic and rent-seeking behaviour, holding back its superior knowledge. Possibly, such behav-iour is encountered quite often by EMNCs. If they are incapable of taking on a powerful posi-tion, they may be restricted in the benefits they can realize from asset-seeking CBAs.
In hypothesis 2 it was argued that the link between asset-seeking CBAs and new do-mestic product launches is positively enhanced for BG affiliates, so that affiliated firms can launch more new domestic products in the post-acquisition period than independent firms. In line with recent theory, BG affiliation has a positive impact in the direct relationship to new domestic product launches (Castellacci, 2015; Chang et al., 2006; Choi et al., 2011; Kim & Lui, 2015; Mahmood & Mitchell, 2004; Mahmood et al., 2011; Wang et al., 2015). However, being affiliated with a BG does not support the post-acquisition phase after a CBA.
33 Especially surprising is the finding that affiliates depict a lower post-acquisition inno-vation performance as a function of their BG’s acquisition experience, at least in the first year after the acquisition. This negative experience transfer (Barkema & Schijven, 2008) hints at the study of Haleblian and Finkelstein (1999), who find a u-shaped relationship between ac-quisition experience and abnormal stock returns. Low levels of acac-quisition experience may not just have a less positive, but even a negative effect when companies engage in inappropriate generalizations. Even though the acquisition experience of the BGs in the sample ranges be-tween zero and eleven, the average BG can look back on less than one acquisition, indicating that some BGs don´t reach an experience threshold that allows them to benefit from learnings. Moreover, the usually high diversification of BGs across multiple industries (Khanna & Yafeh, 2007) increases the likelihood that learnings from dissimilar settings are inappropriately trans-ferred. If affiliates fail to adjust group-level learnings to their distinct setting (Barkema & Schi-jven, 2008) or imitate acquisition routines due to legitimacy reasons (DiMaggio & Powell, 1983), their access to BG acquisition experience may even hamper efficiency in reverse trans-ferring and exploiting strategic assets.
34
6 CONCLUSION 6.1 Key findings
Other than expected, findings suggest that strategic asset-seeking CBAs play a rather subordinate role for new domestic product launches of Indian pharmaceutical firms. The few firms that engage in this means for innovation catch-up (Awate et al., 2012) seem to lack the capabilities to successfully exploit the value embedded in the acquired assets. Moreover, ef-fects may be felt over multiple years after the acquisition. Especially surprising are the results regarding the role of BGs. While the direct impact of BG affiliation on new domestic product launches is positive and in line with theory (Castellacci, 2015; Chang et al., 2006; Choi et al., 2011; Mahmood & Mitchell, 2004; Mahmood et al., 2011), BGs seem to rather constrain their affiliates when it comes to reverse knowledge transfer and post-acquisition asset exploitation. The observed negative acquisition experience transfer (Barkema & Schijven, 2008) suggests that affiliates tend to make inappropriate generalizations from potentially insufficient and dissimilar BG learnings. They may imitate BG post-acquisition routines for legitimacy reasons (DiMaggio & Powell, 1983), which hampers their asset exploitation efficiency if not adjusted to the particular setting at hand (Haleblian & Finkelstein, 1999). BGs only turned out to fulfil their expected role as paragons (Khanna & Yafeh, 2007) when the focus shifted to their international experience. Findings indicate that BGs with high levels of internationalization help their affiliates to identify relevant strategic assets in foreign markets and to overcome challenges regarding institutional and cultural differences.
6.2 Contributions and implications
35 pharmaceutical industry, since prior studies exclusively focused on patent filings in the Chinese market (Anderson et al., 2015; Li et al., 2016). With assessing the role of BG affiliation, this research paper followed the call to provide better insights into which factors help enhance re-verse knowledge transfers (Hitt et al., 2016). The lack of support for the notion that member-ship in a BG positively influences post-acquisition new domestic product launches may be a sign that the institutional development in India already reached a level that makes the benefits of BG affiliation obsolete. For future academic research, this implies the necessity to conse-quently assess the impact of BGs in interaction with the quality of the institutional environment as well as its potential improvement over time.
At the disaggregate level, this paper enhances the understanding of the differential im-pact of particular BG characteristics. Apparently, managers in BG affiliated firms need to trans-fer learnings from prior acquisitions on BG-level with caution. It is important that they adjust routines to their specific setting, and maybe seek external consultation if the internal experience stock is still insufficient. Moreover, it seems valuable for the management of Indian pharma-ceutical firms to foster international experience of their employees, even if their strategy fo-cuses on building a competitive advantage in the domestic market. They could for example send key personnel on assignments abroad and further encourage an exchange among global managers on BG-level.
