The fast-changing global environment in recent years, including competition from multinational enterprises (MNEs) originating in emerging economies (‘‘South’’) and development of the Southern economies, has led to noticeable changes, both inside advanced-economy (which we term the ‘‘North’’) MNEs and in host countries (Awate, Larsen, & Mudambi, 2015; D’Agostino, 2015; Jha, Dhanaraj, & Krishnan, 2015; Mudambi, Piscitello, & Rabbiosi, 2014; Nair, Guldiken, Fainshmidt, & Pezeshkan, 2015; von Zedtwitz, Corsi, Søberg, & Frega, 2015). These include the increasing need to monitor and understand new global trends and hence to engage in multi-sourcing of knowledge inputs, a trend reinforced by rising R&D costs and a shortage of R&D personnel in the industrialized countries (Chen, 2004; Reddy, 1997). This has led to a strategic re-focusing of international innovation to simultaneously maximize the value of home-based knowledge and capture and synthesize local-based knowledge, both in the North (Chen, 2004; Zander, 1999) and in the South (Liu & Chen, 2012; Shin, Kim, & Park, 2016). In addition, market and institutional conditions in the South have substantially improved (though at varying speed), including an increased supply of qualified R&D personnel and skills, government support, emergence of regional innovation systems, preferential policies, and an increasingly sophisticated consumer demand (Guan, Yam, Tang, & Lau, 2009; Li, 2009; Liu, Simon, Sun, & Cao, 2011). The combination of these factors, as well as wider developments in the advancement and availability of information technologies in knowledge management (Mabey & Zhao, 2017; Yu, 2011), are driving the shift in innovation locus towards the South (Haakonsson & Ujjual, 2015; Haour & Jolly, 2014), with the aim of developing more sustainable competitive advantages by aligning internal and external knowledge networks globally. Consequently, in recent years MNE global innovation networks have extended to encompass the South (Jha, Dhanaraj, & Krishnan, 2018; Jha et al., 2015; Liu & Chen, 2012) as is clearly evidenced by a recent study published by PriceWaterhouseCoopers (2015) that identified Asia as the top destination for corporate R&D expenditure, ahead of North America and Europe.

Although this emerging trend is gaining a growing level of attention, research on R&D internationalization to date nevertheless takes a predominantly North-centric view. In comparison, studies addressing this trend using a South-centric lens remain sparse, with a few recent exceptions including Awate et al. (2015), Jain, Nair, and Ahlstrom (2015), von Zedtwitz et al. (2015) and Mudambi and Santangelo (2016). However, none of these studies were intended to provide the level of depth and detail that our article offers. Our historical case study approach is better able to capture the incremental changes in the South over a considerable length of time. It thus sheds new light on the historical development of four distinct waves of R&D internationalization derived from the current literature. As such, the objective of our study is not to test or re-examine the literature-informed waves. Instead it extends current literature on R&D internationalization towards the South by offering historical and contextual evidence of a micro-level study (i.e., AstraZeneca (AZ) in China) to be used for further theory development and empirical testing across other MNEs and/or Southern economies.

We first construct an overarching framework of R&D internationalization, which encompasses innovation in both the North and the South. Extending existing literature, our framework integrates four major waves of R&D internationalization and their distinct R&D objectives through time. It moves from a North-North wave for home-based knowledge augmentation and a North-South wave for home-based knowledge exploitation, to a South-South wave for local-based knowledge exploration and a South-North for home-based knowledge augmentation. Building on this, we apply a historical (longitudinal) lens to the case of AZ in China to trace and analyze changes in its R&D activities from its first entry in 1993 to 2017. In so doing, our article answers the following two research questions: (1) How have AZ’s operations in China evolved between 1993 and 2017 leading the country to play an increasingly active role in AZ’s R&D configuration?; (2) What can be learned from the case of AZ in China in light of the existing literature and theories on R&D internationalization?

Our article makes several key contributions. First, we construct a literature-informed comprehensive framework offering a classification of R&D internationalization waves and their corresponding R&D objectives. This framework explicitly illustrates and integrates four waves of R&D internationalization and their distinct R&D objectives in each wave, depicting the overall process of R&D evolution in the South. Second, in light of this framework we present a more complete picture of the R&D evolution process and the mechanisms underlying this process using the case of AZ China in the period between 1993 and 2017. Our comprehensive case study evidence offers new insights into how the South (in this case, AZ in China) has upgraded its role in R&D internationalization, thus focusing in particular on the last three waves of our framework of R&D internationalization. It thus addresses the limitations of existing research, which lacks a solid micro-level understanding of the incremental changes an MNE may experience inside a fast-changing Southern economy such as China. Third, by exploring the context of China, we offer new insights into the idiosyncratic institutional and market transformation of the country that impacted on AZ’s local R&D evolution. Whilst China is known to represent unique opportunities for MNEs, studies exploring its historical impact on MNE innovation activities remain sparse. Our article is one of the first to apply a longitudinal case study approach to reveal contextual details of R&D evolution in the South. Lastly, our article contributes to the methodological development in research on R&D internationalization research by way of applying a longitudinal historical approach to the study of R&D internationalization, a rather neglected approach (Welch & Piekkari, 2017). We demonstrate the suitability of a single case study in exemplifying ‘‘context’’ in qualitative research (Doz, 2011; Siggelkow, 2007; Welch, Piekkari, Plakoyiannaki, & Paavilainen-Mäntymäki, 2011), when the phenomenon in question is relatively new or difficult to study. The empirical evidence of our study is not intended to generalize across the industry or across other Southern economies, but instead is to be used for further theory development to advance our understanding as to whether the case of AZ in China and the proposed pattern of R&D evolution towards the South are typical.

In the remainder of this article, we first review the literature on R&D internationalization and construct an evolutionary framework. Then, we discuss our methodological approach and the background of our research context. This is followed by our case study findings. Finally, we discuss these findings and conclude with our contributions to theory development, implications for practice, and our recommendations for future research.

The internationalization of R&D: Literature review and conceptual framework

Literature review: R&D locations and objectives

'Dunning (1958), Safarian (1966) and Brash (1966) were among the first to discuss the act of internationalization of R&D displayed by some first-mover MNEs at the time. Soon after, an empirical study by Creamer (1976) provided clear evidence that a variety of overseas R&D activities were performed by MNEs from the North. Cordell (1973) and Ronstadt (1978) offered typologies of different overseas R&D activities. Subsequently, Behrman and Fischer (1980), Håkanson (1981), Hood and Young (1982), Niosi (1999), and Edler (2007) recorded a shift of global innovation activities from MNE headquarters towards the active participation of foreign subsidiaries, which became more prominent as an important source of new knowledge generation (Feinberg & Gupta, 2004; Griffith, Redding, & Van Reenen, 2004).

Initially, these strategically important locations for new knowledge generation were typically either geographically close or institutionally similar countries, so that MNEs could retain close control over key activities while also benefiting from locational advantages (Song, Asakawa, & Chu, 2011; Song & Shin, 2008). Hence, a large majority of MNEs chose to locate their international innovation activities in developed countries, particularly in the US, Europe, and Japan, while also transferring innovations to and back from other developed countries (Edler, 2007; Haakonsson & Ujjual, 2015; Haour & Jolly, 2014; Niosi, 1999). Although these studies provided valuable theoretical and empirical insights at the time, they were predominately underpinned by North-centric developments and thinking, whilet the South remained an ‘‘outsider’’ in the MNE global innovation network (Liu & Chen, 2012). Drawing on the terminology of ‘‘waves’’ from the works of Coe, Helpman, and Hoffmaister (1997) and Jha et al. (2015), we classify this trend as the North-North wave which is mostly reflecting ‘‘home-based knowledge augmentation’’ (HBKA-N) within Northern economies.

