Dual Role of Absorptive Capacity for Exploitative and Exploratory Learning

Boris Radickovic Master of Arts (MA)

Submitted in fulfilment of the requirements for the degree of

Deakin University

June, 2019

Acknowledgement

This PhD study and related research project would not have been possible without the help of some key people that I wish to acknowledge.

Firstly, I would like to express my sincere gratitude to my supervisors Dr Matthew Mount and Dr Heather Round for the continuous support of my research, for thier patience, motivation, and immense knowledge. Their guidance helped me at all times during this research - discussing the topic, building industry connections and writing the thesis. I could not have imagined having better advisors and mentors for my PhD study.

Besides my supervisors, I would like to thank the rest of BL Deakin Business School academics, especially HDR Director - Assoc. Prof. Ambika Zutshi and Prof. Andrew Noblet for support and encouragement during the HDR units and my candidature. I would laso like to acknowledge Prof. Alexander Newman for his insightful comments as a sounding board and interim supervisor following HDR confirmation. My sincere thanks goes to Research Services colleagues (Elizabeth Fitzgerald, Julie Asquith, Vilia Dukas and Penny Love) who provided great support with facilities, workshops and administration.

I have also appreciated inspiring discussions with colleagues at ANZIBA and ISPIM conferences which incented me to widen research from various perspectives, as well as connections with innovation practitioners who kindly participated in this research.

I would also like to thank my fellow EA building colleagues for the stimulating discussions, for the times we were working together before deadlines, and for all the fun we have had in the last three years. Last but not the least, I would like to acknowledge my family, my partner, and all my friends in Australia and overseas, for supporting me emotionally throughout writing this thesis and my life in general.

This research is supported by an Australian Government Research Training Program (RTP) Scholarship.

Table of Contents Abstract ...... 1 Chapter 1 – Introduction ...... 2 1.1 Background and significance ...... 2 1.2 Research problem and research questions ...... 7 1.3 Research objectives and contribution ...... 10 1.4 Research setting ...... 13 1.5 Research design and thesis outline ...... 16 Chapter 2 – Literature review ...... 18 2.1 Innovation – multiple facets ...... 18 2.2 Innovation outcomes...... 21 2.3 Innovation performance ...... 23 2.4 Open innovation ...... 26 2.5 External knowledge search and innovation performance ...... 30 2.6 Dynamic capabilities ...... 33 2.7 Capability development and organizational routines ...... 38 2.8 Knowledge management and Knowledge based view (KBV) of the firm ...... 41 2.9 Absorptive capacity (AC) ...... 45 2.10 Organizational knowledge creation theory ...... 61 2.11 Organizational learning ...... 66 2.12 Intersection – dynamic capabilities, knowledge management and organizational learning ...... 77 2.13 Conclusion and research gap ...... 81 Chapter 3 – Conceptual framework ...... 84 3.1. Overview ...... 84 3.2. Part I – Absorptive capacity ...... 86 3.3 Part II - Hypotheses development ...... 96 3.4 Absorptive capacity and organizational learning (conceptual model) ...... 101 Chapter 4 – Methodology ...... 105 4.1. Research paradigm ...... 105 4.2. Ontology ...... 105 4.3. Epistemology ...... 106 4.4. Research methods...... 108 4.5. Survey methodology ...... 107 4.6. Sample and Data Collection ...... 109 4.7. Measures ...... 110 4.8. Data analysis ...... 114

Chapter 5 – Analysis and Results ...... 121 5.1 Absorptive Capacity – operationalizing and pre-testing measures ...... 121 5.2 Confirmatory Factor Analysis (CFA) ...... 124 5.3 Part II – Data analysis ...... 129 5.3.1 Data cleaning ...... 130 5.3.2 Hierarchical linear regression and Tobit regression results ...... 131 5.3.3 Average marginal effects ...... 137 Chapter 6 – Findings and discussion ...... 146 6.1 Discussion of findings ...... 146 6.2 Theoretical Implications ...... 152 6.3 Managerial Implications ...... 157 6.4 Limitations and Future research ...... 162 6.5 Conclusion ...... 168 Appendix ...... 171 Appendix 1: Ethics approval ...... 171 Appendix 2: Survey questionnaire ...... 172 References ...... 183

Abstract

This PhD study represents a timely contribution to the study of absorptive capacity

(AC) and organizational learning by drawing on a routine-based conceptualization and examining the effects on innovation performance. It continues on the trajectory of research in this important line of enquiry about how AC dimensions are deployed to enable dual forms of learning. Despite the proliferation of the AC construct in the management and organization literature and acknowledgement of its role in organizational learning, our understanding of how the separate stages and dimensions of the construct are deployed for exploitation and exploration remains limited. Prior research has shown that exploitative and exploratory learning rely on contradictory organizational capabilities and routines; yet, AC is often conceptualized and empirically deployed as a uniform construct irrespective of the learning outcome pursued by the organization.

This study develops two competing theoretical arguments to address this conundrum that represent the exploitation-based view of AC and exploration-based view of AC to account for the dual role of the construct for driving performance benefits from incremental and radical innovation. Specifically, it argues that internal stages of AC are critical for driving incremental innovation that is the outcome of exploitative learning, while transformative stages of AC are critical for driving radical innovation that is the outcome of exploratory learning. To test hypotheses, the study draws from a unique dataset of 241 publicly listed firms. The findings provide support for assertions regarding the duality of AC and how its dimensions are deployed differentially according to the learning outcomes, which represents an important boundary condition to the construct that has yet to be theorized or explained.

The results have important theoretical implications for the literature on AC and organizational learning.

Chapter 1 – Introduction

1.1 Background and significance

The rapid diffusion and adoption of affordable advanced information and communication technologies, the lowering of barriers for the movement of people, capital, goods and services across national boundaries have increased the ease with which knowledge can be accessed globally (Cavusgil & Knight, 2015). New knowledge and technologies are rapidly transforming firms and collaborative networks, driving productivity and profits to record highs but also challenging existing business practices (Brynjolfsson & Hitt, 2000;

Ozalp & Kretschmer, 2018) As a result, firms are being challenged to gain benefit from knowledge sharing through network linkages, as the pace of innovation has become faster leading to the shortening of product and service life cycle (Stark, 2015). In such environment, where the intensity of change is constantly increasing, firms have to learn and innovate continually in order to sustain their performance. Therefore, innovation is widely regarded as a critical source of competitiveness in an increasingly changing environment (Dess & Picken,

2001; Kumar, Mudambi, & Gray, 2013). In addition, with tougher competition, technology advances, and shifting customer preferences, it's crucial for firms to facilitate learning through knowledge processes to adapt to the unpredictable environment. Higher-level cognitive efforts and a more deliberate and collective focus on the learning challenge (Zollo & Winter,

2002) can help to penetrate the ambiguity, thus potentially increase innovation performance.

Previous studies described innovation as a knowledge management process

(Madhavan & Grover, 1998) and characterized innovative firms as knowledge creating

(Nonaka & Takeuchi, 1995) where new products embody organizational knowledge. An evolutionary approach (Nelson & Winter, 1982) views innovation as a path-dependent process whereby knowledge is developed through interaction among various actors. 2

Therefore, at its core, innovation is about knowledge and emerges as a result of combining different knowledge sets (Nonaka & Toyama, 2003). It encompasses the process of creation and diffusion of new and economically valuable knowledge in the form of novel products, processes, and organizations (Feldman & Pentland, 2003; Feldman & Rafaeli, 2002) and relates to the discovery of knowledge not known by others (Cowan & Jonard, 2009). The creation and management of knowledge have become crucial factors for innovation and success of organizations (Kogut & Zander, 1996).

Innovation has usually been synonymous with large firms, dependent on in-house research and development (R&D) for improving their products or introducing new ones.

Therefore, scholars have traditionally focused on internal development of knowledge

(Brouwer, Kleinknecht, & Reijnen, 1993; Hoskisson & Hitt, 1988) for enhancing innovation performance. As a result, innovation performance has often been considered the outcome of

R&D activities and examined subject to taxonomies of novelty of innovation which spans from radical to incremental innovation (Soete & Freeman, 1997). In the closed innovation model, R&D is done in-house and serves as the key source for attaining competitive advantage by the means of commercializing technologies internally. Typically, these firms were characterised as having well-developed internal R&D capabilities (Nerkar & Paruchuri,

2005) as part of their strategic assets (Van de Vrande, De Jong, Vanhaverbeke, & De

Rochemont, 2009). Firms with established R&D capabilities have experienced superior performance (Teece, 1986) by being able to introduce new products and services which allow them to retain and even expand their customer base. Although R&D based innovation has been beneficial for firms for a long period of time, it has become increasingly less effective due to structural changes in the global economy (Dunning & Lundan, 2008) and the rise of open innovation model (Chesbrough, 2003).

The process of constant innovation can be severely hampered in the closed innovation model due to the limited internal R&D capabilities (Nerkar & Paruchuri, 2005). 3

As a consequence, firms have faced limitations in sourcing knowledge for introduction of new products and services (OECD, 2005) relevant for outperforming their competitors. In order to overcome their internal R&D capability constraints and invigorate innovation outcomes, successful firms have increasingly become more open to external sources of knowledge (Laursen & Salter, 2006). Such process has led to innovating with large variety of external sources, taking into consideration linkages to and knowledge exchanges with other firms (Subramaniam & Youndt, 2005). The benefits of external knowledge search for innovative performance have been proved in several empirical studies (Grimpe & Sofka,

2009; Katila & Ahuja, 2002; Laursen & Salter, 2006). In addition, evolutionary economics research (Nelson & Winter, 1982) has suggested the firms’ openness to external environment can improve their ability to innovate, enabling them to create new combinations of knowledge. Therefore, firms do not limit themselves to internal knowledge only, but also allow external knowledge to enter their innovation domain.

One of the main challenges that firms face in the process of enhancing innovation performance is the generation of new knowledge (Teng, 2007; Zahra, Filatotchev, & Wright,

2009). Such knowledge contributes to doing things differently, manifested as innovation in products, services, processes, systems, strategies and markets (Teng, 2007). New knowledge may either be generated by the firms in-house or acquired through various external channels.

The key mechanism for increasing the firms’ knowledge base (Mowery, Oxley, & Silverman,

1996), which explains their ability to deal with new external knowledge, has received the major attention in the literature and has been coined as absorptive capacity (AC) (Cohen &

Levinthal, 1990; Todorova & Durisin, 2007; Zahra & George, 2002). AC is defined as the firm’s ability to “identify, assimilate, and exploit knowledge from the environment” (Cohen and Levinthal, 1990, p. 589). Consequently, innovation performance is dependent on AC through the access to and generation of new knowledge from external sources (Laursen &

Salter, 2006; Love, Roper, & Vahter, 2014).

To survive pressure from the external environment, firms need to recognize new external knowledge, assimilate it, and apply it to commercial ends (Cohen & Levinthal, 1990).

Therefore, AC has emerged as an underlying theme in strategy and organization research.

This construct has been conceptualized as path-dependent process, driven by the systems, processes and routines of the organization (Todorova & Durisin, 2007) and consisting of four dimensions -acquire, assimilate, transform and exploit (Zahra and George 2002). In this regard, scholars distinguish between potential (acquire and assimilate) and realized AC

(transform and exploit) (Zahra & George, 2002). While potential AC refers to the ability to value and acquire external knowledge, realized AC reflects the capacity to leverage the knowledge that has been absorbed (Zahra & George, 2002). Knowledge leveraging is a subject to sense-making through deployment of organizational routines (Nelson & Winter,

1982) that allow firms to transfer acquired knowledge as long as a degree of commonality or knowledge overlap exists (Sun & Anderson, 2010). Finally, each AC dimension in turn is dependent on specific learning processes (Lane, Koka, & Pathak, 2006).

Organizational learning (OL) is the dynamic process of creating new knowledge and transferring it to where it is needed and used, resulting in the creation of new knowledge for later transfer and use (Argote, 2012). Therefore, OL supports expanding the knowledge stock of the organisation, thereby keeping its knowledge stock up-to-date and relevant (Kong & Farrell,

2012). It is commonly related to a change in the organization’s knowledge base that occurs as a function of its experience or by learning from the experiences of others (Argote, 2011,

2015). The success of firms dealing with and learning from such experiences can have a significant impact on whether they survive and prosper (Garud et al., 2011). In addition, OL is crucial for firms operating in unpredictable environments to respond to unforeseen circumstances more quickly than their competitors.

The learning process within the organization can be encouraged through careful design of organizational practices and routines, though it may develop into a dynamic

capability in its own right (Zollo & Winter, 2002). It is a process that enables a new shared understanding to evolve within the organization, particularly with regards to the new external knowledge that has been acquired. At the organizational level, learning occurs as routines are progressively refined to yield standard operating procedures to deal with categories of experiences that are expected to occur regularly over time (Garud, Dunbar, & Bartel, 2011;

Nelson & Winter, 1982). Such learning occurs in dual forms - exploitative learning, which refers to the elaboration and deepening of capabilities on an existing path or trajectory

(March, 1991) focused on incremental innovation; and exploratory learning, which refers to experimenting with or establishing new assets and new capabilities (March, 1991), thereby oriented towards radical innovation.

Given their shared focus, natural points of comparison and commonality exist between organizational learning and AC as relevant concepts that underpin innovation.

Previous research argued that AC is a concrete example of OL that concerns an organization’s relationship with new external knowledge (Sun & Anderson, 2010). On the other hand, OL enables more effective responses to dealing with external knowledge from the complex and dynamic environment. Therefore, a firm's openness towards external sources is seen as important mechanism for organizational learning (Lane et al., 2006). Furthermore, organizational learning underpins the firm’s ability to build new competences, reconfigure and realign the external knowledge and inter-firm competencies, essential for generating innovative outcomes (Neely, Filippini, Forza, Vinelli, & Hii, 2001). However, the role of capabilities and related learning processes that enable firms to explore and exploit knowledge

(March, 1991) as the sources of innovation performance remains unclear. Understanding the choices between exploratory and exploitative learning are complicated by the fact that returns from the two options vary not only with respect to their expected values and novelty, but also with respect to deploying the appropriate routines and processes rooted in different

dimensions of AC. Accordingly, the next section elaborates further the research problem and highlights the key research questions of this study.

1.2 Research problem and research questions

Firms increasingly innovate by tapping into external knowledge sources as the critical knowledge is not always available inside the firm (Nonaka, 1994). The more complex the innovation, and the broader the knowledge input that is required, the more likely that ﬁrms will seek external knowledge (Katila & Ahuja, 2002). In that regard, prior related knowledge is conducive to searching and absorbing knowledge from external sources (Lane & Lubatkin,

1998). The benefits of acquiring external knowledge for innovation have been widely demonstrated in previous research (Ahuja, 2000; Belderbos, Carree, & Lokshin, 2004;

Cassiman & Veugelers, 2006; Grimpe & Sofka, 2009; Katila & Ahuja, 2002; Laursen & Salter,

2006). Overall, these studies point to the fact that increased linkages to knowledge flows from various external partners, particularly in uncertain environments, lead to improved innovation performance, which is a subject to decreasing returns after a certain point (Laursen & Salter,

2006). However, though firms draw from external sources of knowledge, they face the challenge of learning and appropriating value from them. While the extant literature focused mostly on inter-firm relationships, it hasn’t sufficiently explained related knowledge and learning mechanisms which impact innovation performance.

Organizational learning helps improve managerial skills in undertaking innovative projects more effectively through time, and it may also develop into a firm-level dynamic capability in its own right (Zollo & Winter, 2002). However, capability development is a time consuming process which must take account of uncertainty and knowledge imperfections

(Pandza, Horsburgh, Gorton, & Polajnar, 2003), particularly in case of acquiring external knowledge. Overall, the dynamic capabilities research shows that experience provides firms with an opportunity for learning, but to make use of that opportunity managerial actions and 7

deliberate learning are needed (Zollo & Winter, 2002). However, the research on organizational learning highlights the trade-off between exploration and exploitation (learning in new domains versus learning in known domains) (March, 1991). Such trade-off is related to the ability of an organization to learn in mature markets where efficiency, control, and incremental improvement are needed in contrast to learning in new domains where flexibility, autonomy, and experimentation are valued (Tushman & O'Reilly, 1996). In addition, previous research views organizational learning as a routine-based process, where routines represent the rules, heuristics, and norms that trigger organizational responses to external stimuli as a means to innovate and change (Zollo & Winter, 2002). Hence, organizational learning depends on practices and routines, patterns of interaction both within and outside the firm, and the ability to mobilise tacit knowledge and promote interaction. Drawing from evolutionary economics, the need to manage routines of variation, selection, and retention

(VSR) has been positioned as essential for navigating dual organizational learning processes

(Nelson & Winter, 1982).

Several organizational researchers have defined learning in terms of acquiring, retaining, and transferring knowledge (Huber, 1991; Robey, Ross, & Boudreau, 2002). On the other hand, despite the growing interest in AC, few have captured the richness and multidimensionality of the concept. Even though Cohen and Levinthal’s seminal work (1990) highlights the multidimensionality of AC (Cohen & Levinthal, 1990), further research measured it as a unidimensional construct, usually using a firm’s R&D spending intensity as a proxy for it

(Flatten, Engelen, Zahra, & Brettel, 2011). Thus, previous studies do not capture the interactions between its pathways and knowledge conversion processes (Nonaka, 1991;

Nonaka & Takeuchi, 1995) that underpin AC. As a consequence, the construct is at the risk of reification (Lane et al., 2006) and the building blocks that constitute it largely remain a black box. One of the very few exceptions to the argument above is the study by Lewin and his colleagues (Lewin, Massini, & Peeters, 2011) that conceptually identifies the configuration

of meta-routines and routines underlying internal and external AC. However, the empirical investigation of such routine-based model is still scarce. While researchers have discussed AC as a process or capability, very few (Jansen, Van Den Bosch, & Volberda, 2005a) have attempted to operationalize and test the individual dimensions (Lane & Lubatkin, 1998). Due to lack of dedicated measures, the feasibility of AC construct remains constrained, which affects the rigour and legitimacy of the empirical studies in the field. Based on this, the study poses the first research question: What are the underpinning elements of the differing dimensions of AC construct at the firm level?

The question of how organizations learn and acquire new knowledge has been one of the most enduring themes in organization and strategic management research (Afuah &

Tucci, 2012; March, 1991; Wang, Van De Vrande, & Jansen, 2017). The extant literature highlights the difference between exploratory and exploitative learning as two distinct forms organizational learning (March, 1991). Exploitative learning is associated with incremental innovation focused on the diffusion, refinement, and extension of the firm’s existing knowledge base; whereas exploratory learning is associated with radical innovation focused on the variation, replacement, and renewal of existing knowledge with new knowledge (Baum, Li,

& Usher, 2000; March, 1991). These dual forms of learning represent a relatively passive, accumulative process of experiential learning from usual experiences, as well as a more deliberate, cognitively taxing process of exploratory learning from unusual experiences respectively (Garud et al., 2011; Zollo & Winter, 2002). Thus, it is clear that exploitative and exploratory learning rely on contradictory organizational capabilities and routines (Benner &

Tushman, 2003; Lewin et al., 2011). Despite the vast literature on AC, we know very little about its underpinning processes and routines that drive dual learning processes oriented towards radical and incremental innovation performance. As such, the question of how AC is conceptualized within the broader organizational learning and dynamic capability research areas, as well as a focus on its measurement is timely for further development of the field.

The concept of AC was originally presented as the means through which organizations structure and manage these contradictory processes of learning to drive innovative performance (Cohen & Levinthal, 1990). Yet, despite the dual nature of exploitative and exploratory learning, AC has largely been conceptualized and empirically deployed as a uniform construct irrespective of the outcomes pursued by the organization (Lewin et al.,

2011). The implicit assumption, therefore, is that AC dimensions of acquire, assimilate, transform, and exploit, and the interactions between them, have an equal and “uniform” effect in driving both outcomes of incremental and radical innovation performance. While arguments against this assumption seem relatively intuitive given the underlying differences between exploitative and exploratory learning, scholars have yet to theorize and empirically investigate the dual role of AC during exploitative and exploratory learning and the effects on innovation performance. This is particularly important since high level of innovation performance can only be retained for as long as the uniqueness of the firm capabilities is sustained in the market, thereby it is subject to the managerial alertness to recognizing and deploying promptly the next bundle of unique capabilities that contribute to sustainable rents generation (Bamiatzi, Bozos, Cavusgil, & Hult, 2016). Thus, the study poses the second research question: What impact do the differing dimensions of AC (acquire, assimilate, transform and exploit) have on driving incremental and radical innovation performance during exploitative and exploratory learning respectively?