6.3 Limitations and future research
36 especially in the BG model. Future studies may test the proposed hypotheses in additional em-pirical settings to overcome this shortcoming.
Second, the level of newness and innovativeness varies considerably across the counted product launch cases. For example, a firm may have already introduced certain products in the pre-sample period in other markets and with that the measured event may rather deserve the label of a repeated than a new product launch. Moreover, even if the introduced products are new in the portfolio of the respective firm, they are not necessarily new to the world market. In fact, the vast majority of products introduced by Indian pharmaceutical firms are generics of branded products going off patent. Future studies with a broader scope should disentangle different types of product launches and distinguish between multiple levels of innovativeness. Third, it needs to be noted that the sample is not limited to multinational corporations with operations outside their home country, but also includes Indian pharmaceutical firms that do business mainly domestically. Hence, their propensity to engage in strategic asset-seeking CBAs may be relatively lower. Due to sample size concerns it was impossible to drop these firms, however future studies based on a larger sample may focus solely on multinational cor-porations.
Fourth, the coding of the acquisitions showed that asset- and market-seeking CBAs cannot be treated as mutually exclusive. Moreover, the method of content analysis is prone to subjective interpretations that could lead to biased data. A proposed improvement for future research is to employ a survey design that offers firm representatives a more fine-grained scale to distinguish the different types of deal rationales (Gubbi & Elango, 2016).
A
REFERENCES
Agrawal, P., & Saibaba, P. (2001). TRIPS and India's pharmaceuticals industry. Economic
and Political Weekly, 36(39), 3787-3790.
Ahluwalia, M. S. (2002). Economic reforms in India since 1991: Has gradualism worked?
Journal of Economic Perspectives, 16(3), 67-88.
Ahuja, G., & Katila, R. (2001). Technological acquisitions and the innovation performance of acquiring firms: A longitudinal study. Strategic Management Journal, 22(3), 197-220. Aiken, L. S., West, S. G., & Reno, R. R. (1991). Multiple regression: Testing and
interpret-ing interactions. Newbury Park, CA: Sage Publications.
Ambos, T. C., Ambos, B., & Schlegelmilch, B. B. (2006). Learning from foreign subsidiar-ies: An empirical investigation of headquarters' benefits from reverse knowledge trans-fers. International Business Review, 15(3), 294-312.
Anderson, J., Sutherland, D., & Severe, S. (2015). An event study of home and host country patent generation in chinese MNEs undertaking strategic asset acquisitions in developed markets. International Business Review, 24(5), 758-771.
doi:10.1016/j.ibusrev.2015.01.007
Awate, S., Larsen, M. M., & Mudambi, R. (2012). EMNE catch-up strategies in the wind tur-bine industry: Is there a trade-off between output and innovation capabilities? Global
Strategy Journal, 2(3), 205-223.
Awate, S., Larsen, M. M., & Mudambi, R. (2015). Accessing vs sourcing knowledge: A com-parative study of R&D internationalization between emerging and advanced economy firms. Journal of International Business Studies, 46(1), 63-86. doi:10.1057/jibs.2014.46 Barkema, H. G., & Drogendijk, R. (2007). Internationalising in small, incremental or larger
steps? Journal of International Business Studies, 38(7), 1132-1148.
Barkema, H. G., & Schijven, M. (2008). How do firms learn to make acquisitions? A review of past research and an agenda for the future. Journal of Management, 34(3), 594-634. Belenzon, S., & Berkovitz, T. (2010). Innovation in business groups. Management Science,
56(3), 519-535.
Blundell, R., & Griffith, R. (1995). Dynamic count data models of technological innovation.
Economic Journal, 105(429), 333-344.
Cameron, A. C., & Trivedi, P. K. (2013). Regression analysis of count data (2nd ed.). Cam-bridge: Cambridge University Press. doi:10.1017/CBO9781139013567
B Castellacci, F. (2015). Institutional voids or organizational resilience? Business groups,
inno-vation, and market development in Latin America. World Development, 70, 43-58. doi:10.1016/j.worlddev.2014.12.014
Chang, S., Chung, C., & Mahmood, I. P. (2006). When and how does business group affilia-tion promote firm innovaaffilia-tion? A tale of two emerging economies. Organizaaffilia-tion Science,
17(5), 637-656.
Chittoor, R., Sarkar, M. B., Ray, S., & Aulakh, P. S. (2009). Third-world copycats to emerg-ing multinationals: Institutional changes and organizational transformation in the Indian pharmaceutical industry. Organization Science, 20(1), 187-205.