A second stream of literature has noted innovation-related knowledge transfer from headquarters to less developed economies in the form of production offshoring paired with on-site technical support and backup required to ensure timely problem-solving (Florida, 1997). Furthermore, the global investment environment witnessed significant changes, such as increasing international competition and advancing information technology (Castellani & Zanfei, 2006; Yu, 2011). We classify this development of limited R&D activities in the South as the North-South wave, which is closely aligned with Vernon’s (1966) product life cycle model whereby existing product knowledge is further exploited in new locations. Consequently, for the first time, the South became a destination for R&D activities, though this was generally viewed as more of an ‘‘operational necessity’’, restricted to production (Cantwell, 1992; Pearce & Papanastassiou, 1997). Drawing on previous literature on the objectives and characteristics of R&D internationalization (Bas & Sierra, 2002; Kuemmerle, 1999a; Liu & Chen, 2012; von Zedtwitz & Gassmann, 2002), this wave can be identified to reflect the main objective of ‘home-based knowledge exploitation’ (HBKE-S) in supporting the development of new markets and foreign production sites.

As home-based knowledge exploiters, MNEs chose to locate their R&D units close to large foreign markets or manufacturing sites to ensure on-time support. The main tasks of these R&D units tended to include assimilating knowledge transferred from the headquarters, modifying existing manufacturing processes, or modifying existing products to meet the local market conditions in the host countries (Athreye, Batsakis, & Singh, 2016; Pearce & Papanastassiou, 1999). By doing so, MNEs could gain economic return from utilizing their existing proprietary knowledge (generated in the North) in overseas locations (such as the South), while home-based R&D units continued to define the product specifications and design the manufacturing processes (Liu & Chen, 2012; Schmiele & Mangelsdorf, 2009).

A third and fourth stream of literature has emerged in recent time which suggests the increasingly crucial role of emerging economies in undertaking innovative activities for MNEs by incorporating the development of global innovation networks that excel at tapping into new centers of knowledge (Castellani, Jimenez, & Zanfei, 2013; D’Agostino & Santangelo, 2012; Dunning & Lundan, 2009; Jain et al., 2015; Mabey & Zhao, 2017; UNCTAD, 2012). What makes these two streams significant is that Southern economies are no longer viewed only as the location for low-cost labor or primary resources, but also potential sources of new knowledge (Prabhu & Jain, 2015). Some of the latest studies further support the growing proportion of FDI now devoted to strategic innovation in emerging economies, particularly Asia (Altenburg, Schmitz, & Stamm, 2008; Athreye, Tuncay-Celikel, & Ujjual, 2014; Bresciani & Ferraris, 2016; Castellani et al., 2013; Prashantham & Dhanaraj, 2015). These innovation investments target both the local market (we term this the ‘‘South-South’’ wave) and the global market (we term this the ‘‘South-North’’ wave). This significant development has been the focus of some of the latest research (e.g., Awate et al., 2015; Jha et al., 2015, 2018; Mabey & Zhao, 2017).

In terms of the South-South wave, the main objective of the MNE concerns ‘‘local-based knowledge exploration’’ (LBKE-S) in support of the development of new products for local and regional markets (Jha et al., 2015). R&D units under the South-South wave are focused on the seeking and development of knowledge useful in providing products to local markets (Jha et al., 2015). In situations where market demand is high and the availability of products is low, an opportunity arises for local-based R&D units to fill this gap by developing market-specific products. This leads R&D units to seek localized knowledge from entities such as research institutions, universities, local firms, and customers (Awate et al., 2015), and combine this knowledge with its existing internal capabilities to realize new or improved products that are market-specific. In terms of the South-North wave, MNEs are mainly engaged in ‘‘home-based knowledge augmentation’’ (HBKA-S) in support of the development of complementary and novel knowledge for the global market. R&D units under the South-North wave are concerned with the development of long-term innovative capabilities by taking advantage of novel or complementary knowledge in foreign locations, which is gaining growing importance and is driven by the pressure of knowledge-based competition (Jha et al., 2015, 2018). To augment their home-based technologies, MNEs set up their R&D units overseas with an aim to access the host countries’ knowledge for their headquarters (Asakawa & Ashok, 2008). One main objective of these units is to engage in knowledge sourcing and to benefit from potential knowledge spillovers (Almeida & Kogut, 1999; Feinberg & Gupta, 2004).

Nevertheless, both these waves reflect a picture that is dramatically different from the conventional view of R&D internationalization. MNEs no longer regard the South only as a target for market expansion and production, accompanied by limited R&D activities, but also as an important location for knowledge augmentation. Reverse knowledge transfer from the South to the North thus becomes a distinct possibility (Govindarajan & Ramamurti, 2011; Govindarajan & Trimble, 2012). This departs from the product life cycle understanding of knowledge exploitation and moves towards a global innovation network arrangement of differentiated and distinct R&D units worldwide (Mudambi & Santangelo, 2016). These R&D units, located in the North and the South, span across a wide variety of R&D activities.

Drawing on the four streams of literature, Table 1 summarizes our classifications of each of the four waves and associated R&D objectives.

Table 1 Classification of the four major waves in MNE R&D internationalization
Full size table

An overarching conceptual framework: The internationalization of R&D towards the South

While the four waves of R&D internationalization have been identified in prior research, they have remained largely independent as most studies focused on issues within the respective waves. In conventional North-North and North-South innovation paradigms, R&D stayed largely in the North while MNE activities in the South were predominantly driven by knowledge exploitation to meet growing local market demand. This reflected a highly hierarchical relationship between the North and the South. However, in the latest development of MNE R&D internationalization, Southern economies such as China (Bruche, 2009; Liu & Chen, 2012) and India (Awate et al., 2015; Jha et al., 2015; Kumaraswamy, Mudambi, Saranga, & Tripathy, 2012) have emerged as high-profile destinations in attracting inward innovation FDI. Some of the latest evidence (e.g., Liu & Chen, 2012; Liu et al., 2011) suggests that MNE R&D mandates in these locations have increasingly gone beyond the traditional pattern of knowledge exploitation. Manea and Pearce (2004), Pearce and Papanastassiou (2006), Dunning and Lundan (2008), and Zhang, Zhao, Bournakis, Pearce, and Papanastassiou (2018) have found that technological dynamism tends to shape the internal organizational structure of the MNE, which often explains the initial and subsequent strategic motivations of overseas subsidiaries. The strategic motivation underlying the establishment of subsidiaries in different countries depends on the locational opportunities and advantages that are present at the time of the initial investment. For instance, initially, market- and efficiency-seeking subsidiaries were mostly established in the South while knowledge-seeking subsidiaries were established in the North. Market- and efficiency-seeking motivations, which are underpinned by the strategic rationale of asset exploitation (Bartlett & Beamish, 2013; Dunning & Lundan, 2008), are intended to benefit from local resource abundance by investing in a particular country or region to supply goods or services to local, regional, and/or global markets. A knowledge-seeking motivation, on the other hand, concerns “the pursuit by MNEs of new technological capabilities, scientific capacity (research facilities) and creative expertise (e.g., dimensions of tacit knowledge) from particular host countries, in order to extend the overall competences (product range and core technology) of the group” (Manea & Pearce, 2004: 4). As the local environmental conditions in the South continue to evolve, benefits of an embedded innovation capacity for developing goods in and for host markets and regions become apparent. As a result, initial motivations of market- or efficiency-seeking can often shift towards local innovation through ‘‘asset augmenting’’ activities (e.g., knowledge-seeking) (Bartlett & Beamish, 2013; Cantwell, Dunning, & Lundan, 2010; Dunning & Lundan, 2008; Zhang et al., 2018).

Recent work by von Zedtwitz et al. (2015) extends the concept of ‘‘reverse innovation’’ by asserting that emerging economies are not simply locations for innovation adoption but also for new knowledge development. This implies that apart from knowledge exploitation (wave 2), local-based knowledge exploration (wave 3) and home-based knowledge augmentation (wave 4) are seen to take place in the South as R&D units become deeply involved in the local context with increased benefits of access to external knowledge, hiring of local personnel, and support of the local government.