1.3 Research objectives and contribution

Due to increased global competition, the ability of firms to respond to market changes, and reduce redundancy of information by using their knowledge and learning capabilities can be critical for sustained competitive advantage (Gold, Malhotra, & Segars,

2001). Ongoing advances in information and communication technologies (ICT) and the use

of digital platforms (Ozalp & Cennamo, 2017) are increasing pressure on firms to learn and innovate constantly in order to compete successfully. Organizational learning is largely dependent on a firm's contacts with external knowledge sources (Lane & Lubatkin, 1998).

Therefore, a firm's openness towards external sources is seen as important mechanism for organizational learning (Lane et al., 2006). There is further evidence in the literature related to the impact of AC on innovation (Kim, Akbar, Tzokas, & Al-Dajani, 2014; Kostopoulos,

Papalexandris, Papachroni, & Ioannou, 2011) as well as support for significant relationship between AC and innovative output (Wolfe, 1994). Since innovations result from recombination of familiar and unfamiliar elements of knowledge (Katila & Ahuja, 2002) the more diverse the set of knowledge sources that firms draws upon, the greater the chances are for the firm to combine knowledge in novel ways. This poses challenge on firms’ ability to deal with and apply the new external knowledge effectively to its products and services.

Therefore, analysing the effects of different types of AC on incremental and radical innovation performance during exploitative and exploratory learning may provide further understanding of how firms are able to take advantage of knowledge from external sources and learn from those interactions.

To date, corresponding measures for dimensions that form external, transformative and internal AC are still lacking. Therefore, the study differentiates between these types of AC and further clarifies the distinctness of their constituting dimensions and underpinning routines. Following the validation of such routine-based model (Lewin et al., 2011) the study examines the role of different types of AC in unlocking the benefits of exploration associated with radical innovation and exploitation associated with incremental innovation (Tushman &

O'Reilly, 1996). It responds to a call for the empirical examination of AC (Felin, Foss, &

Ployhart, 2015; Volberda, Foss, & Lyles, 2010), more specifically the dimensions of AC underpinned by routines (Lewin et al., 2011) as antecedents of organizational learning outcomes. The primary argument of the study is that the AC dimensions and their

underpinning routines work in different combinations to unlock performance benefits of exploitative and exploratory learning in the form of incremental and radical innovation respectively. Thus, the hypotheses predominantly focus on the interactions between individual dimensions during external AC, transformative AC, and internal AC. The study explicitly proffers competing hypotheses from these different components of the construct to account for the variations in its deployment for exploitative and exploratory learning. Such interactions are the key in revealing how firm-level components of AC can be uniquely blended to achieve high level of innovation performance and first mover advantages which tend to have an expiry propensity (Bamiatzi et al., 2016).

Testing these competing hypotheses was a formidable empirical challenge owing to the lack of validated routine-based measures for individual dimensions of acquire, assimilate, transform, and exploit that collectively constitute the AC construct, and difficulty in isolating observable routines for each dimension that aligns to their latent nature. The study overcame this challenge by drawing on the routine-based conceptualization of AC proposed by Lewin et al. (2011) to derive a list of observable routines for each individual dimension. These routines were then pre-tested with senior practicing innovation managers and leading academics before a confirmatory factor analysis was undertaken to operationalize measures for each dimension. For the purpose of empirically testing hypotheses, a unique dataset from surveying professional innovation managers from 241 firms was generated. The analysis provides support for assertions regarding the dual role of AC and extends prior research by theorizing and empirically validating the role of its individual dimensions during external, transformation, and internal stages in unlocking incremental and radical innovation performance. These results are robust with respect to a number of post-hoc checks, controls, and alternative specifications.

Findings of the study provide significant contributions to both theoretical lenses of AC and organizational learning. These are particularly relevant for empirical deployment of AC

dimensions, thereby validating a routine-based model (Lewin et al., 2011) and providing evidence for non-uniformity of the construct. Furthermore, the study contributes to organizational learning theory (March, 1991) by theorizing and confirming the role of internal

AC in driving exploitative learning with incremental benefits, and transformative AC as the key driver of exploratory learning, with positive effects on radical innovation performance.

Therefore, an AC lens adds value to explaining dynamics in organizational routines that form the foundation of the two distinct types of organizational learning (Levitt & March, 1988;

March, 1991). The study responds to the call for investigating “the complexities associated with coordinating interdependencies between and among internal and external AC routines”

(Lewin et al., 2011, p. 94) by showing dual nature of AC and providing a more nuanced perspective on organizational learning.

1.4 Research setting

The study focuses on innovation activities of firms of different sizes across industries in several OECD countries, that are members of global innovation association. This network consists of a number of industry members as important business players that produce and diffuse knowledge, crucial for innovation. With the advent of advanced information and communication technologies and rapid globalization, some of these firms are also considered

“Born-Globals” (BGs) and are inherently entrepreneurial and innovative (Coviello, 2015;

Knight & Cavusgil, 2004). These firms are seen as carriers or source of knowledge, and through their almost symbiotic relationship with clients, some of them are co-producers of innovation (Stauss, den Hertog, van der Aa, & de Jong, 2010). Therefore, such firms display a specific pattern of capabilities that enhance movement of knowledge beyond firm boundaries.

Most OECD members are high-income economies and are regarded as developed countries. Owing to these similarities, firms from other OECD countries provide a useful

basis for comparisons with Australian firms. Similar to other developed economies, Australia is shifting focus from lower skilled and highly labour-intensive industries towards knowledge- intensive and internationally-focused industries, especially in services and advanced manufacturing (Karanikolas, 2013). This vision has been supported by the National

Innovation and Science Agenda, launched by Australian Government in 2015, that provides a framework for Australia's new innovation policy, accompanied by $1.1 billion AUD investment over four years (DIIS, 2015). Due to the downfall of the mining industry, the

National Innovation and Science Agenda is designed to stimulate innovation and entrepreneurship aimed at creating new jobs and enhancing economic growth. However, at the macro level, Australia was ranked the 19th most innovative nation among OECD countries (dropped 2 places comparing to 2015) according to the Global Innovation Index

(GII) that aims to rank economies’ innovation capabilities and results (Cornell University,

2016). The evidence below (Table 1.1) suggests that Australia is less efficient than similarly developed countries in transforming innovation inputs into outputs. Australia’s relative GII weaknesses from innovation input mostly concerns human capital, while from innovation output side it exhibits weaknesses in knowledge and technology outputs.

Table 1.1 Australian scores in the GII 2016 (out of 128)

Overall performance Score (0-100) Rank

Global Innovation Index 2016 53.1 19

Innovation Output Sub-Index 41.3 27

Innovation Input Sub-Index 64.9 11

Innovation Efficiency Ratio 0.6 73

Global Innovation Index 2015 55.2 17

Source: Global Innovation Index 2016 (Cornell University, 2016)

At the micro level, Australian large firms were ranked poorly (21st out of 32 OECD countries) on the proportion of businesses innovating process in 2014 (OECD, 2014). In contrast to large firms, the same source reveals that Australian small to medium-sized enterprises (SMEs) are much more innovative, ranking 5th out of 29 OECD countries on the proportion of businesses innovating process. According to Australian Innovation System

Report (Department of Industry & Science, 2014) Australian firms possess a low ability to absorb and exploit external knowledge for competitiveness. The same source further indicates that absorptive capacity (AC) of firms may be limited by low concentrations of researchers in business and their uneven distribution within the private sector. Furthermore, OECD results suggest that Australian firms have among the lowest levels of collaboration for innovation within the OECD countries (OECD, 2014). Finally, the previous study found that Australian firms tend to downgrade innovation as a priority with a relatively poor level of innovation culture in Australian manufacturing industry (Samson & Gloet, 2014).

The above figures related to Australian firms show their overall poor performance in relevant areas such as poor innovation culture, low absorptive capacity and low level of collaboration for innovation. On the output side, low scores are particularly predominant in the areas of knowledge impact and knowledge diffusion. Overall, the GII indicates shortcomings in the capacity of Australian firms to generate and to bring innovations to application and diffusion. These facts suggest serious challenges that Australian firms face when engaging in innovation activities, particularly with external parties. Therefore, there is an urgent need for evidence to explain the phenomenon considering the presented figures, which serve as the motivator to undertake this research in addition to the research gap found in the academic literature. Findings should provide insights related to key capabilities that are paramount for innovation in response to the above-mentioned challenges, with potential differences across sizes, industries and countries in the research setting.

1.5 Research design and thesis outline

The research design is based on the main research questions, which are addressed within the empirical study based on cross-sectional design. The study applies quantitative research methods (Bryman, 2004) and collects data using the online questionnaire distributed to a sample of 715 firms. The structure of the thesis consists of the six chapters mapped below (Figure 1.1). Following chapter 1 related to introduction to the thesis, chapter 2 provides a comprehensive review of literature/prior studies in relevant areas. It is followed by chapter 3 where conceptual model and the research hypotheses are elaborated by 1) theorizing different types of AC with their underpinning elements and 2) conceptually establishing the link between the key constructs in the model. Then, Chapter 4 explains the methodological choices in the proposed research based on quantitative techniques as well as the philosophical assumption that underpin research. Furthermore, chapter 5 presents the work on validating measures of AC and further data analysis and modelling, studying the relationship among relevant variables by testing the model hypotheses. Finally, the main contributions including the theoretical and practical implications, along with the limitations and avenues for future research are elaborated in chapter 6.

Figure 1.1 Thesis structure (developed for the purpose of the study)

Chapter 1

Introduction

Chapter 2 Chapter 3

Literature Review Conceptual framework

Chapter 4 Chapter 5 Research Methodology Data analysis/modelling

Chapter 6

Discussion & Conclusions

Chapter 2 – Literature review

This chapter focuses on reviewing various academic concepts that have emerged in the literature to date and are relevant in the context of this study. Therefore, it encompasses organizational learning, knowledge management and dynamics capabilities theoretical lenses which have been significant research topics studying innovation. First, this chapter draws on innovation literature, in particular open innovation (Chesbrough, 2003), in order to examine its foundations and relevance of search for external knowledge in facilitating innovation outcomes and performance. Second, it includes the elements of knowledge management and knowledge-based view (KBV) of the firm (Davenport & Prusak, 1998; Nonaka & Takeuchi,

1995) streams of literature to explain the relevance of knowledge creation and the conversion processes. Third, by reviewing the dynamic capabilities theory (Helfat et al., 2007; Teece,

2007; Teece, 2016; Teece, Pisano, & Shuen, 1997; Zahra, Sapienza, & Davidsson, 2006) it explains the relevance of this framework and more specifically conceptualization of AC construct (Lane et al., 2006; Zahra & George, 2002) and its underlining routines. Then, this chapter provides review of organizational learning (Baum et al., 2000; March, 1991) concept relevant for innovation in dynamic and uncertain environments. Finally, it concludes by highlighting intersections and boundaries among various theoretical streams and outlining the research problem and gaps.

2.1 Innovation – multiple facets

Innovation as a research topic emerges across many disciplines, thus evolving as a relevant subject examined from different angles by a number of scholars, economists, scientists, engineers, etc. However, innovation seems to be defined and interpreted in different ways in the literature (Leonard & Swap, 1999; O’Sullivan & Dooley, 2009;

Rothaermel & York, 2013). Possibly, due to the long history, the term innovation is notoriously ambiguous and lacks either a single definition or measure (Adams, Bessant, &

Phelps, 2006). In the well-known “Theory of Economic Development” Schumpeter set the foundation for future innovation studies (Schumpeter, 1934). He argued that, for economic change to occur, it is important to consider innovation-originated market power and entrepreneurial activities rather than the invisible hand and price competition. Schumpeter defined innovation as a process of creative destruction, relating it to growth spurts - the driving forces of a capitalist economy (Schumpeter, 1950). A Schumpeterian perspective tends to emphasize innovation as market experiments and to look for large, sweeping changes that fundamentally restructure industries and markets.

The review of a numerous studies in innovation literature (Damanpour, 1991;

Henderson & Clark, 1990; Radas & Božić, 2009; Rothaermel & York, 2013) shows that most definitions stem from either a process perspective (proceeding from the conceptualization of a new idea to a solution of the problem and then to the actual utilization) or from an outcome perspective (those that see innovation as the final event). For example, according to process perspective “innovation refers to the process of bringing any new, problem-solving idea into use” (Kanter 1983, p. 20). In contrast, Damanpour (1991) presents three well-established categories of innovation: technical versus administrative, product versus process, and radical versus incremental, each of which focuses on innovation as an outcome. The difference between incremental and radical innovation is characterized by the variance in novelty between the innovation and the existing product or process that it improves (Henderson &

Clark, 1990).

While most definitions broadly fall in these two categories, further distinction is made between the traditional innovation and the knowledge-based innovation (Quintane,

Casselman, Reiche, & Nylund, 2011). Whereas traditional innovation approach (Damanpour,

1991) conceptualizes innovation without explicitly considering the underlying knowledge, the 19

knowledge management literature considers knowledge as the essence of the innovation process (Galunic & Rodan, 1998; Nonaka & Takeuchi, 1995). The most generally accepted definition of innovation comes from the OECD’s Oslo Manual (2005) which captures innovations that are new or significant to the firm, as well as new to the world, being broader than simply technical breakthroughs and their application in industry (OECD, 2005). In summary, what is common for all innovation definitions is that innovation is related to making changes to established norms by introducing something new that adds value to the organization. For the purpose of this study, outcome-based approach (Damanpour, 1991) is adopted where innovation is considered a knowledge-based outcome and further distinguished based on the degree of novelty – radical versus incremental. Finally, different views related to innovation definitions are outlined below, showing that the definition of innovation has extended throughout the literature from an early focus on products and exploitation of ideas to development of new methods and the establishment of new management systems. Such evolution has accounted for a value-added novelty, and knowledge as the essence of the innovation process (Nonaka & Takeuchi, 1995) which also affected the definition adopted for the purpose of this study.

- First commercial transaction involving the new product, process, system or device, although the innovation term is used to describe the whole process (Freeman, 1982). - Innovation refers to the process of bringing any new, problem-solving idea into use (Kanter, 1983). - A change that creates a new dimension of performance (Drucker, 2002). - An invention that has been exploited commercially (Martin, 1994). - The multi-stage process whereby organizations transform ideas into new/improved products, service or processes, in order to advance, compete and differentiate themselves successfully in their marketplace (Baregheh, Rowley, & Sambrook, 1996). - A sum of invention and exploitation (Roberts, 1988). - An invention that has been exploited commercially (Martin, 1994).

- The multi-stage process whereby organizations transform ideas into new/improved products, service or processes, in order to advance, compete and differentiate themselves successfully in their marketplace (Baregheh et al., 1996). - The process of making changes, large and small, radical and incremental, to products, processes, and services that results in the introduction of something new for the organization that adds value to customers and contributes to its knowledge store (O’Sullivan & Dooley, 2009). - The commercialization of any new product, process, or idea, or the modification and recombination of existing ones (Rothaermel & York, 2013). - Any novel product, service, or production process that departs significantly from prior product, service, or production process architectures (McKinley, Latham, & Braun, 2014). - The use of purposive inflows and outflows of knowledge to accelerate internal innovation, and expand the markets for external use of innovation, respectively (Chesbrough, 2006) - The embodiment, combination, and/or synthesis of knowledge in novel, relevant, valued new product, processes or services (Leonard & Swap, 1999). - The adoption of an idea or behaviour that is new to the organization, where the innovation can be a new product, a new service, a new technology or a new administrative practice (Hage, 1999). - The implementation of a new or significantly improved product (good or service), or process, a new marketing method, or a new organisational method in business practices, workplace organisation or external relation (OECD, 2005)

2.2 Innovation outcomes

The empirical studies to date has largely relied on readily calculated outcomes of innovation, with few studies examining the link between a more comprehensive organizational effects (Totterdell, Leach, Birdi, Clegg, & Wall, 2002). The most often reported significant explanation of innovation outcomes is R&D performance (Mairesse & Mohnen,

2010). This variable is significant and positive for innovation in almost all studies (Crépon,

Duguet, & Mairessec, 1998; Mohnen, Mairesse, & Dagenais, 2006). Two other variables that are mutually correlated with innovation are patents (Duguet & Lelarge, 2006) and exports

where the export performance of firms depends not only on their R&D efforts but also on their innovation activity (Sterlacchini, 1999). In addition, Fagerberg (1988) showed that the dynamics of the exports of the 15 OECD countries depends upon the technological innovation. Overall, progress in identifying outcomes of an innovation has been hindered by

(1) a predominant reliance on a few, positive outcome measures, (2) a concentration on inputs, and (3) a bias toward positive results (Simpson, Siguaw, & Enz, 2006).

Considering the lack of negative outcomes discussed in the literature, effects of an innovation are generally assumed positive and desirable; yet, a highly innovation-oriented organizational philosophy will have pitfalls. Oslo Manual (OECD, 2005) is an important guide in the field of innovation as it shows the dynamic process of a firm’s innovation outcomes and performance based on the evolution of innovation.. The first edition of Oslo

Manual has been limited to manufacturing firms, their product and process innovation, while the 2nd version included organizational innovation as one of service innovation categories.

Finally, the 3rd Oslo Manual (OECD, 2005) offers the broader classification of innovation and its outcomes, extending it to marketing innovation as well. Thus, product, process, organizational and marketing innovations form part of the 3rd Oslo Manual (Figure 1.2), where technological and service innovation are equally important (OECD, 2005).

Figure 1.2 Innovation connotation evolution

Source: Oslo Manual, Third Edition (OECD, 2005)

Oslo Manual (OECD, 2005) suggests that outcome indicators are directly affected by the length of product life cycles. They are likely to be higher in product groups with short life cycles and innovation is expected to take place more frequently. Thus, innovation surveys can be applied to ask firms to estimate the percentage of turnover affected by various innovations. An overview of indicators of innovation outcomes in the 3rd Oslo Manual

(OECD, 2005) is outlined below (Table 1.2).

Table 1.2 Indicators of innovation outcomes

Product The proportion of turnover due to new or significantly improved products. This Innovation indicator provides important information on the impact of product innovations Outcomes on the overall makeup of turnover (i.e. the share of turnover from new products) and on the degree of innovativeness of the enterprise.

Process The percentage of turnover that is affected by process innovations associated Innovation with improved efficiency, productivity and cost reduction. It can provide an Outcomes indication of how extensive process innovations are in terms of the firm’s total operations. Organizational Factors concerning workplace organization and results of organizational change. Innovation These could be related to customer relations, improving the capture and sharing Outcomes of knowledge and/or improving working environment.

Marketing The percentage of total turnover that is affected by marketing innovations. This Innovation indicator can be related to 1) an estimate of the percentage of turnover due to Outcomes goods and services with significant improvements in product design or packaging and to 2) an estimate of the share of turnover affected by new marketing methods in pricing, promotion or placement.

Source: Oslo Manual, Third Edition (OECD, 2005)

2.3 Innovation performance

Due to relevance of examining performance of firms based on different degrees of novelty of innovation (Soete & Freeman, 1997), the review within this section considers further differences between radical and incremental innovation and the associated 23

performance. The major difference captured by the labels radical and incremental is the degree of novel technological process content embodied in the innovation and hence, the degree of new knowledge embedded in the innovation. Increasing the degree of novelty of innovation can improve competitive advantages and create opportunities for firms to access new markets (McDermott & O'Connor, 2002). Innovation performance results from producing innovation outcomes (new products, processes, organizational and marketing innovations) (OECD, 2005) though examining innovation outcomes has changed along with innovation connotation. In addition, the level of novelty of innovation strongly influences innovative performance (Garcia & Calantone, 2002).

Incremental innovation relates to product line extensions or improving existing product/service as opposed to radical innovation, referring to products/services that are new to both the market and the firm (Radas & Božić, 2009). It is associated with a minor improvement or simple adjustments in current technology (Radas & Božić, 2009, Tushman

& Anderson, 1986) and returns from innovation on existing products and services

(Subramaniam & Youndt, 2005). Therefore, incremental innovation performance refers to financial benefits obtained from innovation that involves minor technology changes and relatively incremental customer benefits (Atuahene-Gima, 2005). The performance effects of incremental innovation are concerned with refining existing products, services, or technologies and reinforcing the potential of established product/service designs and technologies (Ettlie, Bridges, & O'keefe, 1984). The development of incremental innovation is a common sense in the same industry, and usually several different actors within the industry possess knowledge regarding specific aspects of the product/service or process. Firms in the same industry complement each other in co-creating and refining products and services but compete in dividing up markets (Brandenburger & Nalebuff, 1996).