Choi, S. B., Lee, S. H., & Williams, C. (2011). Ownership and firm innovation in a transition economy: Evidence from China. Research Policy, 40(3), 441-452. doi:10.1016/j.re-spol.2011.01.004
Cloodt, M., Hagedoorn, J., & Van Kranenburg, H. (2006). Mergers and acquisitions: Their effect on the innovative performance of companies in high-tech industries. Research
Policy, 35(5), 642-654. doi:10.1016/j.respol.2006.02.007
Cohen, W. M., & Levinthal, D. A. (1990). Absorptive capacity: A new perspective on learn-ing and innovation. Administrative Science Quarterly, 35(1), 128-152.
Cuervo-Cazurra, A. (2012). Extending theory by analyzing developing country multinational companies: Solving the goldilocks debate. Global Strategy Journal, 2(3), 153-167. doi:10.1111/j.2042-5805.2012.01039.x
Dierickx, I., & Cool, K. (1989). Asset stock accumulation and sustainability of competitive advantage. Management Science, 35(12), 1504-1511.
DiMaggio, P. J., & Powell, W. W. (1983). The iron cage revisited: Institutional isomorphism and collective rationality in organizational fields. American Sociological Review, 48(2), 147-160.
Dunning, J. H. (1988). The eclectic paradigm of international production: A restatement and some possible extensions. Journal of International Business Studies, 19(1), 1-31.
Dunning, J. H. (2009). Location and the multinational enterprise: A neglected factor? Journal
of International Business Studies, 40(1), 5-19.
Earley, C., & Mosakowski, E. (2004). Toward culture intelligence: Turning cultural differ-ences into a workplace advantage. The Academy of Management Executive, 18(3), 151. Elango, B., & Pattnaik, C. (2007). Building capabilities for international operations through
networks: A study of Indian firms. Journal of International Business Studies, 38(4), 541-555.
C Fisman, R., & Khanna, T. (2004). Facilitating development: The role of business groups.
World Development, 32(4), 609-628. doi:10.1016/j.worlddev.2003.08.012
Fleming, L. (2002). Finding the organizational sources of technological breakthroughs: The story of hewlett-packard's thermal ink-jet. Industrial and Corporate Change, 11(5), 1059-1084.
Fowler, K. L., & Schmidt, D. R. (1988). Tender offers, acquisition, and subsequent perfor-mance in manufacturing firms. Academy of Management Journal, 31(4), 962-974. doi:10.2307/256347
Gubbi, S. R., Aulakh, P. S., & Ray, S. (2015). International search behavior of business group affiliated firms: Scope of institutional changes and intragroup heterogeneity.
Organiza-tion Science, 26(5), 1485-1501.
Gubbi, S. R., & Elango, B. (2016). Resource deepening vs. resource extension: Impact on as-set-seeking acquisition performance. Management International Review, 56(3), 353-384. doi:10.1007/s11575-016-0282-y
Guillén, M. F. (2003). Experience, imitation, and the sequence of foreign entry: Wholly owned and joint-venture manufacturing by South Korean firms and business groups in China, 1987-1995. Journal of International Business Studies, 34(2), 185-198.
Gupta, A. K., & Govindarajan, V. (1991). Knowledge flows and the structure of control within multinational corporations. Academy of Management Review, 16(4), 768-792. doi:10.5465/AMR.1991.4279628
Haleblian, J., & Finkelstein, S. (1999). The influence of organizational acquisition experience on acquisition performance: A behavioral learning perspective. Administrative Science
Quarterly, 44(1), 29-56.
Hennart, J. (2009). Down with MNE-centric theories! Market entry and expansion as the bun-dling of MNE and local assets. Journal of International Business Studies, 40(9), 1432-1454. doi:10.1057/jibs.2009.42
Hennart, J. (2012). Emerging market multinationals and the theory of the multinational enter-prise. Global Strategy Journal, 2(3), 168-187. doi:10.1111/j.2042-5805.2012.01038.x Hilbe, J. M. (2011). Negative binomial regression (2nd ed.). GB: Cambridge University Press. Hilbe, J. M. (2014). Modeling count data. Cambridge: Cambridge University Press.
doi:10.1017/CBO9781139236065
Hitt, M., Hoskisson, R., Johnson, R., & Moesel, D. (1996). The market for corporate control and firm innovation. Academy of Management Journal, 39(5), 1084-1119.
doi:10.2307/256993
Hitt, M. A., Li, D., & Xu, K. (2016). International strategy: From local to global and beyond.
Journal of World Business, 51(1), 58-73.