Drawing on previous literature (Cantwell & Iammarino, 1998; Cantwell & Janne, 1999), one explanation for this development could be that countries with abundant innovation resources and a favorable supporting environment can influence the mandate of local R&D units (Ito & Wakasugi, 2007; Liu & Chen, 2012). In the case of certain economies in the South such as India and China, this is increasingly evident (Castellani et al., 2013; Dodgson, 2009; Lu, Tsang, & Peng, 2008). For instance, institutional conditions for innovation in these countries are fast improving through the efforts of national and local governments. Several pro-innovation factors can be identified: investment in higher education and the retention of skilled native and foreign researchers and managers (Yang & Jiang, 2007), development of clusters (encompassing local firms and support systems), and the establishment of national systems of innovation (encompassing MNEs, indigenous firms, public and private research institutions, universities, involving in a variety of innovation activities) (Archibugi & Iammarino, 2002; Dodgson, 2009; Lundvall, 1992, 2007). As a result, these economies benefit from an accumulation of innovative capabilities (Lema, Quadros, & Schmitz, 2015). As part of these changes, subsidiaries in these locations are offered the opportunity to improve their innovation capabilities. Indeed, a recent study by Zhang and Pearce (2011) finds that subsidiaries in China are not solely receivers of MNE knowledge, but also generators of new knowledge by working closely with local universities, firms, and research institutions. In particular, they argue that while market-seeking subsidiaries are frequently found in emerging economies, some of these economies have developed impressive innovation capacity for generating new knowledge, contributing de novo to the MNE’s technological trajectory. Thus, new knowledge generated in the South not only better captures the Southern market but can also be valuable to MNE global competitiveness (Mudambi & Santangelo, 2016). The locational hierarchy between the North and the South is thus flattening, whereby the South is starting to become part of MNEs’ global innovation network for knowledge augmentation (Chen, 2004; Liu & Chen, 2012). Nevertheless, this changing relationship between the North and the South, and the shift towards R&D in the South, does not happen instantly; instead, MNE R&D internationalization from the North to the South is likely to go through an evolutionary process. However, to date, studies attempting to directly map this pattern across the waves remain scarce (Govindarajan & Ramamurti, 2011; von Zedtwitz et al., 2015).

In this study, we assert that the waves in R&D internationalization are not static or unrelated. Instead, they are parts of a dynamic development in which the growth of innovation within MNEs has shifted towards the South. In Fig. 1 we illustrate the gradual shift in MNE R&D internationalization from the North to the South. During the first wave (North to North), the global innovation network only spanned across advanced economies, termed the first-tier location with the accompanying R&D objective being mainly HBKA-N. Thereafter, the South was considered a second-tier location predominantly for the purpose of HBKE-S to meet growing market demand. Subsequently, the increased efforts in improving innovation capability and the fast-growing national economies in the South made them an increasingly attractive hub for LBKE-S, thus adding a third-tier location to the global innovation network. In the most recent wave, further strengthening of innovation capabilities in the South has led to HBKA-S and thus a fourth tier emerged. In short, the framework proposes that the four waves found in literature can be considered as four consecutive processes of R&D internationalization while they are also mutually inclusive to represent a multi-layered continuous network of global innovation.

Fig. 1
figure 1

Four-wave framework of MNE R&D internationalization

Full size image

Therefore, this article proposes that MNE R&D internationalization towards the South is likely to show a dynamic pattern, from HBKA-N and HBKE-S to LBKE-S and HBKA-S. Hence, our empirical study takes MNE subsidiaries in the South as the appropriate unit of analysis in light of this latest development in R&D internationalization (Liu & Chen, 2007, 2012). Furthermore, to address the empirical problem noted by Liu and Chen (2012): “the spatial distribution of innovation resources and potential is not even… [in the case of emerging economies]” (1109), we take a country-specific approach to better deal with such an unevenness and account for host-country idiosyncrasies. Moreover, since MNEs’ decisions to undertake R&D in specific locations are conditioned by both firm- and location -related factors, a context-specific view of the pattern of MNEs’ evolution in their local activities is essential. As each MNE is a nexus of knowledge flows across home and host countries (Gupta & Govindarajan, 1991), how they pursue certain knowledge objectives by organizing specific R&D activities with local firms, research institutes, and universities in a host country becomes an important question. Nevertheless, studies adopting an evolutionary perspective to explore the pattern of R&D internationalization in the South remain limited. Jha et al. (2015), Jha et al. (2018) and Kumaraswamy et al. (2012) are rare exceptions; however, all focus on India and cover a maximum period of ten years.

In sum, drawing on the literature on R&D internationalization, this article attempts to address the key question of how MNE R&D activities in the South have evolved to play an increasingly active role within the MNE global innovation network, so that new insights into the latest shift in R&D internationalization from the North to the South can be gained. Here, we take a historical lens to identify empirically the potential pattern of change in subsidiaries’ innovation activities of a specific MNE, AZ, in the case of a specific Southern economy, China. The next two sections first discuss the specifics of the case study approach (AZ in China), and then provide our empirical findings covering the period between 1993 and 2017.

Epistemology and methodology

To achieve a holistic understanding of the overall development pattern of R&D internationalization in our case study, we decided to adopt a longitudinal historical perspective, which echoes a similar approach taken by Jha et al. (2015) and Kumaraswamy et al. (2012). We follow Turnheim and Geels’ (2013) process-tracing approach to explore the evolution of AZ in China from its initial entry in 1993 to 2017. This approach is particularly useful in capturing the evolutionary pattern we seek to unfold in this study as it “takes path dependence seriously and goes ‘inside the black box’ to explain how actions and changing contexts produce event chains” (Turnheim & Geels, 2013: 1754). We adopt an epistemological style commonly used by historians, ‘‘narrative explanation’’, which is particularly strong in capturing complex social interactions through time and event sequences (Abbott, 2001; Gill, Gill, & Roulet, 2018).

In line with Turnheim and Geels (2013), the narrative explanation style that we adopt in this article is not about simply describing separate events, but involves the tracing of the historical process, guided by prior literature and the proposed conceptual framework (Fig. 1). On the other hand, “historians rarely perform formal tests of theories; instead, they judge theories and conceptual frameworks for their usefulness in writing plausible and interesting narratives” (Turnheim & Geels, 2013: 1755). Therefore, using a ‘‘thick’’ descriptive case study (Child, Rodrigues, & Tse, 2012) allows us to derive insights related to existing knowledge (Doz, Santos, & Williamson, 2001: 584). Welch et al. (2011) proposed that contextualization should not be viewed as a limitation of case study research, but, on the contrary, as a bridge to ‘‘reconcile’’ the positivist views on the generalizability of findings, which is regarded as fundamentally opposed to ‘‘explaining’’ and ‘‘contextualizing’’ (743). Applying this to our study, we argue that the contextual details of our case are simply too important to ignore. Thus, in order to fully understand the evolutionary pattern of R&D activities of AZ in China, it is important to complement a ‘‘positivist context-free orientation’’ with a more contextualized explanation, highlighting causal relationships. The objective of our study is thus to provide an analytical narrative that traces the pattern of R&D internationalization identified in the literature, but then emphasizes the impact of context, thus becoming an ‘‘explanatory’’ tool of understanding, hence departing from its initial ‘‘exploratory role’’ (Yin, 2014).

Within case study research, researchers can choose to select either single or multiple cases. The decision on which to use is largely dependent on its appropriateness in dealing with the topic (Fletcher, Zhao, Plakoyiannaki, & Buck, 2018; Yin, 2014). As the topic of this study is about drawing on rich contextual details to unfold an evolution pattern, a single case study of AZ in China is chosen. Siggelkow (2007) emphasized the usefulness of a single case study as an important source for both unique data evidence and theorizing. Our chosen case is empirically appropriate for several reasons. First, UNCTAD’s (2005) ranking of innovation-related FDI shows that China has become a top inward innovation FDI destination of the South. This is confirmed by a recent report by PriceWaterhouseCoopers (2015) that also identifies China as a recipient of some of the largest global R&D investments. Second, among many R&D intensive MNEs, AZ PLC has one of the oldest operating histories in China and was one of the very first to set up a strategic R&D center in the country (Bernstein Research, 2009), thus providing the longest and richest story.