In contrast to incremental innovation, radical innovation and its performance effects are related to major transformations of existing products, services, or technologies that often 24

make the prevailing product/service designs and technologies obsolete (Chandy & Tellis,

2000). Radical innovation represents revolutionary changes in technology and clear departures from existing practice (Radas & Božić, 2009, Tushman & Anderson, 1986). It is usually associated with returns from innovation that significantly change existing products/services

(Subramaniam & Youndt, 2005) and develop the new ones. Such innovation has potential to ensure biggest long-term performance, but it also requires the biggest departure from technologies, processes, and competencies employed before (Tushman & Anderson, 1986).

As Schumpeter stated, radical innovation offers “the carrot of spectacular reward or the stick of destitution” (Schumpeter 1942, p. 87). To achieve radical innovation, firms need to make a substantial investment in knowledge and R&D with potential benefit from great rewards, while for incremental innovation, the investment is considerably lower, but returns on investment are smaller as well (Tushman & Anderson, 1986). Therefore, radical innovation performance refers to financial benefits obtained from an innovation that incorporates substantially different technology and fulfils novel and emerging customer needs.

Firms require adequate learning mechanism to develop innovations, whether incremental or radical. The key learning mechanism associated with incremental innovation is known as exploitative learning (exploitation) and is focused on efficiency, increasing productivity, control and certainty (O’Reilly & Tushman, 2008). On the other hand, radical innovation requires exploratory learning (exploration) (March, 1991) with focus on search, discovery and autonomy (O’Reilly & Tushman, 2008) as well as engagement in variation, experimentation, and play (Gupta, Smith, & Shalley, 2006). Previous research suggested that irrespective of the underlying interpretations of organizational learning, the assumption that learning will improve future performance exists in all instances (Fiol & Lyles, 1985). However, this study argues that more variable approach is necessary to precisely establish such relationship subject to different learning processes (exploitative versus exploratory) and knowledge related mechanisms that might have differing effects on innovation performance

(radical versus incremental). Such approach to examining differing learning and knowledge mechanisms in relation to different types of innovation performance needs to reconsider traditional approach to innovation model.

Internal development of new knowledge may be accompanied by large resource expenses, high risk levels and issues with timing (Eisenhardt and Schoonhoven, 1996, Teng,

2007) which poses limitations to in-house innovation. Therefore, sourcing new knowledge from external sources is a complementary approach (Simsek et al., 2003, Teng, 2007, Zahra et al., 2009) to traditional innovation within a firm. The capacity of a firm to innovate is therefore enhanced by an extended knowledge base accessed through external networks of suppliers, customers, competitors, universities, etc. (Freel, 2003). Networks enable a firm to rapidly ﬁll in speciﬁc knowledge needs without having to spend enormous amounts of resources to develop that knowledge internally or acquire it through vertical integration (Van de Vrande et al., 2009).Through diverse external partnerships and network ties (Schilke and

Goerzen, 2010, Sarkar et al., 2009, Kale and Singh) a firm can obtain valuable and specialized knowledge, competencies and resources complementing or compensating their own limited in-house resources for innovation (Parida et al., 2010). Hence, there is a large body of literature on open innovation that discusses the value of accessing knowledge from external sources and networks, which is reviewed in the next section.

2.4 Open innovation

Due to the systematic encouragement and exploration of a wide range of internal and external sources for innovative opportunities, the concept of open innovation (Chesbrough,

2003) emerged. The term open innovation (OI) was first coined by Chesbrough as he stated that “firms can and should use external ideas as well as internal ideas, and internal and external paths to market, as the firms look to advance their technology” (Chesbrough, 2003,

p. xxiv). According to this concept, firms cannot afford to rely entirely on their own knowledge, but also use the opportunity to commercialize the knowledge from other firms. In addition, internal knowledge not being used in the firm can be exploited outside (i.e. through licensing) (Chesbrough, 2003). Therefore, OI is related to technology exploration (purposive inflows of knowledge) and technology exploitation (reflecting innovation practices to organize purposive outflows of knowledge) (Van de Vrande et al., 2009). To benefit from OI, firms need to identify external knowledge and be able to combine it with firm specific internal knowledge to generate a new product/service (West & Gallagher, 2006).

The initial definition of OI was refined to incorporate the “use of purposive inflows and outflows of knowledge to accelerate internal innovation, and expand the markets for external use of innovation, respectively” (Chesbrough, et al., 2006, p.1). Thus, OI paradigm represents a holistic innovation management strategy that explores and exploits a wide range of sources for innovation opportunities through multiple channels (West &

Gallagher, 2006). At the heart of OI is the assumption that firms cannot conduct all R&D activities by themselves, but instead have to capitalize on external knowledge which can be licensed or bought (Gassmann, 2006). In defining openness, Chesbrough argues that OI is a paradigm that assumes firms should use both external and internal ideas/paths to market, as they look to advance their technology (Chesbrough, 2003). Scholars have investigated how organisational factors such as internal R&D and external ones (e.g. market turbulence) are linked to OI (Dahlander & Gann, 2010) but little attention has been paid to the interplay between the different capacities required for managing knowledge in the OI context

(Brunswicker & Van de Vrande, 2014). One of the exceptions is the study on OI capabilities that identified six knowledge capacities for managing internal and external knowledge: inventive, absorptive, transformative, connective, innovative, and desorptive capacity

(Lichtenthaler & Lichtenthaler, 2009).

Following Chesbrough’s (2003, 2006) pioneering contributions, OI literature has covered many topics. They range from the direction of knowledge ﬂows (inward or outward), to the forms of openness (alliances, joint ventures, networks, etc.) (Blomqvist & Levy, 2006), the parties involved (suppliers, users, competitors, communities) (Foss, Lyngsie, & Zahra,

2013), or the impact of openness on innovation performance (Dahlander & Gann, 2010;

Laursen & Salter, 2006; West & Bogers, 2013). In addition, a number of aspects of OI in

SMEs at the international level have been examined (Wynarczyk, Piperopoulos, & McAdam,

2013), including the relevance of the OI model throughout the value chain and the role of social capital and external partners in enhancing the process. It was also suggested that age is a significant variable to explain the likelihood of SMEs to move from closed to open clusters

(Othman Idrissia, Amaraa, & Landrya, 2012).

The literature distinguishes between three kinds of open innovation: inbound, outbound and coupled (Chesbrough, 2003; Chesbrough, Vanhaverbeke, & West, 2006).

Inbound OI refers to the use of sources outside of the entity or group to generate, develop or implement ideas within a firm (Chesbrough, 2006). It is an outside-in process and involves opening up the innovation process to knowledge exploration (Lichtenthaler, 2011). On the other hand, outbound OI refers to use of external pathways for developing and commercializing innovations (Chesbrough et al., 2006). It is an inside-out process and includes opening up the innovation process to knowledge exploitation (Lichtenthaler, 2011).

Finally, coupled OI refers to combining both by the means of co-creation through alliances, cooperation and joint ventures (Enkel, Gassmann, & Chesbrough, 2009). Some of the examples are cross-licensing agreements where firms transfer a part of their own technology to get access to external knowledge (Grindley & Teece, 1997). Further research examined the concepts of technology exploration and exploitation respectively, linked to inbound and outbound OI and found that exploitation activities are pursued by less than 30% of the

sample, whereas exploration ones were adopted by more than 90% (Van de Vrande et al.,

2009).

Successful OI produces first-mover advantages, superior financial returns, market growth, and market share (Lichtenthaler, 2011). Hence, firms can improve their overall return by adopting OI for a more comprehensive innovation strategy. There are different methods that firms might consider when deciding to undertake OI. For instance, in an ideation contest, a firm (the seeker) facing an innovation-related problem (e.g., a technical R&D problem) posts this problem to a population of independent agents (the solvers) and then provides an award to the agent that generated the best solution (Terwiesch & Xu, 2008).

Another prominent method to gather ideas of users that can complement those of a firm's professionals at the idea generation stage is crowdsourcing (Poetz & Schreier, 2012).

Crowdsourcing refers to outsourcing “the phase of idea generation to a potentially large and unknown population in the form of an open call” (Poetz and Schreier, 2012, p.4). For example, as a part of tournament-based crowdsourcing (also referred to as “broadcast search”) a firm can solve technical problems in form of an open call for solutions to a large network of experts (Lüttgens, Pollok, Antons, & Piller, 2014).

Another stream of OI research addresses innovation markets (Chesbrough, 2007) and value creation in innovation ecosystems (Adner & Kapoor, 2010) with focus on facilitating

OI, especially regarding the ease of interfirm technology transfer. The scholars of this academic stream have examined the role of intermediaries in facilitating technology exchanges

(Chesbrough & Crowther, 2006). In this context, internet marketplaces for IP and technology auctions have received increasing attention as they promise new kinds of intermediaries for interfirm technology transfer (Chesbrough & Crowther, 2006). Furthermore, the evidence showed that the structure of technological interdependence affects firm performance in new technology generations (Adner & Kapoor, 2010). More recent research examined the role of

intergenerational platform‐technology transitions and related capabilities as sources of disruptive innovation at the ecosystem level (Ozalp & Kretschmer, 2018).

2.5 External knowledge search and innovation performance

Searching for external knowledge has been recognized as a crucial part of the innovation (Grimpe & Sofka, 2009; Katila & Ahuja, 2002) since the critical knowledge is not always available inside the firm (Nonaka, 1994). Increased linkages to and knowledge flows from various external partners, particularly in uncertain environments, lead to improved innovation outcomes (West & Bogers, 2011). Moreover, the benefits of collaboration with various actors in the innovation systems (i.e. customers, suppliers, universities) for innovative outcomes have been widely demonstrated (Ahuja, 2000; Belderbos et al., 2004; Cassiman &

Veugelers, 2006). Leiponen and Helfat (2010) found that an increased number of external knowledge sources leads to increased innovation and better financial performance (Leiponen

& Helfat, 2010). Love et al. (2014) point to similar findings by highlighting the “breadth of external innovation linkages” that leads to improved innovation outcomes. Firm that open their innovation processes are more likely to be innovative, though “the benefits to openness are subject to decreasing returns” (Laursen and Salter 2006, p. 132), thus additional search becomes unproductive after a certain point. Overall, there are negative consequences of too much external knowledge search both at the firm level (Laursen & Salter, 2006) and at the individual level (Dahlander, O'Mahony, & Gann, 2016).

Several studies in OI field (Belderbos, Cassiman, Faems, Leten, & Van Looy, 2014;

Cassiman & Veugelers, 2006; Laursen & Salter, 2006) emphasize the need for open search strategy which guides search for external knowledge and collaboration with external parties.

In that regard, external search breadth and depth (Laursen & Salter, 2006) have been commonly used as proxies for measurement of open search strategy (Dahlander & Gann,

2010). External search breadth refers to the number of external sources of knowledge or information in the innovation system, such as suppliers, customers and universities, used by the ﬁrm in its innovative activities (Laursen & Salter, 2006; Laursen & Salter, 2014). On the other hand, external search depth is defined in relation to the extent to which firms draw deeply from the different external sources or search channels of innovative ideas (Laursen &

Salter, 2006; Laursen & Salter, 2014). Subjects to novelty of innovation, Laursen and Salter

(2006) found empirical evidence for the argument that increase of radical innovation performance implies cooperation with only a few external sources but on a high intensity level

(depth). For incremental innovation performance, however, a broad variety of knowledge sources (breadth) was found to be more important as firms focus on fine-tuning the product by means of incremental improvements (Laursen & Salter, 2006). Further research concluded that cooperation and competition between a firm, its suppliers, customers, partners and competitors is “double-edged sword” to innovation performance (Bouncken & Kraus, 2013).

Previous studies have measured the beneﬁts of OI (Spithoven, Vanhaverbeke, &

Roijakkers, 2013), the limits to such beneﬁts (Laursen & Salter, 2006), the use of external sources and collaboration more generally in the innovation process (Grimpe & Sofka,

2009; Leiponen & Helfat, 2010; Love, Roper, & Bryson, 2011) and the net costs of OI

(Faems, De Visser, Andries, & Van Looy, 2010). Majority of these studies have sought to quantify the beneﬁts of value creation to innovation output and ﬁnancial performance, frequently using standard metrics for new product development studies. Examples include rate of new product releases (Boudreau, 2010), product performance (Lau, Tang, & Yam,

2010), revenue growth (Chesbrough & Crowther, 2006), the fraction of revenues attributable to radical innovations (Laursen & Salter, 2006) and the fraction of revenues attributable to new products (Grimpe & Sofka, 2009; Laursen & Salter, 2006). While earlier studies demonstrated positive outcomes of open search strategies (Laursen & Salter, 2006; Leiponen

& Helfat, 2010), more recent studies emphasized that firms need governance mechanisms

(Felin & Zenger, 2014; Foss, Husted, & Michailova, 2010) for orchestrating a system of relationships in order to enjoy the beneﬁts of openness (Gambardella & Panico, 2014).

Overall, such studies provide insights regarding contingency and management conditions for firms’ success with OI, particularly with regards to knowledge flows and their governance

(Foss et al., 2010). Open forms of innovation offer modiﬁed governance attributes that permit access to a broader array of knowledge, and often more powerful motivation for decentralized search, well suited to alternative types of problems (Felin & Zenger, 2014).

Previous research has shown that firms differ with regards to their potential to profit from OI (Lichtenthaler, 2011). Though OI paradigm and related external knowledge search offer prominent opportunities for enhancing a firm’s innovation performance, some scholars argue that the novelty outlined by Chesbrough and colleagues (Chesbrough, 2003;

Chesbrough et al., 2006) is elusive and not representing a dichotomy (Dahlander & Gann,

2010; Trott & Hartmann, 2009). They argue that OI paradigm created a partial perception with regards to the value of engaging external sources in innovation by describing something which is undoubtedly true in itself (the limitations of closed innovation principles) (Trott & Hartmann, 2009). In addition, the application of OI requires that firms are capable of understanding and utilizing the knowledge they acquire.

Lichtenthaler and Ernst (2006, p. 368) assert that “companies have to establish organizational processes and structures that facilitate knowledge transactions”. Thus, a successful OI builds on effective knowledge processes and firms need to provide a supportive context for the latter (Lichtenthaler, 2011; Lichtenthaler & Lichtenthaler, 2009).

In conclusion, based on the review of OI literature it is evident that most of studies pointed to the importance of external knowledge for innovation performance of firms.

Following this line of inquiry several empirical studies analysed the effect of external knowledge sourcing on innovation performance, and while some have showed positive effects on innovation performance (Leiponen & Helfat, 2010, Love et al. 2014) others point 32

to the decreasing effects after a certain point (Laursen & Salter, 2006). However, although recognizing that openness to external knowledge sources involves processes of identifying and selecting appropriate partners, developing routines for information processing and utilization of knowledge (Love et al., 2014), most of these studies lacked to empirically examine the role of such processes (Jeppesen & Lakhani, 2010; Piller & Walcher, 2006) and organizational capabilities that facilitate innovation performance. The reason for it could be potentially due to availability and convenience of databases such as the Community

Innovation Survey (CIS) (Grimpe & Sofka, 2009; Laursen & Salter, 2006) and other datasets that are generally based on an input-output, but have limited information on firm level capabilities and routines in such context. While recognizing the progress made by OI scholars, this study argues that further research is necessary to examine organizational competencies and learning mechanisms focused on blending external and internal knowledge for developing new products, processes and systems. Therefore, the next sections of the literature review look for the answers to these issues in the literature on dynamic capabilities, in particular AC and the organizational routines that underpin the construct.

2.6 Dynamic capabilities

The term “capabilities” comes from strategic management - the careful management of internal and external resources and competences to keep a business successful in the changing environment (Teece et al., 1997). Therefore, a capability is seen as a particular kind of resource that is embedded in the organization, and that improves the efficiency and effectiveness of other firm’s resources (Eisenhardt & Martin, 2000; Teece et al., 1997).

Scholars started using the term “dynamic” more frequently to explain the nature of capabilities in a changing and uncertain environment. This process requires innovativeness - capacity of an organization to produce innovations continuously (Galunic & Rodan, 1998) in

order to keep up with technological advancements as well as agility - fast response to changes in the uncertain environment (Teece, Peteraf, & Leih, 2016). Previously dominant in the literature, resource based view (RBV) suggests that resources are a source of competitive advantage (Barney, 1991; Barney, 1995). However, despite the importance of RBV (Barney,

1991; Barney, 1995) in explaining sustainable competitive advantages in static environments, this theoretical framework lacked appropriateness for vastly fluctuating market conditions.

Thus, dynamic capabilities (DC) (Helfat et al., 2007; Teece, 2007) emerged as a key source of competitiveness of firms, enabling them to change their unique resources over time and adapt to changing market circumstances, given the dynamic nature of their environment (Helfat et al., 2007; Teece et al., 1997).

The initial conceptualization of DC distinguishes the capacity to 1) sense and shape opportunities and threats, 2) to seize opportunities, and 3) to uphold competitiveness by the means of enhancing, combining and reconfiguring the firm’s intangible and tangible assets

(Teece et al., 1997). Therefore, a firm’s ability to integrate, build, and reconfigure internal and external competencies to address turbulent environments is primarily concerned with its capability to innovate (Teece et al., 1997). Other scholars (Kindström, Kowalkowski, &

Sandberg, 2013) provided a support for DC approach finding it useful for improving innovation, particularly in services, across the areas of sensing, seizing and transforming the business environment. Further distinction was made between dynamic and operational capabilities according to the capability hierarchy (Winter, 2003), where operational capabilities are concerned with the firm’s operational functioning and are considered substantive

(ordinary). The recent research demonstrates further the distinction between ordinary and dynamic capabilities (Teece, 2016) by outlining that ordinary capabilities allow the performance of the activities needed to meet current objectives. They involve the performance of administrative, operational, and governance-related functions necessary to the execution of current plans. In contrast, DC involve higher-level activities that can enable an

enterprise to upgrade its ordinary capabilities and to direct its ordinary activities toward high- payoff endeavours. This requires coordinating, or “orchestrating,” the ﬁrm's resources to address and even shape changes in the environment (Teece et al., 2016). While ordinary capabilities are related to efficiency, dynamic capabilities are associated with learning and improving and about being innovative and effective or, in another word, entrepreneurial

(Teece et al., 2016).

The capability literature shows that experience provides firms with learning prospects and that learning mechanisms such as knowledge-related activities are an important pillar of the development of DC (Zollo & Winter, 2002). In the context of DC concept, organizational learning may be treated as the way to integrate DC into the internal processes of the firm.

Eisenhardt & Martin (2000) suggest that DC become more salient through the process of learning that generates new knowledge Accordingly, there are three learning mechanisms, comprising experience accumulation, knowledge articulation, and knowledge codification, through which firms tend to develop DC (Zollo & Winter, 2002). Hence, DC emerge as the result of learning and have been coined as “second-order” competences (Zollo & Winter,

2002).

The DC framework has been evolving rapidly in the course of the last two decades.

One key implication of the DC concept is that firms are competing not only in terms of their ability to exploit existing capabilities, but also on their ability to renew and develop such capabilities (Teece et al., 1997). Furthermore, it was proposed that reside in managerial levers that enable innovation (Elkins & Keller, 2003; Mumford, Scott, Gaddis, & Strange, 2002).

Previous research indicated that an organization’s propensity to innovate or to adopt innovations is a type of dynamic capability which contributes to competitive advantage

(Helfat et al., 2007). For example, dynamic innovation capabilities of continually pre-empting competitors by introducing new products and technologies helped Intel and Rubbermaid have sustained their “evolutionary fitness” in the market for many years (Helfat et al., 2007).

In addition, it was recognized that DC refer both to the ability of a firm to recognize a potential technological shift and to its ability to adapt to change through innovation (Hill &

Rothaermel, 2003). It was further found that strong DC are necessary for fostering the organizational agility necessary to address deep uncertainty, such as that generated by innovation and the associated dynamic competition (Teece et al., 2016).