D Iona, A., Leonida, L., & Navarra, P. (2013). Business group affiliation, innovation,
interna-tionalization, and performance: A semiparametric analysis. Global Strategy Journal,
3(3), 244-261. doi:10.1111/j.2042-5805.2013.01060.x
Johanson, J., & Vahlne, J. (1977). The internationalization process of the firm: A model of knowledge development and increasing foreign market commitments. Journal of
Inter-national Business Studies, 8(1), 23-32.
Khanna, T., & Palepu, K. G. (2006). Emerging giants. (cover story). Harvard Business
Re-view, 84(10), 60-69.
Khanna, T., & Rivkin, J. W. (2001). Estimating the performance effects of business groups in emerging markets. Strategic Management Journal, 22(1), 45-74.
Khanna, T., & Yafeh, Y. (2007). Business groups in emerging markets: Paragons or para-sites? Journal of Economic Literature, 45(2), 331-372.
Kim, Y., & Lui, S. S. (2015). The impacts of external network and business group on innova-tion: Do the types of innovation matter? Journal of Business Research, 68(9), 1964-1973. doi:10.1016/j.jbusres.2015.01.006
La Porta, R., Florencio Lopez-de-Silanes, & Shleifer, A. (1999). Corporate ownership around the world. The Journal of Finance, 54(2), 471-517.
Lamin, A. (2013). Business groups as information resource: An investigation of business group affiliation in the Indian software services industry. Academy of Management
Jour-nal, 56(5), 1487-1509. doi:10.5465/amj.2011.0176
Li, J., Strange, R., Ning, L., & Sutherland, D. (2016). Outward foreign direct investment and domestic innovation performance: Evidence from China. International Business Review,
25(5), 1010-1019. doi://dx.doi.org.proxy-ub.rug.nl/10.1016/j.ibusrev.2016.01.008
Long, S., & Freese, J. (2006). Regression models for categorical dependent variables using
stata (2nd ed.). US: Stata Press.
Luo, Y., & Tung, R. L. (2007). International expansion of emerging market enterprises: A springboard perspective. Journal of International Business Studies, 38(4), 481-498. Luo, Y., & Wang, S. L. (2012). Foreign direct investment strategies by developing country
multinationals: A diagnostic model for home country effects. Global Strategy Journal,
2(3), 244-261. doi:10.1111/j.2042-5805.2012.01036.x
Mahmood, I. P., & Mitchell, W. (2004). Two faces: Effects of business groups on innovation in emerging economies. Management Science, 50(10), 1348-13651.
doi:10.1287/mnsc.1040.0259
E Mathews, J. A. (2002). Competitive advantages of the latecomer firm: A resource-based
ac-count of industrial catch-up strategies. Asia Pacific Journal of Management, 19(4), 467-488. doi:1020586223665
Mathews, J. A. (2006). Dragon multinationals: New players in 21st century globalization.
Asia Pacific Journal of Management, 23(1), 5-27. doi:10.1007/s10490-006-6113-0
Morris, R. (1994). Computerized content analysis in management research: A demonstration of advantages & limitations. Journal of Management, 20(4), 903-931.
Myers, R. H. (1990). Classical and modern regression with applications (2nd ed.). Boston: PWS-KENT.
Nahavandi, A., & Malekzadeh, A. R. (1988). Acculturation in mergers and acquisitions.
Academy of Management Review, 13(1), 79-90. doi:10.5465/AMR.1988.4306790
Nair, S. R., Demirbag, M., & Mellahi, K. (2015). Reverse knowledge transfer from overseas acquisitions: A survey of Indian MNEs. Management International Review, 55(2), 277-301. doi:10.1007/s11575-015-0242-y
Naldi, L., & Davidsson, P. (2014). Entrepreneurial growth: The role of international
knowledge acquisition as moderated by firm age. Journal of Business Venturing, 29(5), 687-703.
Narula, R. (2006). Globalization, new ecologies, new zoologies, and the purported death of the eclectic paradigm. Asia Pacific Journal of Management, 23(2), 143-151.
doi:10.1007/s10490-006-7162-0
Nerkar, A., & Roberts, P. W. (2004). Technological and product-market experience and the success of new product introductions in the pharmaceutical industry. Strategic
Manage-ment Journal, 25(8), 779-799.
Perkins, S. E. (2014). When does prior experience pay? Institutional experience and the mul-tinational corporation. Administrative Science Quarterly, 59(1), 145-181.
doi:10.1177/0001839214523603
Rabbiosi, L. (2011). Subsidiary roles and reverse knowledge transfer: An investigation of the effects of coordination mechanisms. Journal of International Management, 17(2), 97-113. doi://dx.doi.org/10.1016/j.intman.2010.10.001
Ramamurti, R. (2012). What is really different about emerging market multinationals? Global
Strategy Journal, 2(1), 41-47. doi:10.1002/gsj.1025