Case study rigor

In this article, we present a rigorous single case study, relying on the strategies, principles and coding rules described in Cook and Campbell (1979) and Gibbert and Ruigrok (2010). Thus, in Table 2 we present the four validity and reliability criteria, namely internal validity, construct validity, external validity and reliability. These criteria are considered as the ‘‘primary reports’’ along the horizontal axis. We also discuss the research measures taken in our study for each corresponding criterion, the ‘‘secondary reports’’ along the vertical axis, drawing on Yin (2014).

Table 2 Primary and secondary reports considered
Full size table

With reference to Turnheim and Geels’ (2012, 2013) definition of secondary (journal articles and authored books) and primary (newspapers, government reports and company annual reports) data, we draw on a wide variety of secondary and primary sources to highlight and trace activities of AZ during its operations in China. Although these sources cannot always be taken at face value, they do provide some interesting facts for our qualitative analysis, which tracks the gradual changes in the company’s operations and it can be said that “This variety in data sources enables triangulation, and allows for a rich analysis…” (Turnheim & Geels, 2013: 1755). A comprehensive list of data sources is provided in Table 3.

Table 3 Primary and secondary data sources
Full size table

Data analysis protocol

The data analysis process was partially planned and partially emergent as the data collection progressed. Appropriate adjustments were made throughout the study by examining emerging patterns from the raw data, thus making the process interactive. Following a similar approach as in Kumaraswamy et al. (2012), our longitudinal case is divided into three distinctive periods (covering waves 2 to 4 of our framework), using the narrative to depict AZ’s innovation activities in China. Specifically, period one displays an initial focus on market expansion and production in China together with the necessary low-level R&D activities for local adaptation (i.e., a North-South wave), while period two features a new focus on local and regional innovation and production, and period three is characterized by a shifting focus towards global-driven innovation. We then provide interpretations of those R&D activities undertaken in each period, as well as over the course of the three periods mentioned above in relation to our proposed overarching framework.

Case company: Brief history of AZ PLC

The beginning of AZ can be traced back to 1913 when Astra was an independent company based in Sodertalje, Sweden. It was not until the 1930s, however, that the company started its first small-scale research activities. The company acquired a couple of factories in 1939 and 1942, helping to make Astra the largest Swedish pharmaceutical company. From the early 1990s, Astra started to experience significantly increasing costs for the development of new drugs. Believing that the company needed a more international platform for the future, Astra started to look for partners. It was not long before Zeneca stood out as the ideal partner (The Independent, 1998). This partnering was one of the largest-ever European mergers at the time and made the combined company into the fourth largest pharmaceutical company in the world. The merger also signalled the company’s intention to strengthen its innovation and world market share in the long run.

The integration of the two companies’ worldwide organizational activities post-merger called for some major structural changes and, consequently, led to the formation of a new group arrangement, whereby a much-widened geographical scale was realized. Specifically, while a new focus was starting to be placed on the Asia Pacific, Latin America and the Middle-East as younger markets with greater growth potential in the long term, the US, Canada and Europe remained the three largest and most mature markets for the group. Hence, across these six markets, the overall emphasis was still placed on the North and three worldwide strategic innovation centers in Sweden, the UK and the US were responsible for the most advanced drug research for the Northern market. This period of operation can be characterized as a part of the North-North wave, with the main objective of HBKA-N. The next section focuses on exploring R&D activities of AZ in China in the context of the North-South, South-South, and South-North waves, and HBKE-S, LBKE-S, and HBKA-S objectives respectively.

Empirical findings: AstraZeneca’s R&D internationalization in China, from 1993 to 2017

Three distinctive periods in AZ’s internationalization of R&D activities in China can be identified. Below, we discuss each of these periods and the corresponding waves and innovation activities in more details.

Initial expansion in China (1993–2001): The North-South wave and HBKE-S

Since the announcement of ‘‘open door’’ policy in 1978, China has carried out major economic reforms to move towards a more market-based system of trade and economic growth and to open up for foreign investments in low value-adding and factor-endorsed areas–a reflection of the abundance of low-cost labour and the absence of national innovation capabilities. This shift drastically accelerated China’s economic expansion and was the deciding point for Astra to establish its first sales and marketing subsidiary in the country: Astra (Wuxi) Pharmaceuticals Co. Ltd., in Jiangsu Province, not far from Shanghai (Pederson, 2008; World Scientific, 2003).

Jiangsu was an ideal location for Astra for a number of reasons. First, in 1992, many parts of Jiangsu Province were turned into economic zones and Wuxi city established Wuxi New District–one of the largest industrial parks in China. The new district was known for providing strong support for international operations. Second, Jiangsu Province is ideally located next to trade intensive regions, such as Zhejiang, and is home to hundreds of thousands of businesses. Third, the local government offered many incentives to attract FDI. Hence, the overall business environment of Wuxi made it the ideal location in China at the time of Astra’s entry.

In the following year, Zeneca also entered China through a joint venture with Sinopharm, a state-owned corporation. Sinopharm, also known as China National Pharmaceutical Group Corporation, was the largest pharmaceutical and healthcare group in China at the time and already had a well-established network in the country. The initial agreement between Sinopharm and Zeneca was for the sole distribution of Zeneca’s leading drugs at the time (Heathcote, 1994; Zeneca Group PLC, 1994). Similar to Astra, this move was also an indication of Zeneca’s attempt to enter into the Chinese market early as part of their global expansion strategy for existing products. One year into the agreement, Zeneca realized the huge potential of the Chinese market and established its own sales and marketing operations in 1995. Benefiting from the rapid economic growth in China in the following years, Astra and Zeneca both successfully positioned themselves in the Chinese market to be leaders in the sale of specific drugs. They also set up offices in over 20 major cities. However, Zeneca’s Chinese joint venture demerged around the same time as Astra and Zeneca merged.

Following the merger between Astra and Zeneca, the new AZ embarked on its most significant Chinese venture at the time, that is, an accumulative investment of US$270 million (in comparison to a US$121 million investment by GlaxoSmithKline around the same time) in building its first world-class production facility in Wuxi New District, where Astra’s previous Chinese operations had been initiated (Chemical Market Reporter, 2001). The new production plant employed 600 workers and produced 95% of all AZ’s products sold in China. It is worth noting that while there are other pharmaceutical MNEs with local production facilities in China, AZ was seen to have the largest facility in terms of the scale of investment and the facilities created (AstraZeneca PLC, 2001; Wall Street Journal, 2001a). This reflects a strong strategic emphasis on HBKE-S in the country and first sign of the North-South wave.

During this period, AZ also made its first investment in R&D in the country–another step under the North-South wave. Between 1996 and 2001, AZ undertook nine international multicenter clinical trials in the respiratory field in China, with the involvement of over 130 domestic hospitals and institutions. By 2001, AZ had invested US$35 million in clinical trials and conducted 37 clinical research projects involving approximately 20,000 patients at more than 1,000 domestic medical sites. The objectives of these investments were to develop drugs that were effective for Chinese patients with specific diseases. To further its efforts in local innovation, AZ emphasized the importance of strategic partnerships with Chinese research institutions. For instance, AZ and Shanghai Jiaotong University launched a joint research project to investigate the genetic links to neuropsychiatric diseases (AstraZeneca PLC, 2001; Wall Street Journal, 2001b).

Although the research-related activities during this period were not intended to include building its own innovation facilities, AZ was still one of the first foreign companies to commit to market expansion in the form of HBKE-S at the time. This move benefited the company for two reasons. First, China was considered a marketplace where mature drugs could still generate sales long after patent expiration. Second, the common practice in China for drug selection was that once particular branded drugs were accepted by the doctors and patients, there was a strong tendency to stick to them permanently (AstraZeneca PLC, 2010; China Food and Drug Administration, n.d.). Hence, the development pattern of this period signals a heavy overall strategic emphasis on market expansion through HBKE-S rather than any other R&D activities, possibly due to the weak institutional conditions at the time, such as the lack of availability of R&D resources and government support policies.