Empirical studies studying the development process of capabilities started to evolve in

1980s (Eisenhardt, 1989; Fredrickson, 1984) and has grown immensely since then. These studies, which offer support for the DC framework, frequently applied a comprehensive case study methodology (Danneels, 2011; Tripsas & Gavetti, 2000), while recent quantitative studies examined only a narrow aspect of dynamic capabilities. The review on dynamic capabilities (Wang & Ahmed, 2007) outlined that a significant number of empirical studies

(D'Este, 2002; George, 2005; Lehrer, 2000; Salvato, 2003; Woiceshyn & Daellenbach, 2005) dealing with this framework do not explain the concept itself, but rather describe how firm evolution occurs over time. In response to this issue, Wang and Ahmed (2007) suggested DC be viewed according to antecedents, the concept itself and the consequences. They proposed adaptive capability, absorptive capability and innovative capability as the most important components of DC that underpin a firm’s ability to integrate, reconfigure, renew and recreate its resources in line with external change (Teece et al., 1997; Wang & Ahmed, 2007). These three capabilities are correlated, but conceptually distinct (Wang & Ahmed, 2007).

The emerging research on DC provides evidence on their development and long-term consequences on performance and survival (Adner & Helfat, 2003; Helfat et al., 2007;

Macpherson, Jones, & Zhang, 2004; Teece, 2007). By leading the change rate of knowledge,

DC become the organizational capabilities that are positively related to long-term performance. For example, recent empirical study (Rakthin, Calantone, & Wang, 2016) combined the market orientation (Jaworski & Kohli, 1996; Kohli & Jaworski, 1990) and AC

(Lane et al., 2006; Zahra & George, 2002) constructs for explaining effects on performance.

Results showed that AC related to market knowledge positively influenced firm performance by enhancing customer acquisition and retention (Rakthin et al., 2016). On the other hand, according to Helfat et al. (2007, p. 7) performance effects of DC should be measured by means of the evolutionary fitness since “the extent of evolutionary fitness depends on how well the dynamic capabilities of an organization match the context in which the organization operates.” It was further proposed that a deeper understanding is needed regarding different mechanisms that underpin the performance effects of capabilities (Schilke, 2014).

One of the latest DC studies (Laaksonen & Peltoniemi, 2018) applies a critical viewpoint by indicating that dynamic capabilities cannot explain performance, but only changes in performance. This is due to the fact that performance outcomes depend on quality of the ordinary capabilities that the DC alter (Zahra et al., 2006) and the evolutionary nature of those capabilities (Helfat et al., 2007). It was further suggested that a possessing dynamic capabilities as such doesn’t necessarily result in a positive performance (Eisenhardt & Martin,

2000). In addition, there has been a criticism of tautology in a) recognizing firms that portray superior performance, and b) accrediting performance to the DC the firms possess (Priem &

Butler, 2001; Williamson, 1999). It was pointed out that that DC are usually only identified where there is a sustained competitive advantage, and therefore it is tautological to claim that they are properties which lead to the competitive advantage (Priem & Butler, 2001). In order to mitigate that risk, scholars have argued that dynamic capabilities should be rather observed by the changes they cause in a firm’s resource base (Eisenhardt & Martin, 2000; Teece, 2007;

Zahra et al., 2006).

In summary, these conflicting findings in the DC research area lead to the problem where the concept of DC and its link to performance is still subject to ambiguity being insufficiently underpinned by the empirical data (Easterby-Smith & Prieto, 2008). Although there is an extensive progress made in the area, there are still many opportunities to examine the linkages between dynamic capabilities and performance. Much of the research has been

on examining how dynamic capabilities might enable firms to adapt to changes in markets.

Such changes then may or may not result in a superior firm performance. Therefore, DC have much more complex performance effects than initially assumed, varying from insignificant in very stable and dynamic settings to positive in moderately dynamic environments (Schilke,

2014). The concept of DC (Teece & Pisano, 1994) emphasized that in a dynamic environment a firm’s competitive advantage depends on internal processes and routines, focused on renewing its stock of knowledge. Accordingly, performance differences across firms are accounted for by differences in the capacity of firms to accumulate, deploy, renew, and reconfigure resources in response to changes in the external environment (Nelson & Winter,

1982; Teece, 2007). DC or in other words second-order competences (Danneels, 2008) can be seen as composed of concrete and well-known knowledge management activities (Paarup

Nielsen, 2006). Therefore, dynamic capabilities and the associated knowledge management activities create flows to and from the stock of knowledge (Paarup Nielsen, 2006). Related capabilities for knowledge management allow a firm to access external knowledge, internalize it and combine it with an internal knowledge base to create new knowledge (Jantunen, 2005).

2.7 Capability development and organizational routines

Capability development is a generative process and capabilities are identified through retrospective sense making as knowledge of organisational processes and markets evolve

(Pandza et al., 2003). Hence, capabilities emerge via the integration of specialist knowledge across a number of individuals, and are associated with the development of organizational competences and routines (Teece & Pisano, 1994). Organizational routines are considered basic components of organizational behaviour and sources of organizational capabilities

(Nelson & Winter, 1982). Whilst initially characterized as a source of inertia and inflexibility

(Hannan & Freeman, 1984), routines were subsequently reconceptualised as relevant source

of flexibility and change (Feldman & Pentland, 2003). Placed at the core of evolutionary economics theory (Nelson and Winter 1982), routines relate to understanding firm adaptation, innovation, and change under constraints of bounded rationality. Hence, routines are stable patterns of behaviour that embed organizational reactions to variegated, internal or external stimuli. Zollo and Winter (2002) distinguish search routines and view the systematic generation and modification of operational routines as the vital mechanism underlying firms’ dynamic capabilities (Zollo & Winter, 2002). Accordingly, routines constitute the building blocks of organizational capabilities (Winter, 2003) and evolve over time as a result of problemistic search, organizational learning, and past selection and retention processes.

Routines constitute “organizational memory” due to relevance to building the firm’s stock of knowledge, and particularly tacit knowledge (Nelson & Winter, 1982). Furthermore, routines embed regular and consistently practiced patterns of individual and business behaviours that institutionalize individual or organizational knowledge about the firm's ongoing, rent-generating activities (Dosi et al., 1988; Nelson & Winter, 1982). In addition,

“routines reflect experiential wisdom in that they are the outcome of trial and error learning and the selection and retention of past behaviors” (Gavetti and Levinthal, 2000, p. 113). This view is consistent with the traditional view of organizational learning as skill building based on repeated execution of similar tasks that is implicit in much of the empirical literature on learning curves (Argote & Ingram, 2000). In conclusion, routines encode organizational capabilities, serve as a means for storing knowledge, and are also seen as key components of organizational learning (March, 1991).

In addition to organizational routines, metaroutines have been theorized as a mechanism for generating dynamic capabilities (Teece & Pisano, 1994). Metaroutines refer to routines for changing routines (Adler, Goldoftas, & Levine, 1999) and are often caused by external pressures (i.e. from the management team) to improve performance (Feldman &

Pentland, 2003). Thus, a number of firms employ meta-routines such as continuous 39

improvement and total quality management (Hackman & Wageman, 2005) as a means to generate change. In addition, Zollo and Winter (2002) proposed that DC were created through the continued interaction and mutual adjustment of learning mechanisms (i.e. experience accumulation, knowledge articulation and codification) and further guide the systematic creation and transformation of operating routines (Zollo & Winter, 2002).

A dynamic capability is a learned and stable pattern of collective activity through which the firms systematically generate and modify their operating routines in pursuit of improved effectiveness (Zollo & Winter, 2002). Therefore, DC are systematic patterns of organizational activity aimed at the generation and adaptation of operating routines (Zollo &

Winter, 2002). While DC are often characterized as unique and idiosyncratic processes that emerge from path-dependent histories of individual firms (Teece et al., 1997), such capabilities also exhibit common features associated with effective processes across firms

(Eisenhardt & Martin, 2000). Eisenhardt and Martin (2000) highlight the commonalities which imply that DC are “equifinal”, so firms can develop capabilities from many starting points and along different paths (Eisenhardt & Martin, 2000). The specifics of any dynamic capability may be idiosyncratic to a firm (i.e. exact composition of a cross-functional product development team) and path dependent in its emergence, but there is a common “best practice” for particular DC across firms (Eisenhardt & Martin, 2000). However, the existence of common features does not imply that any particular DC is exactly alike across firms.

Eisenhardt and Martin (2000) further point to the importance of repeated practice as a learning mechanism for the development of DC, which enhances development of more effective routines (Eisenhardt & Martin, 2000). The repeated practice per se can contribute to the evolution of DC, but the codification of that experience into technology and formal procedures accelerates the creation of routines (Zander & Kogut, 1995). Accordingly,

Eisenhardt and Martin (2000) suggested a modified concept of DC and an expanded view of routines. They conclude that in moderately dynamic markets, routines in the form of DC are

embedded in cumulative, existing knowledge, while in high-velocity markets, such routines rely extensively on new knowledge created for specific situations (Eisenhardt & Martin, 2000).

Based on the previous studies in this area, the capability building process is illustrated below

(Figure 1.11). The next section further reviews relevant knowledge management literature due to importance of different types of knowledge and related knowledge processes embedded in routines and DC, primarily relevant for conceptualizations and deployment of the AC construct.

Figure 1.11 Capability building process (developed for the purpose of the study)

Interaction and mutual Meta-routines adjustment of ‘learning (higher level mechanisms’ routines) Repeated practice (learning (e.g. codification into Capability mechanism) technology and formal Practiced routines procedures

2.8 Knowledge management and Knowledge based view (KBV) of the firm

Organizational knowledge is a major source of firms’ competitive advantage and is one of the most strategically important resources (Barney, 1991; Grant & Baden-Fuller, 2004;

Nonaka & Von Krogh, 2009). The role of organizational knowledge is particularly important in dynamically competitive environments due to the positive effects of intangible knowledge-related assets and capabilities on firm innovation. Its complexity and tacitness

(Gosain, 2007; Simonin, 1999) are the most studied underlying dimensions of the nature of knowledge. Knowledge complexity is deﬁned as “the number of interdependent routines, individuals, technologies and resources linked to a particular knowledge” (Simonin, 1999, p. 470).

Knowledge tacitness refers to “how easy or difﬁcult it is to codify and articulate the information

that needs to be transferred for speciﬁc knowledge” (Gosain, 2007, p. 259). Both dimensions have a strong inﬂuence on the ease of knowledge acquisition (Narteh, 2008).

Tacit knowledge (Gertler, 2003; Nonaka, 1994) was particularly found important as a source of innovation and competitiveness since it is valuable, rare and costly to imitate (Peteraf,

1993). It encompasses skills, ideas and experiences that people have in their minds and are, therefore, difficult to access because it is often not codified and may not necessarily be easily expressed. Tacit knowledge might be difficult to share with clients and partners (Chadee &

Raman, 2012), so the management of tacit knowledge poses a challenge for firms that engage in acquiring external knowledge sources. This is even more due to the risk of losing tacit knowledge as a consequence of the high employee turnover. Tacit knowledge management is essential for firms within knowledge-intensive industries, especially for those whose activities are widely spread across national boundaries (Anand & Singh, 2000). It was further demonstrated that tacit knowledge contributed to organizational innovation, while explicit knowledge provided a predictable environment and guidance to performing tasks (Venkitachalam & Busch, 2012).

Furthermore, a tacit knowledge index (TKI) was developed to measure a firm’s ability to create and sustain core competence in knowledge management (Harlow, 2008).

The knowledge-based view (KBV) of the firm (Davenport & Prusak, 1998; Nonaka &

Takeuchi, 1995; Wang, Wang, & Liang, 2014) emerged as an extension of the resource-based view (RBV) (Barney, 1991; Barney, Wright, & Ketchen, 2001) of the firm, which provided a theoretical underpinning for the organisation learning and DC research. Starting with Kogut and

Zander (1992), the knowledge based view (KBV) of the firm considers knowledge to be the most important resource of the firm and the main determinant of competitive advantage

(Kogut & Zander, 1992; Nonaka & Toyama, 2003). Therefore, there is a positive link between knowledge, which is valuable, rare, inimitable and non-substitutable (VRIN) (Barney,

1991) and increased performance, thus leading to competitive advantage. The KBV also suggests that firms may generate useful knowledge by combining and applying their existing 42

knowledge and expertise, in order to effectively create new knowledge, as the crucial source of competitive advantage (Grant, 1996). Knowledge characteristics, including codifiability, teachability, complexity, system dependence and product observability, altogether influence the absorption of external knowledge (Zander & Kogut, 1995).

Scholars of this research stream argued that well-developed 1) knowledge acquisition,

2) knowledge sharing and 3) knowledge management capacities are vital for firms that engage in external knowledge sourcing (Dyer & Singh, 1998; Kogut & Zander, 1992; Lichtenthaler &

Lichtenthaler, 2009). First, knowledge-acquisition capabilities consist of processes and mechanisms for collecting information and creating knowledge from internal and external sources (Jantunen, 2005). Even though the importance of knowledge acquisition in innovation is recognized, it is not so clear how effective knowledge-acquisition capabilities are reflected in improved performance (Jantunen, 2005). They may have a more indirect than direct role in promoting innovation (Darroch & McNaughton, 2003), or may be a necessary but insufficient condition for enhancing performance (Zahra & George, 2002). Second, knowledge sharing capacity forms a part of the dynamic cycle and is a key source of competitive advantage for firms (Peteraf, 1993). The more that is invested in building up a knowledge network, the less likely the abandonment of this precious resource is expected

(Bush & Tiwana, 2005). Finally, knowledge management capacity refers to transforming the knowledge capacities and is regarded as a “second-order” dynamic capability, directed at meta-processes that make a firm’s knowledge management add up to more than the sum of the knowledge processes (Zahra et al., 2006; Zollo & Winter, 2002).

Knowledge transfer may lead to some change in the recipient’s behaviour or the development of the new idea that leads to new behaviour (Davenport & Prusak, 1998).

Szulanski (1996) emphasized that “the movement of knowledge within the organization is a distinct experience, not a gradual process of dissemination” (Szulanski, 1996, p. 28). Hence, knowledge transfer is a process of dyadic exchange of knowledge between the source and

recipient units consisting of four stages: initiation, implementation, ramp-up and integration.

However, according to Szulanski (1996) there are a number of impediments related to the knowledge transfer, internal stickiness being the main one. The findings suggest that together with causal ambiguity and relationships between source and recipient units, the recipient’s lack of AC is the most important impediment to knowledge transfer within the firm (Szulanski, 1996). The role of AC of the receiving unit was also emphasized as the most significant determinant of knowledge transfer in a number of other studies (Gupta & Govindarajan, 2000; Lane & Lubatkin, 1998).

In addition, to knowledge management and KBV of the firm, the “knowledge governance approach” emerged as a distinctive approach that links various fields such as knowledge management, organization studies and strategy (Foss, 2007). Thus, knowledge governance (i.e. choosing organizational structures and mechanisms that influence the processes of using, sharing, integrating, and creating knowledge in preferred directions and towards preferred levels) (Foss et al., 2010) has recently become a distinct issue in the management literature. Foss (2007) argued that when organizational issues are discussed in relation to knowledge processes, “organization” predominantly means “informal organization”, that is, networks, culture, communities of practice and the like, rather than formal governance mechanisms” (Foss, 2007, p. 37). He further argued that formal organization may be invoked but is “seldom if ever integrated into the analysis” and in general, “there is a neglect of formal organization” (Foss, 2007, p. 37). According to this perspective (Foss, 2007), knowledge issues are seldom explicitly dealt with from the perspective of organizational design. In contrast, knowledge governance is explicitly identified as organizational design exercises aim at influencing knowledge processes in value-creating directions (Foss et al., 2010).

Overall, successful knowledge management, concerned with the knowledge creation and its transfer in firms, is a prerequisite for organizational learning and innovation. The KBV of the firm (Davenport & Prusak, 1998; Felin & Hesterly, 2007; Nonaka, 1994) provides

valuable insights into the interactive processes through which knowledge is created and exchanged both within and outside firms as an important element of innovation performance.

Knowledge and expertise are embedded in the firm and carried out through multiple mechanisms of knowledge conversion (Nonaka, 1994) that include routines, policies, systems, documents, etc. These conversion modes capture the idea that tacit and explicit knowledge are complementary and can expand over time in creative activities by individuals and groups

(Nonaka, 1994; Nonaka, Von Krogh, & Voelpel, 2006). Therefore, organizational knowledge creation theory has emerged not only to explain the nature of knowledge assets and strategies for managing them, but also to complement the KBV of the firm and the theory of DC by explaining the dynamic processes of organizational knowledge creation (Nonaka, 1991, 1994;

Nonaka & Takeuchi, 1995; Nonaka & Von Krogh, 2009; Nonaka et al., 2006). The next section outlines various approaches to studying the key capability that received major attention in the literature, coined as absorptive capacity (AC) (Cohen & Levinthal, 1990), as relevant for dealing with external knowledge and blending it with internal knowledge base.

2.9 Absorptive capacity (AC)

The knowledge-based view (KBV) of the firm strongly influenced studies related to

AC due to its importance to developing and increasing a firm’s knowledge base. This view considers AC the key construct for sourcing new external knowledge and blending it with a firm’s internal knowledge base (Escribano, Fosfuri, & Tribó, 2009; Jantunen, 2005). It is a path-dependent process in which new knowledge must be transferred across different levels of learning (individual, group and firm levels) (Sun & Anderson, 2010). Creation of new knowledge is subject to conversion, recombination, and exchange of knowledge from external sources, thereby the concept of AC plays a crucial role in this process. In addition to the KBV of the firm (Kogut & Zander, 1992), DC theory (Teece et al., 1997) also influenced the

development of AC and its role in explaining why some firms are better than others in capturing value (Vargo, Maglio, & Akaka, 2008) from external knowledge and collaboration with external parties. By building on a KBV of the firm (Kogut & Zander, 1992) and DC theory (Teece, 2007; Zollo & Winter, 2002), scholars examined AC as important moderating or mediating capability in different fields (i.e. joint ventures and alliances) (Koza & Lewin,

1998), as well as for balancing exploitation and exploration (Andriopoulos & Lewis, 2009).

Absorptive capacity emerged from earlier learning transfer theory (Ellis, 1965) that dealt with positive and negative sides of learning and experience, with an emphasis on both stimulus and response. Since the introduction of the AC construct, researchers have attempted to explain its different aspects and as a result, two important approaches emerged.

The traditional approach considered AC as a static resource in firms and used R&D investments, the number of patents and educated persons as measures of AC (Escribano et al., 2009; Huang & Rice, 2009). However, this approach was challenged by further research that viewed AC through the lens of capability-based (Flatten et al., 2011; Lane et al., 2006;

Todorova & Durisin, 2007). Overall, this approach indicates that AC as a static resource in firms rather than a capability does not reflect the complexity of AC dimensions and the content of knowledge. Due to criticism of the traditional approach that associates AC to

R&D, further studies extend the concept through employee skills and motivation (Minbaeva,

Pedersen, Björkman, Fey, & Park, 2003), prior knowledge (Lane, Salk, & Lyles, 2001), relevance of the knowledge, similarities between organizational structures, and shared research communities (Lane & Lubatkin, 1998) as well as through organizational routines

(Lewin et al., 2011). Owing to a number of distinct approaches to studying AC, the extant literature encompasses rich but conceptually different conceptual frameworks which are reviewed further and summarized below (Table 1.3).

Table 1.3 Prior studies - absorptive capacity (developed for the purpose of the study)

Construct/dimensions Authors Contribution

AC (recognition, assimilation, (Cohen & Levinthal, 1990) Introducing the concept in an application) organisational context

Relative AC and inter- AC (acquisition, dissemination, (Lane & Lubatkin, 1998) organizational learning commercialisation)

AC (acquisition, assimilation, AC as a four-dimension transformation, exploitation) (Zahra & George, 2002) dynamic capability Potential and Realized AC

AC (recognition, assimilation or (Todorova & Durisin, Reconceptualization of AC transformation, exploitation) 2007) (process with feedback loops)

Relationship between MNC AC (employees’ ability and (Minbaeva et al., 2003) subsidiary human resource motivation) management (HRM) practices, absorptive capacity, and knowledge transfer AC – the extent to which partners developed overlapping knowledge (Dyer & Singh, 1998) Relational view (Partner bases and interaction routines specific absorptive capacity)

AC (recognition and assimilation (Tsai, 2001) Network perspective of AC through transformation and exploitation) Extension of process based AC AC (exploratory, transformative (Lane et al., 2006) (organizational learning lens) and exploitative learning)

AC antecedents/micro- Antecedents of AC and foundations of potential and (Volberda et al., 2010) environmental conditions realized AC

AC (external exploration, Open innovation capability (Lichtenthaler, 2009) internal retention, and internal view of AC exploitation)

Process based configuration of AC (assimilate before and after (Patterson & Ambrosini, AC – interactive process/flow acquire) 2015) sequence

AC (internal and external) – (Lewin et al., 2011) Routine-based model of AC micro-foundations

1) Seminal papers (Cohen & Levinthal, 1989; Cohen & Levinthal, 1990)

In their seminal paper, Cohen and Levinthal define AC as the “ability to identify, assimilate, and exploit knowledge from the environment” (Cohen and Levinthal, 1989, p.589).