Second period of expansion in China (2002–2008): The South-South wave and LBKE-S

Since 2002, there were two major developments in China. One was the continuation of a series of innovation programs by the government, including the International S&T Cooperation Program of China; National Key Technologies R&D Program; and the Torch Program for supporting commercialization of high tech products (China Ministry of Science and Technology, n.d.). The other was the fact that the country officially became a member of the World Trade Organization (WTO). This had two major effects–encouragement for increased inward FDI and greater transparency in legal protection for foreign investors. Consequently, AZ deepened its innovation involvement in China (AstraZeneca PLC, 2008; Ministry of Commerce, P.R.C., 2005). Despite China being viewed by most foreign investors as a more attractive destination for production and sales rather than research, extensive local partnering (Table 4) and the establishment of its own Chinese-based research institutions demonstrated AZ’s shift of focus in its operation in China from HBKE-S towards LBKE-S, which is indicative of the South-South wave.

Table 4 List of local partnerships
Full size table

Specifically, AZ took a major step in 2003 by forming its most significant partnership at the time with Peking University’s Guanghua School of Management to fund the China Centre for Pharmacoeconomics and Outcomes Research. This was the first in a series of research and educational programs aimed at supporting reform of China’s healthcare system. During the first three years of the partnership, AZ provided over US$360,000 in sponsorship to help set up and establish research, advisory and training programs for government officials, hospital executives and pharmacy directors.

AZ took another major step in its local development by forming a strategic partnership with Peking University Third Affiliated Hospital in 2007 to set up the company’s first Clinical Pharmacology Unit in the country (Asiabiotech News, 2007). In the same month, AZ and Guangdong Provincial People’s Hospital jointly established a research laboratory with a focus on translational science. The unit was intended to augment existing clinical research capabilities and undertake Phase I clinical research, including clinical pharmacology and safety evaluations steps necessary for launching new medicines in China. This move accelerated the launch and patient access to new medicines in China. In 2008, AZ reached out to the Shanghai Institute of Materia Medica to establish the Drug Safety Evaluation Joint Research Centre in Shanghai Pudong Science Park of the Chinese Academy of Science. This was also intended to accelerate the speed to market for drugs in the country. In the same year, the agreement between AZ and Pharmatech Wuxi came to a successful end, achieving its intended targets two months ahead of schedule, but AZ decided to extend its contract with the company to continue to synthesize compounds according to AZ’s designs to further expand the company’s global compound collection.

In terms of establishing its own research institutions, AZ set a number of historical records. For example, AZ was the first company of its kind to launch a localized Clinical Research Unit for the East Asian region in Shanghai in 2002. The significant number of high quality medical talent available in the city was the reason for the location decision. The unit’s aim was to oversee East Asian clinical research in mainland China, Hong Kong, Taiwan and Korea, with the clinical research conducted by more than 40 high-calibre Chinese researchers. This signalled that the country would become a critical component of AZ’s future global trials and it had the potential to significantly reduce the time it took for drugs to be made available to Chinese patients (AstraZeneca PLC, 2002; Ministry of Commerce, P.R.C., 2005).

Another important development came in 2006 when AZ invested more than US$70 million in the establishment of the Innovation Centre China, which again was the first of its kind in the country (AstraZeneca PLC, 2006; China Pharmaceutical Industry Investment Promotion Report, 2014; Ministry of Commerce, PRC, 2006). It was also one of the most advanced scientific research centers of the Group outside of the UK, Sweden or the US and provided firm evidence of the company’s dedication to research in China, as well as its ambition to have a strong innovation presence in the region. Although the beginning of AZ China did not seem to suggest so, the formation of this innovation center illustrated that this consideration played a major role in contributing to AZ’s global knowledge base. Apart from innovation activities, AZ further expanded its production scale in 2006 with an additional US$35 million investment in its Wuxi site, which also obtained China and European Union Good Manufacturing Practice certification for exporting to the European Union and other countries adopting the same standards. This development was the foundation of AZ China’s important global role later.

In summary, by 2,009, AZ employed more than 2900 staff locally in China, had a network of more than 20 marketing and sales offices, a global manufacturing site in Wuxi, clinical research facilities and multiple collaborations and partnerships with local academic and medical institutions. Their major projects during this period reflected an emphasis not only on market expansion through HBKE-S, but also on LBKE-S to deepen its involvement in the Chinese market in the long run. Apart from improved institutional conditions (e.g., the availability of R&D resources and government support policies), the company’s increased local knowledge and experience is likely to have been a reason for the changing emphasis.

Third period of expansion in China (2009–2017): The South-North wave and HBKA-S

This period witnessed the most significant development in terms of localized and global innovation. It also saw the strengthening of the Chinese government’s policy and funding support for more advanced innovation activities, including National S&T Major Projects for priority industries, the National S&T Innovation Base Program for establishing new research laboratories and bases, and the establishment of the National Natural Science Foundation of China for improving the funding environment for basic research (China Ministry of Science and Technology, n.d.). During this period, AZ China deepened its involvement in LBKE-S but also initiated activities associated with HBKA-S. In seeking LBKE-S, AZ opened its new Chinese site in Shanghai Zhangjiang Hi-Tech Park (which was commonly known as ‘‘China’s Medicine Valley’’) in September 2009, which included AZ’s new Chinese corporate headquarters for China and the Asia Pacific, its marketing operations, its Innovation China Centre, and academic training and regional functions. It became one of only three corporate headquarters of AZ PLC alongside the USA and the UK (AstraZeneca PLC, 2009; China Pharmaceutical Industry Investment Promotion Report, 2014). By the end of this financial year, AZ’s accumulated sales in China were more than US$800 million. AZ was confident that with a population of 1.3 billion in China, a burgeoning middle class and an increasing prevalence of Western-style diseases, the Chinese market offered the largest opportunity for growth, and was convinced that new product offerings were the way to further strengthen its position (Reuters, 2010).

Hence, in order to broaden and improve its product offerings, AZ continued its aggressive innovation localization in a number of ways. First, AZ strengthened its efforts in internal innovation capability building in China. For instance, by 2011, its Innovation China Centre had already built strong capabilities in translation sciences for oncology research for the Chinese market. As a result, the center embraced the new mission to deliver candidate drugs and ultimate Proof of Concepts and valuable drugs to address the significant unmet medical needs of patients in China, such as for liver and gastric cancer (McKinsey, 2012).

Second, AZ also brought in external capabilities to complement its core internal expertise already available at its China center. One example was an acquisition agreement with a privately-owned generics manufacturing company (BeiKang Pharmaceutical Company Ltd) in 2011, which gave AZ access to a portfolio of injectable drugs used to treat infections suffered by Chinese and Asian patients. This move reinforced the company’s commitment to bringing original and high quality branded generic drugs to the Asia-region market to increase their accessibility and affordability for patients who were previously underserved.

More importantly, AZ continued to forge extensive and advanced research collaborations with local partners in China for medical breakthroughs. This move was made to not only address the medical needs of the Southern market, but also to seek opportunities to meet global market needs. A list of some of the major collaborations during this period is provided in Table 5. During this period, AZ China entered into several research collaborations for developing drugs for the global market. This move signalled AZ China’s first move towards becoming a global contributor of innovation the start of the South-North wave. For instance, in 2011, AZ and Chi-Med China joined together to co-develop and commercialize a novel cancer therapy called Volitinib (which later turned out to be successful) and, in 2014, it signed an agreement with Tianjin Medical University for cardiovascular research collaboration to explore novel targets against cardiac fibrosis. Meanwhile, products such as Tagrisso and Brilinta were the outcomes of R&D collaboration between AZ China and AZ’s R&D sites worldwide (Greig, 2016; Zhou et al., 2017). By the end of 2014, AZ had recorded over 1500 patents with the State Intellectual Property Office in China (SIPO, 2015). Thus, the increased product offerings and strengthened research collaborations with local partners in the third period show a greater emphasis on local innovation for the local and global markets. This was built on more favorable local conditions and AZ’s greater local knowledge and experience.

Table 5 List of collaborations
Full size table

From 2016 onwards and in alignment with the introduction of the Healthy China 2030 initiative (Biopharmadive News, 2018; Tan, Liu, & Shao, 2017), AZ undertook several major advanced R&D projects which further reflect the importance of China as an important R&D location for meeting the needs of both Chinese and global markets. An example worth noting is the strategic alliance AZ formed with WuXi AppTec to produce next-generation medicines by undertaking advanced research in the area of new small molecule development, which is part of AZ’s global R&D network and core to AZ PLC (Washington Business Journal, 2017). This project aimed to address the needs of both Chinese and global patients in the main therapy areas of respiratory, inflammation and autoimmunity, cardiovascular and metabolic diseases, as well as oncology. This partnership required not only financial investment but also recruitment of at least 50 world-class scientists to establish a new R&D hub for small molecule research, making China an important location alongside AZ’s UK and Sweden R&D centers (AstraZeneca PLC, 2016; Financial Times, 2015). By 2017, AZ showed a further 7% growth in sales compared to its competitors. In 2017 alone, a number of key developments took place, illustrating the increasingly important role of AZ China in the global innovation network. For one, the formation of the strategic joint venture of Dizal pharmaceuticals with the Future Industry Investment Fund for discovery, development, and commercialization of new medicines for unmet global needs cemented the expansion of AZ’s R&D in China. This occurred at a time when AZ’s competitors were consolidating their operations by closing major R&D facilities (Fiercebiotech News, 2017). In late 2017, the research collaboration with Chi-Med led to a global clinical trial for a rare type of kidney disease found worldwide (Chi-Med, 2017; Reuters, 2017). Further, the research collaboration with Ironwood Pharmaceuticals, Inc. led to an application to the China Food and Drug Administration for the development and marketing of linaclotide (Ironwood Pharma News, 2018), whereas the launch of PT010 for COPD and asthma in China and Japan, ahead of their launch in Europe and the US, signalled the rise of China as a strategic location for knowledge creation (Fiercebiotech News, 2017). Most recently, AZ China responded to Chinese institutional reforms in the implementation of the ‘‘digital China strategy’’ by initiating collaborations with Wuxi National Hi-Tech District and Jiangsu Mobile as well as with China’s technology giants, Tencent and Alibaba, to provide ‘‘smart’’ healthcare services driven by artificial intelligence and internet-enabled technologies (Biopharmadive News, 2018; China Daily News, 2018).

To summarize the three periods analyzed above, Fig. 2 illustrates the major changes in AZ’s activities in China between 1993 and 2017. In the context of the proposed waves framework (Fig. 1), we can identify a case-informed evolutionary pattern. Period one can be characterized as the North-South wave for the main purpose of market expansion through HBKE-S, reflecting a market-seeking motivation. Period two maintained activities of period one while the focus shifted to LBKE-S for meeting distinct local and regional market demands, a reflection of the South-South wave where knowledge-seeking activities emerge. Finally, period three maintained activities of the first two periods while the focus shifted to HBKA-S for meeting global market demand, a reflection of the South-North wave. This became a new variant of knowledge-seeking activities aimed at developing and producing new products that were expected to supply the global market.

Fig. 2
figure 2

R&D activities of AZ China between 1993 and 2017

Full size image

Discussion

AZ’s R&D evolution in China

Having presented AZ’s innovation activities in China in three consecutive periods over the course of more than two decades, we now reflect on each of the three major waves and the overall evolution pattern identified in the case study in light of the proposed framework (Fig. 1). We highlight our key findings in order to develop new insights into the latest development in R&D internationalization towards the South.

Period one (North-South wave and HBKE-S)

During the period of 1993 to 2001, China remained a location with two distinct characteristics: an emerging market and an abundance of low-skilled labour for low-cost production. Furthermore, we highlighted that although China became a location for AZ’s R&D activities, these activities were mainly an ‘‘operational necessity’’ (Bartlett & Beamish, 2013; Cantwell, 1992; Pearce & Papanastassiou, 1997) and were restricted to product and production-related HBKE-S (Bas & Sierra, 2002; Kuemmerle, 1999a; Liu & Chen, 2012; von Zedtwitz & Gassmann, 2002). It could be argued that findings from this period also reflected the wider environmental conditions in China at the time, which could be characterized as having significant weaknesses in the institutional conditions for more advanced R&D activities. During this period, the country showed little or no signs of the availability of R&D resources or supportive government policies.

Specifically, our case study findings show that most of the activities of AZ China during this period were related to assimilating knowledge transferred from the headquarters and modifying existing products to meet the local market conditions in China through extensive localized clinical trials. These activities of product adaptation and localization were achieved in the form of in-house R&D support and collaboration with local Chinese hospitals to ensure safety and suitability of drugs to Chinese patients. Similarly, Zhang and Pearce (2011) and Zhang et al. (2018) note that market-seeking subsidiaries in China incorporated a strong element of product adaptation in local operations. Thus, in order to meet the idiosyncratic demand conditions in China, local sourcing of technology became an essential precondition. Furthermore, our empirical findings also concord with the works of Schmiele and Mangelsdorf (2009) and Liu and Chen (2012) in that MNEs can gain economic returns from utilizing their existing proprietary knowledge (generated in AZ’s global R&D centers) in overseas locations (in this case, China).

Hence, we argue that AZ’s internal motive for entering China during this period was market expansion, which was strongly facilitated by two co-existing factors at the time: ownership advantages and local conditions (Bartlett & Beamish, 2013; Dunning, 1998). Specifically, the long-established R&D base in the North provided AZ the knowledge-based competitive advantages it needed to target markets of the South in pursuit of market-seeking investments (Cantwell et al., 2010; Kuemmerle, 1999b). In contrast, indigenous firms in China were severely lagging behind in terms of innovation. At the same time, China’s economic liberalization started to show positive signs, such as growing consumption (Fan & Wang, 2001). This development in the local institutional conditions provided AZ with the opportunity to potentially take over the market by way of offering drugs that were produced with knowledge from the North and labor from the South.

Period two (South-South wave and LBKE-S)

While AZ continued to further invest in market expansion activities in China during this period (2002–2008), benefiting from fast economic growth, the company also showed first signs of local-based R&D activities. That is, on top of its existing R&D activities relating to local adaptation in period one, the company initiated new projects into advanced drug testing and discovery specifically targeted at local diseases in China. This is in line with the work of Zhang and Pearce (2011) who found that MNEs in China are not solely receivers of headquarters knowledge, but also generators of new knowledge by working closely with local universities, firms, and research institutions in their effort to augment their market-seeking activities. Following our proposed framework, this development in the case of AZ China signals an important evolutionary step along the continuum of AZ China’s pattern of R&D internationalization, where the North-South wave (identified in period one) shifts towards the South-South wave through LBKE-S. China, as an exemplar of the South, is no longer viewed only as a source of low-cost labour, but also a new source of R&D knowledge (Awate et al., 2015; Jha et al., 2015; von Zedtwitz et al., 2015). Knowledge-seeking in the South emerges as a priority for the company in this wave. Subsidiaries in the South have started to play the role of generating new knowledge, as also evidenced by Zhang et al. (2018: 10) who indicated that “the main sourcing of new technologies is thus expected to be well diversified and come primarily from the [emerging-economy] subsidiary… as well as from subsidiary collaborations with local firms and research institutions”.

The case of AZ China thus shows that this period could be characterized as the seeking and development of knowledge useful in providing new products to the local market. There are two fast-changing forces that contribute to this trend: improvements in both market and institutional conditions in China. In terms of market conditions, the growing market demand and limited availability of effective products called for localized innovation. In terms of institutional conditions, increased availability of R&D resources and supportive government policies provided better conditions for undertaking localized innovation. The imperative to adapt to specific conditions in the Chinese market also concords with the propositions by Zhang and Pearce (2011) that suggest a variant of market-seeking investments whose aim was to allow for the adaptation of these technologies “to local tastes or needs, to cultural mores and to indigenous resources and capabilities” (Dunning & Lundan, 2008: 70). Though such adaptation also occurred in period one, the way it was implemented at the subsidiary level in period two, and the different sources of technology that were invoked to secure it, may establish a more important role. For instance, our findings show that AZ was able to benefit from increased availability of high quality local medical talent and set up many R&D collaborations with local universities and firms, facilitated by government initiatives. Our findings further show that AZ’s localized innovation in China enabled the company to better understand local diseases, conduct trials, and produce more effective drugs for Chinese patients. Consequently, the company was able to retain a secure market share and sustainable economic returns.