In their further review of the construct, they put the emphasis on the collective abilities of a firm that prior knowledge provides “to recognize the value of new information, assimilate it, and apply it to commercial ends” (Cohen and Levinthal 1990, p. 128). According to this view,

R&D is at the centre of firms’ innovative processes and is related to both learning and innovation. Thus, AC is a function of a firm’s prior related knowledge and R&D investments which are crucial to AC development. Furthermore, this approach to AC pointed to the relevance of external knowledge for the construct, and considered it cumulative, emphasizing the need for a firm to invest in its AC on a constant basis (Cohen & Levinthal, 1989; Cohen

& Levinthal, 1990).

In addition, Cohen and Levinthal (1990) suggested that AC and the associated learning processes are applicable at a firm level, though they are not simply the sum of employees’ AC in the firm. The main argument is that the reason why some firms are able to value, understand, and apply new knowledge more effectively than the others is because they have already invested in nurturing their AC (Cohen & Levinthal, 1990). Accordingly, if a firm aims to make effective use of the knowledge sources to increase its level of AC city, it needs to focus on the interface between the external environment, the firm as a whole, and its constituent units (Cohen & Levinthal, 1990). In summary, these seminal papers (Cohen &

Levinthal, 1989; Cohen & Levinthal, 1990) positioned AC as a relevant concept in the literature and laid the groundwork for further theoretical developments and empirical studies.

2) Inter-firm and intra-firm processes and model of relative AC (Lane & Lubatkin, 1998;

Mowery et al., 1996; Szulanski, 1996; Veugelers, 1997)

Further studies have fallen in line with Cohen and Levinthal’s (1990) view of AC as both an antecedent and an outcome of learning. These studies increasingly examined the stickiness of knowledge and impediments to transfer of best practices between units in the firm (intra-firm) (Szulanski, 1996) and between different firms (inter-firm) (Mowery et al.,

1996). Findings suggested that success in transferring best practices is dependent on knowledge-based factors such as the AC of the recipient (Szulanski, 1996). In an inter-firm context, Mowery et al. (1996) found that a firm’s capability to absorb technology from an alliance partner was in part dependent on its AC (Mowery et al., 1996). In addition, Veugelers

(1997) suggested that a firm’s AC enabled it to learn from external R&D co-operation. This learning is then invested and exploited through internal R&D efforts, which in turn drives the greater external R&D co-operation and further learning (Veugelers, 1997).

The model of relative AC (Lane & Lubatkin, 1998) was the first attempt to theoretically extended Cohen and Levinthal’s concept (1990) as it proposed three common elements in view of the technology transfer and the success of the inter-organizational relationship. Accordingly, a firm's ability to learn from another firm depends on the similarity of both firms' knowledge bases, organizational structures and compensation policies, as well as dominant logics (Lane & Lubatkin, 1998). The model of relative AC is seen as a learning dyad-level construct and is therefore associated with interorganizational learning (Lane &

Lubatkin, 1998). Lane et al. (2001) further refined the AC construct initially conceptualized by

Cohen and Levinthal (1990). Therefore, “the first two components, the ability to understand external knowledge and the ability to assimilate it, are interdependent yet distinct from the third component, the ability to apply the knowledge” (Lane et al., 2001, p. 1156).

3) Reconceptualization of AC – Zahra and George (2002)

The most important reconceptualization of AC since seminal papers (Cohen &

Levinthal, 1989; Cohen & Levinthal, 1990) is that of Zahra and George (2002). This view 49

emphasizes the systems, processes, routines and structure of the organization that allow firms to identify, assimilate, transform and exploit external knowledge (Zahra & George, 2002).

Accordingly, the initial AC phase “recognizing the value” was substituted with “acquire” dimension. Also, the influence of the regimes of appropriability was relocated, and the concepts of transformation, activation triggers, and social integration mechanisms (Zahra &

George, 2002) were added to the construct. This view criticized previous studies for using proxy measures (i.e. R&D intensity, number of scientists in R&D departments) that “have been rudimentary and do not fully reflect the richness of the construct” (Zahra & George

2002, p. 199).

In line with DC theory, Zahra & George (2002) argued that AC is a dynamic capability and they distinguished between a firm's potential and realized AC. Potential AC, which includes knowledge acquisition and assimilation, captures efforts expended in identifying and acquiring new external knowledge and in assimilating knowledge obtained from external sources (Zahra & George, 2002, p. 189). Realized AC, which includes knowledge transformation and exploitation, encompasses deriving new insights and con sequences from the combination of existing and newly acquired knowledge, and incorporating transformed knowledge into operations (Zahra & George, 2002, p. 190). The fluidity of the “from” – “to” nature of DC was addressed in the AC model (Figure 1.3) in terms of potential AC (acquisition and assimilation) and realized AC (transformation and exploitation) dimensions. In addition, potential and realized AC play separate, though complementary roles. For instance, a firm might have the ability to acquire and assimilate external knowledge, and renew its knowledge base, though it might not be able to exploit the knowledge by integrating it into the operations (Zahra & George, 2002).

Figure 1.3 Model of absorptive capacity

Source: Absorptive capacity: A review, reconceptualization, and extension (Zahra & George, 2002)

4) Reconceptualization of AC (process with feedback loops) (Todorova & Durisin, 2007)

Todorova and Durisin (2007) challenged AC model of Zahra and George (2002) and defined AC as a firm’s ability to recognise the value of external knowledge, acquire, assimilate or transform, and exploit external knowledge. These scholars argue that although Zahra and

George (2002) conceptualize AC as a dynamic capability that fosters organizational change and evolution, they do not use thinking in cycles, thus, failing to capture the dynamics and complexity of the phenomenon (Todorova & Durisin, 2007). Therefore, they propose that development of AC is a path-dependent process that should capture the dynamics of the construct through the addition of feedback loops. Such feedback loops are noted between the absorbed new external knowledge and prior organizational knowledge as the antecedent of

AC (Todorova & Durisin, 2007).

In addition, a reintroduction of the initial dimension - recognizing the value of external knowledge (Cohen & Levinthal, 1990) was suggested, in addition to an alternative understanding of “transform” dimension and impact of social integration mechanisms.

Todorova and Durisin (2007) observe knowledge assimilation and transformation as two alternative processes instead of sequential relationship between them as proposed by Zahra and George (2002). As a result, they suggest that future research in this field should focus on more complex mechanisms of relationships between assimilation and transformation, considering the multilevel aspects and high complexity of organizational learning (Todorova

& Durisin, 2007). Finally, their model (Figure 1.4) highlights moderating role of power relations in the exploitation of knowledge - the difference in power between various actors can influence the AC and related processes either within the organization or between consumers and other stakeholders (Todorova & Durisin, 2007).

Figure 1.4 Absorptive capacity reconceptualization

Source: Absorptive capacity: Valuing a reconceptualization (Todorova & Durisin, 2007)

5) Organizational mechanisms of AC in MNC context (Minbaeva et al., 2003)

While previous studies focused on the importance of AC for knowledge transfer,

Minbaeva et al. (2003) extend AC construct by examining the types of organizational mechanisms (mainly related to HR) that increase levels of AC. The results of their study in multinational corporations (MNC) context indicate that investments in employees' ability and motivation through the extensive use of HRM practices contribute to knowledge transfer of firms (Minbaeva et al., 2003). Thus, employees' ability and motivation constitute the firm's level AC, which is treated as endogenous variable. Their findings show that managers can improve AC of their firms by applying specific HRM practices focused on employees' ability

(training and performance appraisal) and employees' motivation (performance-based compensation and merit based promotion). Furthermore, Minbaeva et. al. (2003) related the prior knowledge base (employees' ability) and intensity of efforts made by the firm

(employees' motivation) to the concept of potential and realized AC (Figure 1.5). While potential AC constitutes high level of employees' ability, realized AC is associated to high level of employees' motivation (Minbaeva et al., 2003).

Figure 1.5 Conceptual model

Source: MNC Knowledge Transfer, Subsidiary Absorptive Capacity, and HRM (Minbaeva et al., 2003) 53

6) AC antecedents and integrative framework (Jansen et al., 2005a; Volberda et al., 2010)

Jansen et al. (2005a) provide evidence of the distinct effects of organizational antecedents on the components of AC. They show that coordination capabilities, such as cross-functional interfaces, participation in decision-making, and job rotation, enhance potential AC, while systems capabilities related to formalization and socialization strengthen realized AC at the business unit level (Jansen et al., 2005a). Finally, integrative AC framework

(Figure 1.6) was developed (Volberda et al., 2010) which considered the moderating effect of environmental conditions on the relationships between the micro-foundations of AC

(managerial/intra/inter-organisational antecedents and prior knowledge), its process dimension (acquisition, assimilation, transformation and exploitation) and tangible

(innovation, R&D and firm performance) and intangible (intra-organisation transfer of knowledge, knowledge search and inter-organisational learning).

Figure 1.6 An integrative framework of absorptive capacity

Source: Perspective-absorbing the concept of absorptive capacity: how to realize its potential in the organization field (Volberda et al., 2010) 54

7) Organizational learning lens on AC – Lane et al. (2006)

Lane et al. (2006) provided a reification of AC construct by abolishing regimes of appropriability from it and consolidating transformation and exploitation dimensions. In Lane et al.’s (2006) conceptual reaffirmation, prior knowledge is not highlighted as in previous conceptualizations, and is instead added to ‘recognize the value’ dimension. They also limit

AC to specific contexts or industries. More importantly, considering AC as a capability and a higher order resource seems to be more consistent with the resource-based view (RBV) suggesting that superior performance mainly originates from higher order resources which are difficult to obtain and imitate, and built over time (Makadok, 2001). In addition, other concepts such as characteristics of firm members’ mental models, structures and processes as well as environmental conditions, knowledge characteristics, learning relationships and firm’s strategy were added to AC model (Lane et al., 2006).

By considering previous conceptualizations of the construct and emphasizing its process perspective (Lane et al., 2006) a more detailed definition of the construct was suggested. Therefore, AC was defined as a firm’s ability to utilize externally held knowledge through three sequential processes: (1) recognizing and understanding potentially valuable new knowledge outside the firm through exploratory learning, (2) assimilating valuable new knowledge through transformative learning, and (3) using the assimilated knowledge to create new knowledge and commercial outputs through exploitative learning. These distinct learning types are illustrated in the model below (Figure 1.7) (Lane et al., 2006). Transformative learning subsumes the knowledge assimilation and transformation processes and is the process of effecting change in a frame of reference. To that end, shared meaning and shared focus are developed among individuals, teams, and organizations belonging to diverse settings. Therefore, transformative learning dictates how efficiently and effectively a firm can create meaning from its knowledge base.

Figure 1.7 The reification of absorptive capacity

Source: The reification of absorptive capacity: A critical review and rejuvenation of the construct (Lane et al., 2006)

8) Relational view and network perspective on AC (Dyer & Singh, 1998)

The relational view on AC emerged (Dyer & Singh, 1998) and further conceptualized partner specific AC referring to the firm ability to recognize and assimilate valuable knowledge from a particular alliance partner (Dyer & Singh, 1998). Partner specific AC is a function of a) the extent to which partners have developed overlapping knowledge bases and b) the extent to which partners have developed interaction routines that maximize the frequency and intensity of socio-technical interactions. This capacity requires implementing a set of inter-organizational processes that allow collaborating firms to thoroughly identify valuable knowledge and then transfer it across organizational boundaries. Previously, it was suggested that the ability of a receiver of knowledge to unpack and assimilate it is largely a function of whether or not the firm has overlapping knowledge bases with the source

(Szulanski, 1996). Therefore, this argument was further linked to partner specific AC as well.

In addition, partner specific AC is enhanced through process where individuals within the alliance partners get familiar with each other’s knowledge bases and develop understanding about where critical expertise resides (Dyer & Singh, 1998).

The knowledge transfer study further suggested that the AC of the receiving unit is the most significant determinant of internal knowledge transfer in MNCs (Gupta &

Govindarajan, 2000). In addition, Tsai (2001) developed a network perspective of AC and suggested that organizational units could produce more innovations leading to higher

performance if they occupy central network positions. This allows them to access new knowledge developed by other firms, though they remain primarily dependent on their own

AC and ability to successfully replicate the new knowledge (Tsai, 2001). Accordingly, a unit may be able to access new knowledge, but not enhance its innovation and performance if it does not have capacity to absorb such knowledge. The results highlight that a unit's innovative capability is significantly increased by its centrality in the intra-organizational network, which provides opportunities for shared learning, knowledge transfer, and information exchange (Tsai, 2001).

9) OI view on AC (Biedenbach & Müller, 2012; Lichtenthaler & Lichtenthaler, 2009)

Open innovation (OI) research (Lichtenthaler & Lichtenthaler, 2009) further extends the process-based view of AC (Lane et al., 2006), which focuses on the external exploration, internal retention, and internal exploitation of knowledge (Figure 1.8). The main difference between this AC model and earlier model of Zahra and George (2002) is the fact that

“transform” dimension does not occur after “assimilate”, but new knowledge is rather assimilated through combination with existing knowledge through transformative learning.

Accordingly, recent research pointed to the relevance of transformative capability as a complementary dimension of exploratory and exploitative learning processes (Biedenbach &

Müller, 2012; Lichtenthaler & Lichtenthaler, 2009; Rezaei-Zadeh & Darwish, 2016). Such learning processes are complementary in their nature and their effects on innovative output depend on one another.

The OI academic stream highlights that due to time lags in developing markets, complementary knowledge and technologies, firms might not be able to apply new assimilated knowledge for commercial purposes immediately. Therefore, firms should be able to retain such knowledge over time to reactivate it when appropriate. Beyond AC, this OI capability

framework emphasizes that all three processes (knowledge exploration, retention, and exploitation) should be organized inside and outside a firm’s boundaries. In addition, the concept of “desorptive capacity” (Lichtenthaler, Ernst, & Lichtenthaler, 2005; Lichtenthaler,

Lichtenthaler, & Ernst, 2004) was introduced. It refers to a firm’s capability to exploit knowledge externally (i.e. external intellectual property commercialization), which is complementary to internal knowledge exploitation in a firm’s own products. Desorptive capacity consists of the process stages of identifying external knowledge exploitation opportunities and subsequently transferring knowledge to the recipient (Lichtenthaler, 2009).

Figure 1.8 Overview of the framework

Source: A Capability-Based Framework for Open Innovation: Complementing Absorptive Capacity (Lichtenthaler & Lichtenthaler, 2009)

10) Process based configuration of AC (Patterson & Ambrosini, 2015)

The empirical study in the context of biopharmaceutical firms (Patterson &

Ambrosini, 2015) supports most of the aspects from the theoretically derived models by

Zahra and George (2002) and Todorova and Durisin (2007). Its findings point to the fact that

“assimilate”, “transform” and “exploit” dimensions of AC are interlinked. However, in contrast to previous studies, the interesting element of this model (Figure 1.9) is that assimilate” dimension was positioned before and after “acquire” and “search for” was added to the “recognize value” dimension (Patterson & Ambrosini, 2015). This research highlights that “assimilate after acquire” is an important concept in itself which requires to be actively

managed (Patterson & Ambrosini, 2015). This study also draws attention to the risk of not sharing information either between firms or within the project development teams themselves. Moreover, if in a co-development project with another firm the “assimilate” processes doesn’t work properly, then the “transform” and “exploit” phases might be greatly affected and cause impediments to effectively transform and exploit the new knowledge

(Patterson & Ambrosini, 2015). The “assimilate” dimension is closely related to social integration mechanisms, the alliance and collaboration and project/program management capabilities (Patterson & Ambrosini, 2015).

Figure 1.9 Newly configured absorptive capacity construct

Source: Configuring absorptive capacity as a key process for research intensive firms (Patterson & Ambrosini, 2015)

11) Conceptual routine-based model of AC (Lewin et al., 2011)

Firms differ in their ability to exploit the external knowledge sources since AC can be understood as a firm-specific dynamic capability which is built over time (path-dependency) based on organizational routines (Winter, 2003). Therefore, it is important to examine the routine structure of AC, and their characteristics (tacit, firm specific, and idiosyncratic) (Lewin et al., 2011). Developing routines that enhance resource recombination and knowledge complexity enables firms to recognize and assimilate more complex knowledge from external sources. However, in acquiring new external knowledge, firms tend to be constrained by their

own existing knowledge base and past routines. Thus, in pursuit of new knowledge sources it is advantageous for the firms to obtain unfamiliar knowledge in areas that are still somewhat closely related to its existing activities (Teece, 1986). Drawing from organisational routines found in evolutionary economics (Nelson & Winter, 1982) and DC theoretical stream (Teece et al., 1997), further research (Lewin et al., 2011) considered shifting from examining AC as a static mechanism to one that acknowledges the dynamics of AC routines. Accordingly, Lewin et al. (2011, p. 95) highlight internal meta-routines, that “strictly refer to processes that occur within the boundaries of the firm”, and external meta-routines, which “refer to processes and practices for acquiring and assimilating knowledge from the external environment”. Such metaroutines are related to variation, selection and replication processes and further includes two meta-routines at the interface between internal and external AC (management of adaptive tension and transferring knowledge back to the organization (Lewin et al., 2011). Finally, this

AC model (Figure 1.10) suggests that metaroutines are expressed in different or similar ways, and in different combinations, as actual practiced routines. Such observable routines embody codified and tacit knowledge, which evolve through different mechanisms and processes including problemistic search, trial and error and learning processes (Lewin et al., 2011). The next section explains further how capabilities develop through organizational routines.

Figure 1.10 Internal and External AC Metaroutines

Source: “Microfoundations of Internal and External Absorptive Capacity Routines” (Lewin et al., 2011) 60

2.10 Organizational knowledge creation theory

Firms create knowledge primarily through the dynamic process of conversion between explicit and tacit knowledge (Nonaka, 1994). Nonaka and Takeuchi (1995) advocated for a view where transitions between explicit and tacit knowledge are critical for transfer of knowledge. On the other hand, scholars argue that explicit and tacit are two distinct forms of knowledge and neither is a variation of the other (Cook & Brown, 1999). These two dimensions of explicit/tacit and individual/social knowledge have been combined by Spender

(1996), who created a matrix of four different elements of an organization’s intellectual capital: social explicit knowledge, social tacit knowledge, individual explicit knowledge and individual tacit knowledge (Spender, 1996). Similarly, it was proposed that organizational knowledge creation takes place in communities of informal and formal interaction as groups of individuals find ways to communicate tacit and explicit knowledge (Nonaka, 1994). Thus,

“organizational knowledge creation can be viewed as an upward spiral process, starting at the individual level moving up to the collective (group) level, and then to the organizational level, and sometimes reaching out to the inter-organizational level” (Nonaka, 1994, p. 20).

Knowledge conversion explains, theoretically and empirically, the interaction between tacit and explicit knowledge and how they interact along a continuum through four different modes in SECI framework (Nonaka, 1994). First, tacit-to-tacit conversion (socialization) takes place when tacit knowledge of individuals is shared with others through shared experiences.

Second, tacit-to-explicit conversion (externalization) occurs when an individual or a group articulates the foundations of individual tacit knowledge. Then, explicit-to-explicit conversion

(combination) takes place when an individual or a group combines discrete pieces of explicit knowledge into a new whole (Nonaka, 1994). Finally, explicit-to-tacit conversion

(internalization) takes place when new explicit knowledge is shared throughout the firm and other members begin to use it to broaden, extend, and reframe their own tacit knowledge

(Nonaka, 1994). In summary, knowledge conversion (Figure 1.12) requires socialization – creation of tacit knowledge through shared experience (Nonaka & Takeuchi, 1995), externalization – making tacit knowledge explicit, combination – consolidating multiple sources of explicit knowledge and internalization – deconstructing explicit knowledge into tacit knowledge. The next section provides an overview of these knowledge conversion processes in more detail.

Figure 1.12 Dynamic process of conversion between explicit and tacit knowledge

Source: A dynamic theory of organizational knowledge creation (Nonaka, 1994)

1) Socialization

Socialization process of knowledge creation refers to transfer of tacit knowledge through shared experience (Nonaka & Takeuchi, 1995). Experience provides individuals with the opportunity to create knowledge through trial•and-error learning (Rerup & Feldman,

2011), while interruptions in experience provide opportunities for knowledge transfer. Tacit knowledge is transferred through rich communication media such as observation rather than through more explicit media (Nadler, Thompson, & Boven, 2003). An organizational form of tacit knowledge is embedded in routines, organizational culture and cognitive schemes. The gaps in the current knowledge base may in turn facilitate firms to identify and acquire the external knowledge that is not available in-house. Therefore, socialization conversion (tacit to tacit knowledge) often occurs through shared experience/interactions with other actors

beyond firm boundaries. It also takes place in communities of informal and formal interaction as groups of individuals find ways to communicate tacit knowledge (Nonaka, 1994). Through common knowledge practices, personal relationships and established ways of interacting, community members may also develop a common sense of identity (Wenger, McDermott, &

Snyder, 2002). Apart from the access to external sources, firms can encourage knowledge sharing inside the firm through conversation, open communication, and other informal events that enhance interaction among employees, activate discussions and the transfer of tacit knowledge. The new knowledge becomes a foundation for a new spiral of knowledge conversion (Nonaka, Toyama, & Konno, 2000) when shared with others through the mode of socialization (Nonaka et al., 2000).

2) Externalization

Tacit knowledge can be converted to explicit knowledge through externalization mode of knowledge conversion (Nonaka, 1994), which is also known as the process of codification (Zander & Kogut, 1995). This process involves more formal interactions such as expert interviews or activities focused on documentation of lessons learned from a project, which require reflection (Zollo & Winter, 2002). Tacit knowledge may become routinized in organizational practice (embedded knowledge) and inculcated in the organization as basic assumptions, beliefs and norms (encultured knowledge) (Blackler, 1995). However, there are risks of making all knowledge explicit and eliminating the tacit personal elements, which can even be destructive to knowledge overall. In addition, tacit knowledge transfer meets several individual and organizational barriers, among them stickiness being the most important

(Szulanski, 1996; Szulanski & Jensen, 2004). The notion of internal stickiness implies the difficulty of transferring knowledge within the organization. In subsequent research

(Szulanski & Jensen, 2004), the role of the template was examined in overcoming stickiness and it was found that the template was a key in replicating routines. In addition, reﬂection

routines (Zollo & Winter, 2002) are intended to enable ﬁrms to formally update their capabilities at speciﬁed intervals or as an integral step in a process (Lewin et al., 2011).

Overall, they are crucial in establishing systems of information analysis that enable codifying tacit knowledge and converting it into explicit as a part of the externalization mode of knowledge conversion (Nonaka, 1994).

3) Combination

Once knowledge is shared, firms need combinative capability (Kogut & Zander, 1992) to synthesize knowledge, which is similar to the integration concept (Grant, 1996) and the conﬁguration concept (Henderson & Clark, 1990). Combination is the process of converting explicit knowledge into more complex and systematic sets of explicit knowledge (Nonaka &

Takeuchi, 1995). As knowledge is dynamic, constantly changing, and evolving, knowledge management systems must also be robust and flexible to take frequent updates from all parts of the firm. In relation to that, careful and deliberate classification or indexing that both capture what the knowledge-giver has to say and the key terms the knowledge-seeker will use

(McInerney, 2002). Combination gives rise to systematized explicit knowledge, e.g. explicitly stated technologies, product specifications and manuals. It contributes to a significant progress in the creation of intellectual capital occurs when bringing together knowledge from different sources and disciplines (Nahapiet & Ghoshal, 1998). In that way, explicit knowledge is combined, edited, and processed to form new, more complex and systematic sets of explicit knowledge (McInerney, 2002).

4) Internalization

Internalization is the process of absorbing explicit knowledge, thus creating new individually held tacit knowledge (Nonaka, 1994). The acquired knowledge of individuals has to be converted into a transferable form and distributed internally in order to be used in business 64

(Jantunen, 2005). There are two different ways that lead to the absorption of explicit and thus the creation of tacit knowledge: (1) personal encounters in which knowledge is generated from real world experiences such as day-to-day work, and 2) simulation and experimentation in which knowledge is generated through learning-by-doing or trial-and-error learning (Hoegl

& Schulze, 2005b). These processes may be carried out through several processes such as: learning when observing other organizations, intentional search and monitoring, etc. (Huber,

1991). Explicit knowledge such as product concepts or the manufacturing procedures have to be actualized through practice and action (Nonaka et al., 2000). By accessing the databases, reading the documents or manuals created through previous phases of knowledge conversion, employees can internalize the explicit knowledge and enrich their tacit knowledge base

(Nonaka et al., 2000).

Overall, knowledge creation theory explained important knowledge conversion processes that are closely related to organizational learning, together being of paramount importance for firms to operate in the knowledge economy (Davenport & Prusak, 1998).

While organizational learning can be approached from several different theoretical angles,

Nonaka and Takeuchi (1995) addressed highly important research gap, since most of organizational learning theories were lacking to consider that knowledge development constitutes learning. Nonaka and Takeuchi (1995) have outlined the limitations of organizational learning theories observing that even after a number of years in this research area, a comprehensive view of what constitutes organizational learning has not been established. While the literature on organizational learning is voluminous and the key issue of knowledge creation (Nonaka, 1994; Nonaka & Takeuchi, 1995) was finally addressed, the field is still short of empirical data, debate, and agreement on the intersections and boundaries between organizational learning and knowledge related theories. For the purpose of getting closer to the puzzle, the next section further reviews extensive organizational learning literature in more detail.

2.11 Organizational learning

Organizational and management scholars have always been interested in how organizational learning (OL) occurs. A common assumption in the strategic management literature is that organizations learn and adapt, which enhances their ability to survive, thereby enhancing the sustainable competitive advantage (Crossan & Berdrow, 2003). For attaining the most beneficial results, learning processes need to be aligned in a coherent way so that the culture, systems, structures and procedures support the strategic orientation of an organisation (Crossan, Lane, & White, 1999). Accordingly, OL theory is typically considered from the perspective of Nonaka’s (1991) knowledge creation theory (Hoegl & Schulze,

2005a). Furthermore, organizations learn by capturing the experience of other organizations through the transfer of encoded experience in the form of technologies, codes, procedures, or similar routines (Levitt & March, 1988). While the OL literature is quite broad, two dimensions appear with certain consistency. The first is related to the content of learning, while another important dimension refers to the extent of cognitive development (Rogoff,

1990) and explains the level at which this development takes place. Overall, there has been a largely cognitive perspective on OL that has aimed to understand how people think about complex problems, solve them, and in addition, avoid errors (Argyris & Schön, 1997).

Due to different units of analysis, learning context and learning content, OL as a firm level concept can be excessively wide and multi-faceted. The origin of the problem is the initial definition of OL as the growing insights and successful restructurings of organizational problems by individuals reflected in the structural elements and outcomes of the organization itself (Simon, 1969). In this initial definition learning was associated with development of insights on the one hand, and structural and other action outcomes on the other. Accordingly, scholars have referred to learning as either a) new insights or knowledge (Argyris & Schön,

1997; Hedberg, 1981), b) new structures (Chandler, 1962), c) new systems (Jelinek, 1979), d)

actions (Cyert & March, 1963) or a combination of the above mentioned. From such variety of perspectives, one of the common definitions (Fiol & Lyles, 1985) summarizes OL as the

‘development of insights, knowledge and associations between past actions, the effectiveness of those actions, and future actions’ (Fiol & Lyles, 1985, p. 811).

The organizational learning (OL) literature has comprised an extensive variety of units of analysis when defining what constitutes the construct, including individual learning (Argyris

& Schon, 1978), team learning (Huang & Li, 2012), and learning embedded into systems and processes of the organization (Bontis, Crossan, & Hulland, 2002). Such variety of units of analysis has contributed to the breadth of the concept. Hence, scholars point to the need to consider the flow of learning across all levels rather than focusing on a single unit of analysis

(Crossan, Lane, White, & Djurfeldt, 1995; Marsick & Watkins, 1996). Some agreement exists about the need to make distinctions between individual and organizational learning, though

OL cannot be simply studied as the sum of individual members’ learning. Organizations, unlike individuals, develop and maintain learning systems that not only influence their immediate members, but are then transmitted to others by way of organization histories and norms (Kilmann & Mitroff, 1976). Learning from groups must be transferred to the organization, which results in a change in the organizational schema through the institutionalization of new structures, systems, processes and routines. These are referred to as the memory bins of the organization (Walsh & Ungson, 1991). Thus, research provides a conceptual framework of OL as a multilevel phenomenon, suggesting that learning in organizations occurs at the individual, group and organizational level (Inkpen & Crossan,

1995).

Due to the above-mentioned differences, it is hard to pinpoint OL concept to a single and all-encompassing definition, which is the reason why the theory of OL is often treated as a black box. One approach to it can be tracked to the beginning of the scientific management movement through controlled tests and experiments (Kanigel, 1997) to generate knowledge 67

about the principles underlying the causes of a given phenomenon. A number of scholars have conceptualized the knowledge that emerges from such an approach as being cumulative but within the confines of a specific overall paradigm (Knorr-Cetina, 1981; Kuhn,

1970). This is because prior knowledge shapes the problems addressed, the instrumentation used, and the kinds of solutions found to particular categories of phenomena (Garud et al.,

2011). In this scientific approach, learning is based on generalizations from samples to populations, using objective measurements and statistical tests. However, this approach has its shortcomings due to the prevailing lack of practical guidance and insufficient focus on categories of experience.

Another approach to studying OL emphasizes experiential learning to generate knowledge about how to carry out an activity. At the individual level, experiential learning occurs as people increasingly refine their skills to deal with predetermined tasks or technologies through a process of learning by doing (Argote & Ingram, 2000; Dutton &

Thomas, 1984). At the group and organizational levels, such learning occurs as routines are progressively refined to yield standard operating procedures to deal with categories of experiences that are expected to occur regularly over time (Nelson & Winter, 1982). In addition, firms increase proficiency at adapting their processes to address new opportunities as they accumulate experience in entering new niches but face early hurdles in extending their experience base (Eggers, 2012). Routines are transformed at the same time as the organization learns which of them to pursue, and discrimination among alternative routines is affected by their transformations (Levinthal & March, 1981). Hence, routines are transmitted through socialization, education and imitation, and can change over time as a result of interpretations of history (Levitt & March, 1988). However, relatively little is known about the details by which organizational experience is accumulated into the structure of routines, but it is clearly a process that yields different kinds of routines in different situations and is only partly successful in imposing internal consistency on organizational memories.

One of the most comprehensive views of OL is that of Crossan et al. (1999). These authors synthesized previous work on OL and postulated a ‘4I model’ which considered the socio-psychological processes of learning at each level of learning, especially when learning is transferred across these levels. The four socio-psychological processes they considered are intuition, interpretation, integration and institutionalization. Organizational learning is embedded in the socio-psychological processes of learning, which enables new learning to traverse levels of analysis in a feed-forward flow (Crossan et al., 1999), thereby enhancing the exploration activity of the organization. The feedback flow is the continued exploitation of historical learning through established systems, processes, practices and routines of the organization. These feed-forward and feedback learning flows makes OL path dependent, and are further influenced by the systems, processes, and structures of the organization (Crossan et al., 1995). The 4I model offers a valuable synthesis of OL by looking at the transfer of learning from the individual to the group to the organizational levels and provides insight into some of the internal drivers that facilitate this learning transfer.

Further noteworthy contributions to the field of OL theory (Argyris & Schon, 1996) emphasize that OL is a product of organizational inquiry. Accordingly, whenever expected outcome differs from actual outcome, an individual or group will engage in inquiry to understand and, if necessary, solve this inconsistency. In the process of organizational inquiry, the individual will interact with other members of the organization and learning will take place as a direct product of this interaction which often goes well beyond defined organizational rules and procedures. Their approach to OL theory is based on the understanding of two

(often conflicting) modes of operation: 1) Espoused theory - refers to the formalized part of the organization, pointing to the various instructions regarding the way employees should conduct themselves in order to carry out their jobs and 2) Theory-in-use – employees rely on interaction and brainstorming, so employees solve problems and learn in the loose, flowing, and social way. The fact that there is a mismatch between these two approaches is potentially

problematic for firms in finding the proper balance to create an environment conducive to learning, while maintaining the structured approach to it.

Argrys and Schon’s (1996) theory on congruence and learning identify three levels of learning which may be present in the organization: 1) Single loop learning - consists of one feedback loop when strategy is modified in response to an unexpected result (error correction), 2) Double loop learning - learning that results in a change in theory-in-use, thereby the values, strategies, and assumptions that govern action are changed to create a more efficient environment and 3) Deuterolearning - learning about improving the learning system itself, which is composed of structural and behavioral components which determine how learning takes place, therefore deuterolearning is learning how to learn (Argyris & Schon,

1996). A dichotomy between single-loop and double-loop learning is in the fact that the first one involves the production of matches, or the detection and correction of mismatches, without change in the underlying governing policies or values, while double-loop learning requires re-examination and change of the governing values. Single-loop learning is usually related to the routine, immediate task, while double-loop learning is related to the nonroutine, the long-range outcome (Argyris & Schon, 1996). Therefore, effective learning should include all three, continuously improving the organization at all levels. However, while any firm will employ single loop learning, double loop and particularly deuterolearning are a far greater challenge outcome (Argyris & Schon, 1996).

Levitt and March (1996) expand further on the dynamics of OL theory. Their behavioural interpretation of OL builds on three classical observations drawn from behavioural studies of organizations. The first is that behaviour in an organization is based on routines (Cyert & March, 1963; Nelson & Winter, 1982). The second observation is that organizational actions are history-dependent, as routines are based on interpretations of the past more than anticipations of the future due to development of collective understandings of history. Thus, organizations are perceived as learning by encoding inferences from history 70

into routines that guide behaviour (Levitt & March, 1988). Learning from past experiences can also be supported by codifying the lessons learned and storing them in manuals, knowledge databases etc. (Alavi & Leidner, 2001). The interpretations of experience depend on the frames within which events are comprehended and developed through story lines

(Daft & Weick, 1984). Therefore, organizations adapt to experience incrementally in response to feedback about outcomes. The third observation is that organizations are oriented to targets as their behaviour depends on the relation between the outcomes they observe and the aspirations they have for those outcomes. Accordingly, OL is viewed as routine-based, history-dependent, and target-oriented.

The emphasis on routines distinguish the formulation of OL from approaches that deal primarily with individual learning within single organizations (Argyris & Schon, 1978) and place OL closer to behavioural lens. Such routine-based conception of learning assumes that the lessons based on experience are maintained and accumulated within routines as rules, procedures, technologies and beliefs are well-maintained through systems of socialization and control and are further retrieved through mechanisms of attention within a memory structure

(Levitt & March, 1988). Routines and beliefs change in response to direct organizational experience through two key mechanisms. The first one is trial-and-error experimentation, which highlights the likelihood that a routine will be used is increased when it is associated with success in meeting a target, decreased when it is associated with failure (Cyert & March,

1963). The second mechanism is related to organizational search as an organization draws from a pool of alternative routines, adopting better ones when they are discovered (Levitt &

March, 1988)

In line with the routine-based approach to OL, Huber (1991) elaborates further how organizations store a great deal of hard information on a routine basis, sometimes for operating reasons and sometimes to satisfy the reporting requirements of other units or organizations. In addition, he distinguishes four constructs related to organizational learning 71

(knowledge acquisition, information distribution, information interpretation, and organizational memory) (Huber, 1991). According to Huber (1991) knowledge acquisition is the process by which knowledge is obtained, referring to this type of organizational learning through acquisition as ‘grafting’. Grafting is a form of external learning or learning from others and relates to knowledge acquisition through access to new members (Huber, 1991).

Accordingly, organizations learn from others in order to incorporate external knowledge and past experience into their future strategy. By perceiving knowledge sharing as a process of learning through grafting, individuals have an impact on the process of acquisition and as a result, affect the outcome. Therefore, learning from experiences of grafting could enable better knowledge acquisition strategies, pointing to the strategic impact of knowledge transfer by knowledge sharing individuals (Kogut & Zander, 1996).

Within previously elaborated views on organizational learning, areas covered include how organizations learn from direct experience, from the experience of others, and how organizations develop conceptual frameworks for interpreting such experience (Huber, 1991;

Levitt & March, 1988). The initial focus of the organizational learning scholars was on studying scientific and experiential learning processes(Baum et al., 2000; Kanigel, 1997) that make it possible to accumulate and refine useful stocks of knowledge in relation to specific categories of experiences. These approaches construe learning as a progressive refinement of knowledge, based on generating improved responses to known categories of experiences (Garud et al., 2011). Overall, these views highlight benefits of learning approaches as manifested in improved responses to well-recognized situations. In particular, specific and known categories of experiences provide individuals and groups with an efficient means for identifying appropriate knowledge and historically effective responses (Bowker &

Star, 1999). However, such learning implicit in many organizations needs to be supplemented by an approach able to handle unusual experience that fall either outside or between known categories. The alternative to previous approaches - a narrative perspective (Garud et al.,

2011) has emerged in cases when organizations are faced with unusual experience that requires people to play active interpretive roles to determine the meaning of an unusual situation and to work out an appropriate response (Garud et al., 2011). Through such an approach, firms need to engage various actors through continual reflection with regards to the implications on unfolding possibilities, which enables them to continually learn from unusual experiences. In addition, to facilitate the experience accumulation processes, firms will have to incur costs due to the time and energy required for people to meet and discuss their respective experiences and beliefs (Ocasio, 1997).

In a classic paper on organizational learning, March (1991) distinguishes between exploration and exploitation as motives for organizational adaptation (March, 1991).

Exploitation refers to the elaboration and deepening of existing capabilities and to incremental improvement in efficiencies on an existing path or trajectory (March, 1991). It is associated with mechanistic and tightly coupled organizational structures and the utilization of existing routines and experimental refinement, and the returns from exploitation are rather predictable, close, and positive (Baum et al., 2000; March, 1991). On the other hand, exploration refers to experimenting with or establishing new assets and new capabilities

(March, 1991). While, the strategic intent of exploitation is to obtain residual revenue and incremental enhancement of other competencies, the strategic intent of exploration is the discovery of new opportunities, which have the potential to dramatically affect a firm’s performance (Gupta et al., 2006; Sidhu, Commandeur, & Volberda, 2007). While firms ‘make- and-buy’ knowledge and identify new technologies in the process of exploration they tend to

‘keep-and-sell’ knowledge and focus on profitable applications and markets for a firm’s own technologies on the path of exploitation (Cassiman & Veugelers, 2006). Usually, exploration activities pave the way for firm’s opportunities to exploit and vice versa, exploitation sets the foundation for exploration (Rothaermel & Deeds, 2004).

In organizational learning literature, the problem of balancing exploration and exploitation is exhibited in distinctions made between refinement of an existing technology and invention of a new one (Levinthal & March, 1981; March, 1991). The knowledge cycle proceeds, in March’s (1991) terms, from an exploration phase to an exploitation one, potentially feeding back into a new exploration phase. In evolutionary models of organizational forms and technologies, discussions of the choice between exploration and exploitation are framed in terms of processes of variation, selection and replication (VSR)

(Winter & Nelson, 1982). In such a framework, three questions have become important to address: (1) how variation comes about, (2) how selection takes place, and (3) how what has been selected in one period is transmitted to the next period (Winter & Nelson, 1982).