Hence, we argue that although AZ’s internal motive for continuing to establish itself in China during this period was local responsiveness, this was achieved through knowledge-augmenting rather knowledge-exploiting R&D (Bartlett & Beamish, 2013; Dunning & Narula, 2005). Specifically, the long-established R&D base in the North was no longer the sole source of competitive advantage (Kuemmerle, 1997). Instead, working with local counterparts during period one enabled AZ to learn more about both the market and the institutional environment (Zhang & Pearce, 2011). Local knowledge access thus became an important channel for enhanced ownership advantage. While in the first period there is a clear dichotomy between MNE ownership advantages (in the form of established home-based knowledge inputs transferred to China) and location advantages (in the form of distinctive demand conditions and the supply of cheap labor and technical skills), the enhanced ownership advantage was the outcome of the interactive collaboration between the company and local knowledge sources, aiming at knowledge-seeking investments (Bartlett & Beamish, 2013). Demand for better products increased significantly in this wave whilst upgrading of knowledge capabilities and improvement of China’s legal framework also progressed well. Thus, AZ’s shifting internal motive was facilitated and supported by the dynamic and interdependent evolution in ownership and location advantages contributing towards “the dynamic competitiveness and locational strategy of firms, and particularly the path dependency of the upgrading of their core competencies” (Dunning, 2000: 167).

Period three (South-North wave and HBKA-S)

The analysis of period three (2009–2017) data shows the most striking development of AZ China to date. The company intensified its South-South and LBKE-S commitment by innovating for not only China, but also the wider Asian region, both through utilizing its existing internal R&D capabilities as well as by collaborating with and acquiring local companies (e.g., Beikang Pharmaceutical Company Ltd). More significantly, during this period, the company showed the first signs of undertaking world-leading R&D for the purpose of addressing knowledge gaps in drug discovery for some of the world’s most complex and challenging diseases (via its Innovation China Centre). This finding is in line with existing literature, which suggests that MNEs establish R&D units in locations where they can access host countries’ knowledge for their headquarters (Asakawa & Ashok, 2008). The main objective of these units is to engage in knowledge sourcing and benefit from potential knowledge spillovers (Almeida & Kogut, 1999; Feinberg & Gupta, 2004). At this point, the South-North wave and HBKA-S in support of the development of complementary and novel knowledge for the global market became evident. This development supports the work of Govindarajan and Ramamurti (2011), Govindarajan and Trimble (2012) and Jha et al. (2015) in that R&D units under the South-North wave display the development of long-term innovative capabilities by taking advantage of novel or complementary knowledge in foreign locations. In the case of AZ China, this development shows that MNE subsidiaries located in non-conventional locations can gradually develop strong innovation capabilities and evolve towards being a more active and differentiated player in MNEs’ global innovation networks (Mudambi & Santangelo, 2016). We argue that the likely cause of this evolution as reflected in the AZ China case is further accumulation of the firm’s knowledge on how to best maximize the opportunities offered by improved institutional conditions, such as local R&D firms, R&D talents and preferential policies. For example, our findings of AZ China show that the development of its distinct knowledge base is contingent upon the availability of local R&D resources. The company was able to form long-term partnerships with local universities (e.g., Tianjing Medical University) and research institutions (Chi-Med China), and strategic joint ventures with government funding bodies (Dizal Pharmaceuticals), filling knowledge gaps in the existing global pharmaceutical market. This concords with Liu and Chen (2012) in that a MNE host-country R&D mandate in the South is likely to be influenced by the changing local conditions, changes that in the AZ case had much to do with increased efforts of national and local governments to establish pro-innovation support (Archibugi & Iammarino, 2002; Lundvall, 1992, 2007). This is evidenced by their investment in higher education and training of R&D personnel (e.g., AZ was able to utilize a skilled scientific personnel available in various hospitals and research institutions), the development of science clusters (e.g., AZ was able to benefit from locating in China’s Medicine Valley of Shanghai Zhangjiang Hi-Tech Park), and specific policy support (e.g., AZ was able to deepen its involvement in the development of China’s national healthcare system with the strong support and facilitation from the government). Another recent investment AZ China made was in the smart healthcare services by collaborating with local technological firms. With respect to AZ’s R&D evolution in China during the third period, it is thus fair to argue that the North-and-South location hierarchy for MNE R&D activities is flattening. The South is seen to become an important part of MNEs’ global innovation network for knowledge augmentation (Chen, 2004; Liu & Chen, 2012).

Hence, we argue that AZ’s internal motive for deepening its involvement in China during this period was worldwide learning (Bartlett & Beamish, 2013; Zhang et al., 2018). This strategic shift from market expansion (as found in period one) and local responsiveness (as found in period two) can be attributed to further changes in ownership advantages and local conditions when compared to the two earlier periods. Through the launch of a series of national S&T programs, a supportive and attractive National Innovation System (NIS) was promoted. Campbell (2013:4) noted that “the private sector (or semi-private sector) has now taken the lead in innovation from SOEs, and relies on foreign funding along with government support for technology projects. Government-supported research institutions are now spread throughout the country, but innovation leadership in key centers, especially Beijing, Shanghai, and Guangzhou, has passed to private firms.” It was clear that in such a supportive environment AZ was able to gain local knowledge at a greater scale and scope than before, and combine local and global knowledge for enhanced worldwide ownership advantages. It provided AZ the knowledge-based competitive advantages it needed to target not only the Chinese, but also the global market.

In summary, the discussion of the three consecutive waves of innovation activities in the case of AZ China clearly suggests a dynamic and path-dependent pattern (from North-South to South-South, and to South-North) that concords with our proposed framework (Fig. 1). The contextual details of each period reveal evidence that AZ’s initial intention of setting up operations in the South (in this case China) was not concerned with strategic knowledge sourcing. It was only through both market and institutional transformation and organizational learning that innovation activities started to be considered seriously (Campbell, 2013). In this very case, our findings imply a possible developmental pattern of co-evolution that has taken place between AZ China and China’s various institutions. Drawing on the co-evolutionary perspective (Cantwell et al., 2010; Child et al., 2012; Rodrigues & Child, 2008), we showed that under conditions of limited organizational knowledge and weak institutions, the activities of AZ China started with the necessary knowledge transfer from headquarters for local production in the early period. This process also ‘‘naturally’’ contributed to the innovation capability of local counterparts who were involved in AZ’s operations in China at the time. As market and institutional conditions improved (particularly with much stronger involvement and support from the Chinese government) during the second period, innovation activities were upgraded to encompass localized product development and customization for the Chinese market. Meanwhile, this deepening relationship between China and AZ resulted in the Chinese environment becoming increasingly supportive to AZ in the way of new opportunities. This in turn led AZ to invest more capital in China in order to develop R&D capabilities, creating a two-way development pattern for both the Chinese pharmaceutical industry and AZ China. During the most recent period, apart from continuous improvement in market and institutional conditions, AZ’s deepened involvement in, and knowledge of, China enabled the company to undertake novel and complementary R&D activities to meet global market needs. At the same time, AZ China contributed to the Chinese pharmaceutical industry’s innovation capability development with the support and opportunities offered by the government.

The empirical evidence of the three periods (Fig. 2) illustrates that each period is contingent upon the preceding period and AZ’s relationship with China. The explanation for this gradual development is likely to be twofold. First, at the firm level, factors such as organizational experience and knowledge of the host-country environment are likely reasons for AZ’s growing confidence and R&D commitment in China. At the locational level, factors such as changes to market and institutional conditions of the host country are another likely contributor to the evolution of AZ’s R&D activities in China.