Linking these questions to an organizational learning lens, Zollo and Winter (2002) point to the fact that exploration activities are primarily carried out through cognitive efforts aimed at generating the necessary range of new intuitions and ideas (referred to as variation) as well as selecting the most appropriate ones through evaluation and legitimization processes. On the other hand, exploitation activities rely more on behavioural mechanisms encompassing the replication of the new approaches in diverse contexts and their absorption into the existing sets of routines for the execution of that particular task (Zollo & Winter, 2002) Therefore, firms are faced with the challenging choice between exploitation and exploration or combining them for improving the innovation performance. Ultimately, it is a strategic choice between the application of what is already understood and mastered to newly generated knowledge and solutions and, vice versa, the enrichment of the things already mastered by new knowledge, approaches, and ideas (Katila & Ahuja, 2002).

In order to balance exploration and exploitation (Tushman & O'Reilly, 1996) it is important that firms have the ability to simultaneously pursue both incremental and discontinuous innovation and change (Tushman & O'Reilly, 1996). Prior research has advocated for organizational-level responses to these tensions in the form of structural

separation of exploratory from exploitative activities (Tushman & O'Reilly, 1996) or the creation of internal governance mechanisms that support the simultaneous pursuit of both

(Gibson & Birkinshaw, 2004). The extant literature further emphasizes contextual ambidexterity as the organizational behavioural capacity to simultaneously demonstrate exploration and exploitation across an entire system (Ghoshal & Bartlett, 1994; Gibson &

Birkinshaw, 2004). The contextual approach conceives of ambidexterity as emerging through a business unit’s organizational context, involving the combination of performance management with stretch targets combined with supportive values and processes to help individuals reach these targets (Gibson & Birkinshaw, 2004).

Ambidextrous organizations theory (Raisch, Birkinshaw, Probst, & Tushman, 2009;

Tushman & O'Reilly, 1996) provides a proven model for forward-looking firms in simultaneous pursuit of radical and incremental innovation goals. There has been further evidence supporting this claim not only through structural separation, but also through leadership and contextual solutions (Birkinshaw, Zimmermann, & Raisch, 2016; Raisch et al.,

2009). Other research has investigated the attempts of firms to balance exploration and exploitation through the alliances they maintain. Lavie and Rosenkopf (2006) establish contingencies of firms aiming at balancing exploration and exploitation through different alliances, and Holmqvist (2004) describes how firms focus either on internal exploitation and coalesce with external parties for exploration purposes, or vice versa (Holmqvist, 2004; Lavie

& Rosenkopf, 2006). While some researchers emphasize that multiunit firms need to separate exploration from exploitation in organizational units (Benner & Tushman, 2003), others argue that organizational units could become ambidextrous (Gibson & Birkinshaw, 2004).

Although, the interplay between exploration and exploitation is recommended, it is worth noting that under well-specified conditions, specialization rather than duality could be entirely feasible (Benner & Tushman, 2003).

The extant literature further suggested that the combination of exploration and exploitation is associated with longer survival (Cottrell & Nault, 2004), better financial performance (Markides & Charitou, 2004) and improved learning and innovation (Holmqvist,

2004; Rothaermel & Deeds, 2004). Other related studies, while not addressing ambidexterity directly, examine how organizations pursue exploration and/or exploitation, especially in relation to the firms’ ability to innovate (Holmqvist, 2004). In summary, these studies and related research on AC provide the lens for looking at the relationship between dynamic capabilities and ambidexterity. These empirical findings explain the conditions under which dynamic capabilities are most valuable and reinforce the importance of ambidexterity as a dynamic capability (O’Reilly & Tushman, 2008). Ambidexterity as a dynamic capability is not itself a source of competitive advantage but facilitates new resource configurations that can offer a competitive advantage (O’Reilly & Tushman, 2008).

In conclusion, examination of the voluminous OL literatures indicates that, while much has been learned about OL, there is a lack of synthesis of work from different research groups. Similarly, it was found that much has been learned about organizational search, but that there is insufficient empirical work and integration with which to create a more mature literature. It has become common in the literature to emphasize learning in the design of organizations, to argue that some important improvements in organizations can be achieved by deploying capabilities to learn quickly and precisely (Kong & Farrell, 2012; Starbuck, 1992;

Zollo & Winter, 2002). Organisations that create knowledge on an ongoing basis are likely to develop dynamic and unique capabilities that potentially underpin continuous organisational learning. Such capabilities include knowledge and learning capabilities, which can occur at individual, group and organisational levels, helping firms to adapt, learn and optimize resources (Teece et al., 2016). Moreover, capability building is enabled through exploratory and exploitative learning processes, thus it is clear that organizational learning complements dynamic capabilities. Due to affinity of the dynamic capability literature with organizational

learning and knowledge management as well, the next section deals with the intersection and boundaries amongst these theoretical streams in more detail.

2.12 Intersection – dynamic capabilities, knowledge management and organizational learning

The turbulence of the business environment has ensured focused attention on organizational knowledge as dominant sources of innovation. Scholarly interest in the topic largely stems from the complexity that characterizes balancing paradoxical tensions across various organizational interfaces, which involves the management of dual knowledge processes. Firms are increasingly looking for knowledge outside their organisational boundaries and are developing more outward-looking strategic approaches to R&D to source at least some knowledge of potential value from the broader environment in which they operate. The more innovation projects build on truly new knowledge, the more difficult it is for a firm to really understand the new knowledge requirements and hence, the less likely is innovation success (Katila & Ahuja, 2002). The three key concepts 1) dynamic capabilities, 2) knowledge management and 3) organizational learning reviewed in this chapter are used in research for the purpose of managing innovation in dynamic and uncertain environments.

The literature on these three streams of literature is characterized by the use of very diverse terminology, where concepts often overlap, but it is unclear how all the pieces fit together

(Vera, Crossan, & Apaydin, 2011). A significant progress in identifying an integrative model and establishing conceptual boundaries with relationships between the constructs has been made in order to avoid the risk of continuously “reinventing the wheel” (Easterby‐Smith &

Prieto, 2008; Vera et al., 2011).

Knowledge management (KM) emerged as a field that studies management of knowledge about a firm, its operations, competitors, customers and supply chain (Siemieniuch &

Sinclair, 2004). It is considered a vital strategic tool for guiding knowledge processes towards

accomplishing sustainable competitive advantage (Davenport & Prusak, 1998). However, knowledge by itself does not necessarily contribute positively to innovation performance and it can be beneficial only when firms have well-developed capabilities for the management of such knowledge (Kogut & Zander, 1992). The KBV suggests that the foundation of a firm’s innovation consists of development, deployment and systematic orchestration of intellectual assets and capabilities (Martín-de-Castro, Delgado-Verde, López-Sáez, & Navas-López, 2011).

Such capabilities are related to process of transforming knowledge into new products, services and systems (Lawson & Samson, 2001) which takes place under great uncertainty, thus making the innovation performance highly unpredictable. Hence, a more dynamic approach has emerged as response to uncertainty and external change, highlighting dynamic capabilities as a key source of competitiveness of firms (Teece et al., 1997)

The concept of DC has become the guiding framework for explaining the organizational knowledge and innovation. This framework focuses on explaining why certain firms have more success than others in achieving superior performance and building competitive advantage within dynamic markets (Teece et al., 1997; Zollo & Winter, 2002).

While DC highlight the renewal of resources by reconfiguring them into new capabilities

(Teece et al., 1997), KM focuses on providing answers on how to create, retain, transfer and use explicit and tacit knowledge (Davenport & Prusak, 1998). There is a link between these two concepts (KM and DC), which is not fully covered in the literature and empirically examined (Easterby-Smith & Prieto, 2008). On the other hand, prior research established that innovation is a positive outcome of effective DC and OL relation, contributing to sustainable competitive advantage (Lawson & Samson, 2001, Easterby‐Smith & Prieto, 2008). The purpose of innovation is to enable a firm to produce new value and gain competitive advantage, while DC enable a firm to purposefully adapt its resource base (Teece et al., 1997) to capture opportunities from environment. Therefore, DC are expected to increase firm’s ability to innovate, and are seen as an underlying factor that influence innovation success.

The intersection between DC and KM highlights the relevance of organizational learning (OL), which is related to the concept of knowledge management (KM) being concerned with the organization’s ability to create and transfer knowledge. However, KM tends to emphasize the static stocks of knowledge held by an organization and the characteristics of that knowledge, rather than the dynamic processes through which knowledge is developed by organizations (Vera et al., 2011) which is characteristic of the DC framework. Cepeda and Vera (2007) delineated the connections between KM and DC stating that: (1) capabilities are organizational processes and routines rooted in knowledge; (2) the input of DC is an initial configuration of resources and operational routines; (3) DC involve a transformation process of the firm’s knowledge resources and routines; and (4) the output of

DC is a new configuration of resources and operational routines (Cepeda & Vera, 2007).

As DC involve change, they involve learning - change in cognition and change in behaviour, and as DC build up routines and resources, they involve knowledge as the most valuable firm resource. However, while DC are focused on the modification of operational routines, OL mechanisms enable the creation and modification of DC (Easterby-Smith &

Prieto, 2008; Winter, 2003). Therefore, DC arise from learning as they constitute the firm’s systematic methods for modifying operating routines. To the extent that the learning mechanisms are systematic, they could be regarded as second order competences (Zollo &

Winter, 2002). The extant literature on DC has mostly an experiential emphasis and has examined the role of accumulated experience as the origin of organizational capabilities (Zollo

& Winter, 2002), the focus on experiential wisdom (Gavetti & Levinthal, 2000), the role of different experiential routines (Eisenhardt & Martin, 2000), experiential learning (Levinthal &

March, 1993), routines as the experiential lessons of history (Levitt & March, 1988) as well as shared experience (Nonaka, 1994).

In KM, OL is established as an important organizational means for continuous improvement of knowledge creation and utilization. Knowledge creation, retention, and

transfer processes can be found at the intersection of the complementary KM and OL fields

(Vera et al., 2011; Wu & Chen, 2014). A common conceptualization of OL differentiates exploitation as firms’ learning processes to improve existing products for existing customers and markets, and exploration as the learning efforts of firms to develop new products for new customers and markets. As the field has evolved, OL and KM researchers agree that organizations are more than the sum of individuals and that by acknowledging the existence of non-human repositories of knowledge, the capacity to learn, to know, and to have a memory (Walsh & Ungson, 1991) can be attributed to firms. The vital role in facilitating such learning is played by AC as the key dynamic capability that enables effective organizational learning through knowledge processing mechanisms (Cohen & Levinthal, 1990). The intersection among different academic concepts is illustrated below (Figure 1.13) and is characterized by the need to further establish the link between AC and different OL mechanisms (exploration and exploitation) as overlapping areas.

Figure 1.13 Boundaries and overlaps of the dynamic capabilities and knowledge management

DYNAMIC ORGANIZATIONAL KNOWLEDGE CAPABILITIES (DC) LEARNING (OL) & AC MANAGEMENT (KM)

w e

e a

Adapted from the study “Dynamic capabilities and knowledge management: an integrative role for learning” (Easterby‐Smith & Prieto, 2008)

2.13 Conclusion and research gap

Organizations dependent on innovation to survive and need to continually improve how they deal with and learn from any new knowledge sources. From early academic efforts, such as distinctive competence (Selznick, 1957), to the more recent and advanced notions of organizational routines (Nelson & Winter, 1982), absorptive capacity (Cohen & Levinthal,

1990), combinative capabilities (Kogut & Zander, 1996) and, finally dynamic capabilities

(Teece et al., 1997) there has been a vast amount of studies dealing with innovation and learning phenomena. In addition, organizational learning has always been a central issue affecting the functioning of organizations, and the role different mechanisms and techniques play in these learning processes is a continually evolving topic. By carefully delving into the underlying experiential learning theory, capabilities literature emphasized the issues of too much experience and associated redundancy, learning myopia or rigidity (Leonard‐Barton,

1992; Levinthal & March, 1993). However, experiential learning theory assumes that all knowledge is externally determined and that it already exists, such that there is no “wiggle- room for an organizational contribution (beyond past, accumulated external and experiential inputs)” (Felin and Foss, 2011, p. 12)

Search for external knowledge is arguably quite complex and difﬁcult, involving uncertainties and characteristics such as the tacitness, complexity, rivalry, and indivisibility of knowledge (Nonaka et al., 2000), which may not be conducive to its detection and transfer

(Zollo & Winter, 2002). The empirical literature on open innovation and search for external knowledge has been concerned mostly with downstream performance outcomes of knowledge search patterns (Patel & Van der Have, 2010). One group of work focuses on the rate, or number of new products that firms introduce, or innovativeness (Katila & Ahuja,

2002; Sidhu et al., 2007). Other work has examined the impact of search patterns on the magnitude of innovations (Ahuja & Morris Lampert, 2001; Rosenkopf & Nerkar, 2001) and

the effects of search behaviour on firm-performance (He & Wong, 2004; Rothaermel &

Alexandre, 2009). However, while these streams of literature have been concerned with firm- level outcomes from external knowledge search and have established important findings, far less is yet understood about the role that knowledge processes and capabilities play along the spectrum exploitation versus exploration.

Firms can use their prior knowledge as an input for innovations, or they may pursue external knowledge acquisition to extend their existing knowledge base (Capron & Hulland,

1999). Since its inception, the concept of AC has been closely linked with notions of organizational learning. Yet the precise nature of the relationship between these two concepts hasn’t been established. The extant literature considers AC as a dynamic capability that influences the firms' ability to create and deploy knowledge necessary to build other organizational capabilities (Jansen, Volberda, & Van Den Bosch, 2005b; Lane et al., 2006).

Even though research on AC offers “valuable insights into the organizational preconditions of successful knowledge integration, it leaves considerable conceptual and empirical aspects unresolved” (Volberda et al., 2010, p. 943). Therefore, many conceptual and empirical ambiguities remain due to limited work on the underlying process structures and the dimensions which shape it (Jiménez-Barrionuevo, García-Morales, & Molina, 2011; Patterson

& Ambrosini, 2015). Foss et al., (2010) highlights the need for research on feedback loops as a gap in previous empirical studies of AC and also argued that discussing AC merely as a capacity without discussing the actual processes that link it to outcomes variables such as patents, innovation and performance cannot be regarded as an integrated approach (Foss et al., 2010).

Absorptive capacity depends on processes and routines within the organization that enable it to share, communicate, and transfer individual-level learning to the organizational level. Routinization of organizational activities embeds capabilities into organizational memory, creating a unique configuration of firm resources (Knight & Cavusgil, 2004).

Research on routines, however, is hampered by a lack of empirical measurement relevant for validation of the construct, which poses a challenge for studying routines empirically (Becker,

Lazaric, Nelson, & Winter, 2005) especially those that underpin AC due to its intangible nature. Despite recent attempts to examine processes and routines that underpin AC (Lewin et al., 2011; Patterson & Ambrosini, 2015), the empirical testing of such routine-based model is still scarce. Understanding the choices between exploration and exploitation as two different types of OL are complicated by the fact that returns from the two options vary not only with respect to their expected values, but also with respect to deploying the appropriate routines embedded in AC. In addition, most of previous studies have empirically deployed

AC as a uniform construct. Consequently, scholars are no closer to understanding how organizations can deploy differing AC dimensions to manage the contradictory tensions of radical and incremental innovation effectively. It remains therefore unclear which bundling of

AC dimensions and underlying routines unlock exploitative and exploratory learning to drive incremental and radical innovation performance.

Chapter 3 – Conceptual framework

This chapter describes an overarching framework of the study by linking absorptive capacity and organizational learning theoretical lenses as outlined in an initial overview. The chapter is then divided in two parts: the first part focuses on theorizing AC as the construct and its distinct sub-process stages, while the second part outlines hypothesis development and conceptual model that links different types of AC to organizational learning and effects on innovation performance. Therefore, “exploitation view of absorptive capacity” and

“exploration view of absorptive capacity” are positioned as two competing theoretical arguments. These parts of the conceptual framework correspond to objectives of answering the research questions posed earlier in chapter one.

3.1. Overview

The survival of firms is a direct reflection of their ability to pursue enough exploitation to ensure the viability today, while engaging in exploration to ensure viability tomorrow (Levinthal & March, 1993; Lewin & Volberda, 1999). Focusing on exploration exclusively may lead firms to become trapped in what the literature has called “frenzies of experimentation, change, and innovation” (Levinthal and March, 1993, p. 105). Firms focusing on acquisition and assimilation of new external knowledge (i.e., potential AC) are able to continually renew their knowledge stock, but they may suffer from the costs of acquisition without gaining benefits from exploitation (Jansen et al., 2005a). On the other hand, firms that only focus on exploitation may fall into a competence trap where core competences become core rigidities (Leonard‐Barton, 1992). In other words, they might not be able to acquire new and external knowledge required for addressing rapidly changing environment. . For AC to take place, firms need prior knowledge to understand the new

knowledge that is absorbed (Cohen & Levinthal, 1990). But, for a firm to increase the level of

AC, it must possess the capability to transform and utilize external knowledge in order to enhance its core competencies. However, firms are not passive recipients of knowledge, and therefore commonly contribute their own knowledge to encourage knowledge exchange and collaboration with external parties (Ahuja, 2000). Such process is also likely to increase the firm level AC as its own knowledge contribution collaborative network (Tsai, 2009) will increase its ability to absorb knowledge from it. Thus, AC is crucial in explaining why some firms are better than others in creating and capturing value from in-sourcing external knowledge and collaboration with innovation partners.

Previous studies have examined the relevance of dedicated processes and internal mechanisms to effectively deal with external knowledge (Bianchi, Cavaliere, Chiaroni, Frattini,

& Chiesa, 2011; Foss, Laursen, & Pedersen, 2011). Lewin and Massini (2004) showed that innovative firms have a far superior capacity for learning than firms that are imitators. They highlight a strong link between a firm’s capacity for innovation and its AC (Lewin & Massini,

2004). This process occurs under uncertainty where firms may take cues from their external environment (Ozalp & Kretschmer, 2018) and induce learning by transferring knowledge and best practices back to the organization (Lewin et al., 2011). Szulanski’s (1996) work on the role of tacit knowledge in organizational learning argued that transferring best practices is important to the firm’s learning and competitive advantage. In relation to such transfer, the ability of a receiver of knowledge to unpack and assimilate it depends on whether or not the firm has overlapping knowledge bases with the source (Szulanski, 1996). Therefore, enhancing organizational learning via AC is a multi-stage process, as firms tend to identify the complementarity between their own knowledge and the acquired knowledge to integrate both for producing innovation outcomes.

Overall, it is evident that success of innovation is largely determined by the level of

AC and firm’s ability to upgrade it continuously through organisational learning. However,

there is a lack of empirical evidence in academic literature on 1) measures that constitute routine-based sub-process stages of AC and 2) how distinct types of AC unlock exploitative and exploratory learning in the form of incremental and radical innovation performance, which is a subject of further investigation in this study. Building on the conceptualization of

AC as a dynamic capability embedded in the systems, processes and routines of the organization (Lewin et al., 2011) and the fact that AC shares a high degree of conceptual affinity with OL, the study next 1) examines different AC dimensions and stages as the sub- process of the construct underpinned by organizational routines in order to operationalize the construct and 2) hypothesizes integration of AC and OL in relation to incremental and radical innovation performance as “exploitation view of absorptive capacity” and “exploration view of absorptive capacity”.

3.2. Part I – Absorptive capacity

Absorptive capacity represents a form of dynamic capability (Teece et al., 1997) that allows organizations to adapt to changing environmental circumstances (Nelson & Winter,

1982) through the acquisition, assimilation, transformation, and exploitation of external knowledge (Cohen & Levinthal, 1990; Zahra & George, 2002). Synonymous with these four dimensions of the AC construct, scholars have distinguished between distinct sub-process stages of AC as a means to explain the mechanisms of organizational learning. Zahra and

George (2002), for example, distinguish between potential and realized AC to differentiate between external and internal stages of the organizational learning process respectively.

Potential AC refers to the interaction between acquire and assimilate dimensions, which are embodied by capabilities and routines that support the sensing and seizing of external knowledge (Lewin et al., 2011; Teece, 2007). In acquiring new external knowledge, a firm tends to be constrained by its own existing knowledge base and past routines. This means that

it is advantageous to the firm to obtain unfamiliar knowledge in areas that are still somewhat closely related to its existing activities (Teece, 1986). Realized AC, on the other hand, refers to the interaction between transform and exploit dimensions, which are embodied by capabilities and routines that support the realization of learning benefits offered by potential AC

(Volberda et al., 2010).