In other words, drawing on the co-evolutionary perspective (Cantwell et al., 2010; Child et al., 2012; Rodrigues & Child, 2008), the AZ China case demonstrates the importance of the linkage between local conditions (external factors) and firm knowledge and motives (internal factors) as potentially co-existing determinants of R&D-related FDI in the South. The local market potential and production benefits in period one attracted AZ to set up sales and manufacturing operations in China, benefiting from HBKE. As AZ became more embedded in the local market with improved market knowledge and an increasingly innovative institutional environment (particularly with the introduction of national S&T programs), they were able to utilize some of their established local networks (such as partnerships with local hospitals and universities) to carry out higher-value-adding R&D activities. The strategic objective for AZ China thus shifted towards LBKE-S during period two. Further deepening of AZ’s relationship with local counterparts (local firms, research institutions, hospitals, universities, and patients) and continuously improving institutional conditions (government support such as the innovation programs introduced in period two and three) provided AZ China the opportunity to start seeking HBKA-S during period three. Each of the three periods was successful in achieving AZ’s objectives in China and success in each period paved the way for the next period as both firm knowledge and the local environment continuously improved. Thus, we argue that it is the necessary complementarity between internal firm knowledge and the external local environment that has influenced AZ’s objectives and performance in China throughout the three periods. The AZ case thus perfectly illustrates the co-evolutionary developmental pattern introduced by prior research (Cantwell et al., 2010; Child et al., 2012; Rodrigues & Child, 2008).

Contributions to R&D internationalization theory and literature

Our longitudinal case study of AZ’s historical development over the past two decades, with specific reference to the context of China, has offered a rich, contextualized example of a possible emerging pattern in MNE global innovation network arrangements. It can be argued that R&D internationalization by MNEs of the North to economies in the South is too idiosyncratic to be explained by established theories. Most of current understanding is predominantly rooted in the context of advanced economies and far fewer studies have explored the contextual details of R&D evolution from the perspective of the South. Against this background, our study makes a number of relevant contributions.

First, we contribute to the current literature on MNE global innovation networks by constructing a framework that integrates four distinct waves and corresponding R&D objectives to create a more complete picture of the internationalization of R&D (Table 1). Prior developments in an advanced-economy context did not offer the opportunity or need for such definitions. However, the recent shift towards knowledge sourcing from new locations in the South makes such a systematic classification important to inform future research. Moreover, our case study findings are found to concord with the four major waves and the corresponding R&D objectives. Furthermore, whereas in general, North-centric waves (i.e., North-North and North-South) are established phenomena, research addressing South-centric waves (South-South and South-North) is only recently emerging (e.g., Awate et al., 2015; Jha et al., 2015, 2018; Liu & Chen, 2012) and still sparse. Hence, our historical longitudinal case study of AZ in China presents new evidence to underpin further discussions on South-centric waves.

Second, we contribute to current theories of MNE R&D internationalization by constructing an overarching conceptual framework (Fig. 1) depicting an evolving pattern of MNE R&D activities shifting from the North to the South. The concordance between the waves framework and our case study findings (summarized in Fig. 2) suggests that the R&D internationalization pattern is composed of three consecutive (and mutually inclusive) shifts in MNE objectives and operations, from HBKE-S (North-South wave) to LBKE-S (South-South wave), and to HBKA-S (South-North wave). The conceptual framework extends our current thinking in understanding the overall development of the MNE global innovation network arrangement by incorporating the latest trends of R&D in the South (Dhanaraj, 2016; Govindarajan & Ramamurti, 2011). Our study highlights a multi-layered and ever more globalized innovation network arrangement in MNEs in search of sustainable competitive advantages. Additionally, it also reveals the possible role of MNE organizational experience and knowledge in capturing local opportunities offered by the fast-changing market and institutional conditions in the Southern economies.

Our third contribution lies in the context of our study, whereby we focused on China by applying a historical lens to exploring this location for MNE R&D. While China is known to represent unique opportunities for MNEs, to our knowledge no studies have traced the evolution process of R&D by an MNE in the country as closely as we did. Our article is one of the few to apply a longitudinal case study approach to reveal contextual details of R&D evolution in the South. In line with Meyer, Mudambi, and Narula’s (2011) and Liu and Chen’s (2012) views on locational differences and the exemplar offered by Wang, Hong, Kafouros, and Wright (2012), we offer insights into the idiosyncratic transitions in China from 1993 to 2017, reflecting its importance and distinctiveness as a research context. Hence, our study highlights the need to appreciate contextual differences among the Southern economies when studying R&D internationalization.

Finally, our article contributes to the recent methodological discussions on qualitative research (e.g. Doz, 2011; Welch et al., 2011). In particular, we illustrate the suitability of a historical longitudinal approach and a single case study design in exemplifying ‘‘context’’ in qualitative research. The case of AZ China illustrates one way of conducting contextually rich research when the target phenomenon is new or difficult to study (e.g. because of company secrecy). The richness of contextual details revealed in this article offers useful empirical insights, and we therefore call for more research to adopt a similar approach to investigating new developments in international business. This is in concert with Verbeke and Kano’s (2015: 415) suggestion that: “A business history lens can illuminate the behaviour of developed-economy multinationals.”

Managerial implications

Our article also has a number of managerial implications. For one, the success of AZ China highlights the potential importance of changing institutional conditions in the case of Southern economies, as well as the possible role of time and experiential knowledge in understanding and deepening R&D-related involvement in the host country. Thus, an important lesson to be learned by other MNEs wishing to tap into the knowledge pool of the Southern economies (e.g., China) is not to rush into making major R&D commitments at the outset, but instead, to proceed through a more gradual process of learning and trial and error. While many MNEs might have some prior operating experience in the South, the national innovation environment is likely to be constantly changing and therefore should be considered as a completely new territory, requiring effective and accumulated learning. A more incremental approach allows for sufficient time to gain a better understanding of the availability of R&D resources, government support policies, and ways to establish and maintain effective partnerships with local research counterparts.

Limitations and recommendations for further research

Despite its important contributions, our study has several limitations that suggest avenues for future research. First, there are some common concerns associated with our case study design. Despite following a methodological approach found in other studies, the use of a single longitudinal case has the limitation that it reflects the development of the chosen case per se and therefore should always be very carefully considered in relation to others (Child et al., 2012). It is thus likely that another case would reveal a somewhat different picture to the narrative we provided in the present article. However, our initial objective was not to provide generalizable outcomes, but rather to offer a contextually rich analysis of a particular case in light of the recent developments, and based on this, to construct a conceptual framework that can serve as a springboard for future research to investigate this increasingly important topic. To extend our understanding of innovation in the South, it would be of great value if future research could explore the actual mechanisms of the evolution process, as well as the key success factors for the internationalization of R&D towards the South. It would also be insightful to investigate the implications of the South-North wave on MNEs, institutions, and countries, as this wave of development makes the global innovation scene much more complex. Next, it would be interesting to see whether or not AZ China’s evolutionary pattern has been followed by other MNEs in either China or other Southern economies. Although we do not see AZ’s innovation activities in China as an oddity or outlier, as figures from UNCTAD (2012, 2016) and a study by D’Agostino (2015) clearly show a strong trend of innovation FDI inflows to the South, whether other MNEs have adopted a similar approach in their engagement with Southern economies is an interesting question to pursue. Confirmatory studies (based on larger samples) or comparative studies (across industries or countries) informed by our research findings would be a useful empirical extension of the present study. We hope that the clear delineation of the R&D internationalization waves in our study, using the North and South paradigm, will motivate further empirical research to generate new insights into the global innovation network arrangement of MNEs.

Conclusion

Using a detailed historical case study, we presented a contextually rich story of the evolution of an MNE’s innovation activities in a Southern economy and evidenced the fast-growing importance of Southern economies as a new destination for knowledge sourcing. We revealed subsequent waves of North-South, South-South, and South-North in the case of AZ China and identified the latest evolutionary pattern of MNE internationalization of R&D, from North-South (HBKE-S) to South-South (LBKE-S), and to South-North (HBKA-S). We argue that this development has critical strategic implications for the configuration of MNE global innovation networks, as they leverage the availability of lower-cost R&D personnel in the South for local, regional, and global competitiveness.