Firms differ in their ability to deal with the external knowledge sources since AC can be understood as a firm-specific dynamic capability which is built over time (path-dependency) based on organizational routines (Winter, 2003). These higher-level routines are expressed within organizations by configurations of empirically observable practiced routines that are idiosyncratic, tacit and firm specific (Lewin et al., 2011). However, they also exhibit common features associated with effective processes across firms (Eisenhardt & Martin, 2000), which allows them to examine the commonalities across a sample of firms. Developing routines that enhance resource recombination and knowledge complexity (Galunic & Rodan, 1998) enables a firm to deal with more complex knowledge from external sources. This study argues that the dynamic capability view of AC is an example of OL in relation to external knowledge, and that each dimension of AC is a learning capability generated by specific processes and routines of the organization (Crossan et al., 1999). Zollo and Winter (2002, p. 340) define organizational capability as a “learned and stable pattern of collective activity through which the organization systematically generates and modifies its operating routines in pursuit of improved effectiveness”. Furthermore, routines are argued to be the most fundamental unit and building block of organizational capabilities (Winter, 2003). A set of organizational routines and strategic processes by which firms acquire, assimilate, transform, and exploit knowledge for the purpose of value creation (Zahra & George, 2002) is elaborated further.

Acquire

The acquisition phase refers to a firm's capability to identify and acquire critical knowledge beyond its boundaries. This is achieved through the broad and intensive use of external sources of knowledge through inter-organizational relationships (Love et al., 2014) of a diverse nature. Therefore, of considerable importance in this initial AC phase are the relations the firms maintain with different external agents that may act as knowledge sources (Grimpe

& Sofka, 2009). Overall, the acquisition dimension includes organizational meta-routines for identifying knowledge from the external environment, scanning and monitoring changes and developments from it, exchanging knowledge with partners, suppliers and competitors, as well as learning processes that occur in those interactions.

Assimilate

The assimilation phase refers to a firm's capability to analyse, process, interpret and understand the information obtained from external sources (Szulanski, 1996). Zahra and

George (2002) consider assimilation to be the process through which knowledge can be interpreted and understood owing to the existing cognitive structures. The role of qualified human capital is crucial with regards to the assimilation process of specific and tacit knowledge (Minbaeva et al., 2003). Hence, the assimilation dimension encompasses formal and informal organizational routines and activities related to transferring knowledge back to the organization, sharing internally knowledge from inter-firm relations, as well as reflection, replication and updating regimes.

Transform

The transformation phase is defined according to the routines responsible for the storage and recovery of knowledge when deemed convenient, as well as its suitable combination and the recognition of potential new uses (Jansen et al., 2005a). Therefore, the firms need to be able to

provide the interested parties with accumulated expressions of knowledge, allowing access to, and efficient use of the organizational memory through information systems and technologies or other alternative instruments (Volberda et al., 2010). As such, the transformation dimension enables firm’s combinative processes, replacing old practices and integrating new superior capabilities, including routines for sharing superior practices within and across subunits.

Exploit

The exploitation phase refers to a firm's capability in applying new knowledge, and thus achieving its set goals (Lane & Lubatkin, 1998). It involves a series of routines directly linked to the market demand that allow a firm to refine, extend, and leverage existing competencies or to create new ones by incorporating acquired and transformed knowledge into its operations (Lane & Lubatkin, 1998; Todorova & Durisin, 2007; Zahra & George, 2002).

Overall, these routines are related to reflecting on and updating existing categories of products, technologies, and processes as well as applying external knowledge to manage internal selection regimes and allocate resources for commercial gain (Distel, 2017; Lewin et al., 2011).

Extending previous AC conceptualization, Lane et al. (2006) add a further transformative sub-process stage of AC, which refers to the interaction between assimilate and transform dimensions that bridges the external AC (i.e., potential) and internal AC (i.e., realized) interface. According to Todorova and Durisin (2007), the oscillation between assimilate and transform dimensions at the transformative stage of AC is a core strategic imperative for internalizing external knowledge through the progressive integration and reconfiguration of knowledge (Patel, Kohtamäki, Parida, & Wincent, 2015; Teece, 2007).

They argue that this iteration familiarizes firms with new knowledge frames by stimulating the development of new “cognitive schemas to absorb knowledge that is less compatible with their prior knowledge” (Todorova and Durisin, 2007, p. 778). In the following sub-sections,

the study discusses these three different sub-processes of AC sequentially and the observable routines that underpin them. Such sub-processes of AC are direct counterparts of the sensing, seizing and reconfiguring capabilities (Teece et al., 1997). Therefore, 1) sensing, related to identification and assessment of opportunities and threats, is expressed through external AC;

2) seizing, related to the mobilization of resources to address opportunities and threats, is expressed through internal AC; and reconfiguring, related to the continuous renewal of a firm’s tangible and intangible assets, is expressed through transformative AC.

1) External absorptive capacity

External absorptive capacity refers to a firm's capability to acquire and assimilate externally generated knowledge through diverse inter-organizational relationships, routines and learning processes that allow to identify, analyse, process and understand the knowledge obtained from external sources (Szulanski, 1996; Zahra & George, 2002). Therefore, the external stage of AC emphasizes a firm’s ability to acquire and assimilate critical knowledge beyond its boundaries through diverse inter-organizational relationships (Laursen & Salter,

2006; Love et al., 2014). Acquisition refers to an ability to identify and recognize critical sources of external knowledge and hinges on the development of routines for scanning and monitoring the external environment (Lewin et al., 2011). Such routines include the utilization of internal gatekeepers that serve as formal or informal boundary spanners for the organization (Cohen & Levinthal, 1990), active engagement in external knowledge searches through market research or interactions with external partners (Laursen & Salter, 2006), and secondary patent search and analysis (Cohen, Goto, Nagata, Nelson, & Walsh, 2002), as well as other forms of low-cost environmental probing as a means to engage in external scanning.

Assimilation, on the other hand, works in conjunction with acquisition during the external stages of AC, and refers to the firm’s ability to analyse, process, interpret, and understand knowledge that is externally acquired (Zahra & George, 2002, p. 189). This 90

involves developing routines for interpreting and understanding externally acquired knowledge, such as formal training programs and utilization of inter-firm communities and steering groups (Lewin et al., 2011). Due to its sticky nature some knowledge might require further refinement through processes of codiﬁcation of knowledge and past experience

(Levitt & March, 1988), retrospective sense making, or the imitation of competitors’ superior practices (Szulanski, 1996). Thus, external AC involves iterative cycles of acquire and assimilate to search for and understand external knowledge, which mainly occurs through trial-and-error learning (Rerup & Feldman, 2011) for interpreting external knowledge.

2) Transformative absorptive capacity

The second, transformative stage of AC refers to a firm’s ability to assimilate and transform external knowledge to develop new or extend existing stocks of knowledge, replacing old practices and integrating new superior capabilities (Zahra & George, 2002). This stage involves recursive cycles of understanding newly acquired knowledge through assimilation routines, and then projecting these new insights into new organizational schemas through transformation routines (Harris, 1994; Lewin et al., 2011).While the benefits of external knowledge that is first acquired and assimilated during the external stages of AC offer transformative potential, such knowledge is often difficult to transfer due to its

‘stickiness’(Szulanski, 1996), especially if such knowledge is distant and unfamiliar. To overcome this challenge, firms must possess transformation capabilities as the building blocks that create new core competencies through the continued interaction and mutual adjustment of learning mechanisms (Zollo & Winter, 2002).

These capabilities allow them to pivot on existing institutionalized understandings of knowledge in light of newly assimilated knowledge. Such capabilities are built through exploratory learning activities, primarily concerned with variation - generating the necessary range of new intuitions and ideas as well as selection of the most appropriate ones through 91

evaluation and legitimization processes (Nelson & Winter, 1982; Zollo & Winter, 2002). This involves developing routines for transferring knowledge back to the organization and internal knowledge sharing (Lewin et al., 2011) such as purposeful efforts for communicating and distributing knowledge among various stakeholders and departments, knowledge integration practices, and utilization of cross-functional teams. These routines enhance possibilities for creative and novel recombination of knowledge that stimulate exploring and learning new ways (March, 1991) as well as the development of new organizational understandings.

3) Internal absorptive capacity

Internal absorptive capacity refers to a firm's capability to transform and exploit newly generated knowledge by refining, extending, and leveraging existing competencies or creating new ones by incorporating transformed knowledge into its operations (Lane & Lubatkin,

1998; Zahra & George, 2002). Therefore, the final internal stage of AC emphasizes a firms’ ability to transform and exploit knowledge to drive its commercial application. This stage involves the deployment of recursive cycles of transformation with exploitation capabilities, which collectively allow the firm to incorporate assimilated knowledge into its existing operations (Lane & Lubatkin, 1998). Such capabilities are built through exploitative learning activities which rely on behavioural mechanisms encompassing the replication of the new approaches in diverse contexts and their absorption into the existing sets of routines for the execution of that particular task (Nelson & Winter, 1982; Zollo & Winter, 2002). This type of learning takes place inside a firm’s focal knowledge domain and relates to refinement, choice, efficiency and improvement (March, 1991). The process of exploitative learning involves developing routines that enable for reflecting on and updating existing categories of products, technologies, and processes as a result of external search, and applying external knowledge to guide decisions to allocate resources for commercial gain (Distel, 2017; Lewin et al., 2011).

Thus, during this stage, knowledge is transformed into a form that fits the organization’s 92

operations so that it can be continuously refined and exploited (Zahra & George, 2002) as a means for its commercial application. The table below summarizes theoretical background and conceptual framework related to AC across its locus, dimensions and observable firm- level routines, which is used for subsequent operationalization of the construct (Table 1.4).

Table 1.4 Summary of AC theoretical background, its dimensions and routines (develop for the purpose of the study)

AC Locus AC dimensions AC routines

• Gatekeepers (Cohen & Levinthal, 1990) Acquire • Probing (Brown & Eisenhardt, 1997)

(Identify/Recognize) • Mining patent literature and industry trade magazines

• Market research, end user surveys, informal interactions with industry actors

Refers to a firm's capability • Co-development relationships (Koza & Lewin, 1998)

to identify and acquire • Collaborating with “lead users” (Von Hippel, 1986) externally generated • Collaborating with suppliers

knowledge that is critical to • R&D partnerships its operations (Lane & • Networking with outside organizations, universities, research institutions Lubatkin, 1998; Todorova • Unﬁltered information from key clients to CEO (Lewin et al., 2011) & Durisin, 2007; Zahra & • Open source (e.g., innocentive.com) (Lewin et al., 2011) George, 2002) • Occupying leadership roles in standard setting industry organizations

(Rosenkopf & Nerkar, 2001)

External

Assimilate

(Interpret) • Sharing within company knowledge acquired in interﬁrm relations

(Rosenkopf & Nerkar, 2001) • Pacing the partner (Koza & Lewin, 1998) Refers to the firm's routines • Engaging external secondments, sabbaticals & free lancers (Criscuolo, 2005) and processes that allow it to analyse, process, • Problemistic and local search interpret, and understand • Learning from good and bad experience the information obtained • Learning from managing alliances (Zollo & Winter, 2002)

from external sources • Learning programs (in-house & external training) to increase the knowledge base of the firm (Minbaeva et al., 2003) (Szulanski, 1996; Todorova • “Copy exact” principle to leverage optimization of processes across units & Durisin, 2007; Zahra &

George, 2002)

AC Locus AC dimensions AC routines

Transformative

• Central provision of information on value of specific new practices and on Transform their implementation: brochures, liaisons between corporate groups, BU and (Integrate/Combine) facilities (Lenox & King, 2004) Refers to a firm’s capability • IT-based knowledge codification system to store and manage knowledge, and to develop and refine the retrieve it for future needs (Lewin et al., 2011) routines that facilitate the • Visit other firm’s divisions to integrate knowledge (Jansen et al., 2005a) combination of existing • Cross-functional project teams (Freeman, 1987) internal knowledge and •Solicitation of scientists and engineers to propose and pursue innovative ideas potential external (Lewin et al., 2011) knowledge (Hughes & • Open office plan chosen to foster informal interactions Wareham, 2010; Sun & (Lewin et al., 2011) Anderson, 2010; Todorova • Technology Forum and Technical Council (Lewin et al., 2011) & Durisin, 2007; Zahra & • Brainstorming sessions to bring together persons with different technical/ George, 2002) market knowledge (Lewin et al., 2011)

Internal Exploit (Commercialize) • Shared sense of PC ecology boundaries to determine project funding (Lewin et al., 2011) Refers to the firm’s routines • Seeking market signals (Lewin et al., 2011) that allow a firm to refine, • Development of prototypes that perform at least as well as what is available extend, and leverage on the market (Lewin et al., 2011) existing competencies or to • Autonomy of middle management to support and allocate resources to create new ones by projects outside CEO’s vision (Rotemberg & Saloner, 2000)

incorporating acquired and • Internal rate of change greater than external rate of change transformed knowledge into (Lewin et al., 2011) its operations (Lane & • One/two comparison benchmark/Comparison to industry best in class vs Lubatkin, 1998; Todorova industry average (Massini, Lewin, & Greve, 2005) & Durisin, 2007; Zahra & • Stretch goals (Porras & Collins, 1997) George, 2002

3.3 Part II - Hypotheses development

The Exploitation-Based View of Absorptive Capacity

Exploitation is associated with the commercialization of existing knowledge inside and outside of the firm’s boundaries (Argote, McEvily, & Reagans, 2003) through combining external sources of knowledge that are closely related to the organization’s existing capabilities

(Stuart & Podolny, 1996). According to prior studies, such learning occurs through engaging in local searches, experiential refinement, and step wise improvements within an existing knowledge paradigm (Benner & Tushman, 2002; Gavetti & Levinthal, 2000; Katila & Ahuja,

2002). Thus, exploitation is associated with the continued elaboration and deepening of organizationally tacit knowledge along established trajectories of improvement characteristic of incremental innovation (Levinthal & March, 1993). This repeated, step-by-step reuse and extension of the organization’s tacit knowledge base leads to the development of routinized behaviours that are stored as procedural memory (Cohen & Bacdayan, 1994), such that efforts to acquire and assimilate, and assimilate and transform external knowledge for the purposes of incremental innovation become quasi-automatic (Zollo & Winter, 2002). Thus, exploitation primarily occurs in the proximity of existing knowledge and previous solutions (Stuart &

Podolny, 1996) that increase the chances of applying workable improvements to existing products.

This mechanism of exploitative organizational learning is therefore the outcome of reliable and predictable search behaviours within external knowledge domains that are local and inherently familiar to the organization (Katila & Ahuja, 2002; Levinthal & March, 1981;

Nelson & Winter, 1982). Such familiar patters increase the likelihood for combining new external knowledge (Zollo & Winter, 2002) and applying best practices to make knowledge more efficient to exploit, and to accelerate its implementation (Zander & Kogut, 1995). The

study argues, therefore, that the interaction between capabilities of acquire and assimilate that constitute the external stages of AC, and assimilate and transform that constitute transformative stages of AC, do not contribute towards performance gains in incremental innovation that result from exploitative learning. This is because the organization’s response to local knowledge is administratively embedded within taken-for-granted practices and routines, such that no cognitive effort nor specialized capabilities to support the purposeful identification, acquisition, interpretation, and understanding of external knowledge sources

(Grimpe & Sofka, 2009; Katila & Ahuja, 2002) are required at the external and transformative interface for knowledge that is already internally tacit (Argote & Ingram, 2000; Nonaka,

1994).

According to Zollo and Winter (2002: 341), “incremental improvements can be accomplished through the tacit accumulation of experience and sporadic acts of creativity.”

This suggests that exploitative learning outcomes are more likely to occur through the interactions between internal AC dimensions of transform and exploit, as organizations are required to cycle through capabilities and routines that lead to new interpretations of known knowledge and the subsequent institution of those new interpretations into existing operations and cognitive schemas (Zahra & George, 2002). This process fosters the cohesion of knowledge, communication and efficiency of knowledge exchange throughout units

(Galunic & Rodan, 1998) allowing to transform and exploit new external knowledge.

Additionally, as these routines are well-practiced and predictable, they permit closely coordinated exploitation of knowledge in pursuing collective objectives (Grant, 1996).

Therefore, internal AC dimensions of transform and exploit will create synergistic effects in that they drive incremental innovation performance gains. Accordingly, the following hypothesis is posed:

Hypothesis 1 (H1): Internal absorptive capacity characterized by the interaction between transform

and exploit is positively associated with incremental innovation performance. 97

The Exploration-Based View of Absorptive Capacity

While exploitative learning is the outcome of tacit accumulation of organizational experience (He & Wong, 2004; Zollo & Winter, 2002) this approach does not translate to exploratory learning, which is characterized by risk-taking, discovery, and variation (March,

1991) in light of unusual experiences that contradict, or represent a significant break from, known categories of experience (Garud et al., 2011; Patel et al., 2015; Winter, 1984). Under such conditions, the organization’s existing cognitive schema and routines that support their re-enactment are incongruent with the external environmental reality of change, and firms must purposely invoke mechanisms that go beyond the semi-automatic stimulus-response process and tacit accumulation of experience (Zollo & Winter, 2002) in order to learn from distant sources of unknown knowledge. According to Ahuja and Morris Lampert (2001), failure to break from these quasi-automatic responses in light of unknown experiences leaves firms at risk of three types of competency traps: familiarity, maturity, and propinquity. These traps collectively occur owing to a bias towards the firm’s existing knowledge base (Zahra &

George, 2002). Hence, familiarity traps result from a heightened proclivity towards refining existing knowledge; maturity traps result from a need for predictability and reliability inherent in existing knowledge; and propinquity traps result from a disposition toward searches within local knowledge domains (Ahuja & Morris Lampert, 2001).

To avoid such traps, organizations must scan for a variety of external opportunities from distant, unfamiliar knowledge domains as markets and technologies change, which requires purposeful and conscious efforts to acquire and assimilate (Ahuja & Katila, 2001;

Szulanski, 1996), as well as distinct capabilities to transform the firm’s existing knowledge base (Eisenhardt & Martin, 2000). Such capabilities are underpinned by exploration routines that occupy the crucial nexus (Feldman & Pentland, 2003) between assimilate and transform stages of AC, thus reducing uncertainty and volatility of the external environment (Becker &

Knudsen, 2005). Radical innovation may involve a higher degree of discontinuity in the sources of innovation, since previously used knowledge might become obsolete in the new context (Christensen, 2013). Therefore, it is through capabilities of assimilation and transformation at the transformative stages of AC that the firm is able to first understand and

“assimilate” external, unfamiliar knowledge; and second, able to “transform” and renew the firm’s existing knowledge structures and schemas to be capable of perceiving unfamiliar knowledge to stimulate radical innovation (Todorova & Durisin, 2007). This transformation process changes the character of knowledge through bisociation, which occurs when new knowledge is created through the bricolage of ‘old’ and ‘new’ cognitive frames of reference

(Koestler, 1966; Zahra & George, 2002) that redirect organizational attentional processes

(Ocasio, 2011) away from the quasi-automatic stimulus-response mechanisms that are relevant for exploitative learning.

At the transformative stage, organizations need to transfer and recombine the new knowledge with its existing stocks of knowledge (Afuah & Tucci, 2012), thus experimenting and making sense of unexplored interdependencies among knowledge fragments. According to Todorova and Durisin (2007), the oscillation between assimilate and transform dimensions at the transformative stage of AC is a core strategic imperative for internalizing external knowledge through the progressive integration and reconfiguration of knowledge (Patel et al.,

2015; Teece, 2007). The integration of existing tacit knowledge with entirely new knowledge as experienced from unpredictable or rare events enables firms to facilitate exploratory learning (Subramaniam & Youndt, 2005). This suggests that exploratory learning outcomes are more likely to occur through the interactions between transformative AC dimensions of assimilate and transform. Therefore, the transformative stage of AC with focus on exploratory learning is likely to result in a change in theory-in-use and govern action (Argyris

& Schön, 1997) to create a more pioneering outcomes as a means of achieving radical innovation.

While the study asserts on the one hand that this transformative stage of AC will induce positive effects on radical innovation performance, it argues that such benefits to performance will be eroded if high transformative AC is coupled with high levels of exploitation, as the new knowledge generated through bisociation during transformative stages will be reversed when casted through existing exploitation routines at internal stages.

Such routines are invariable, repetitious, and handle lower frequencies of unexpected and novel events (Withey, Daft, & Cooper, 1983), addressing narrow range of problems

(Volberda, 1996). In addition, exploitation routines limit the search for new external knowledge and lead to a narrow scope of information processing. Consequently, internal AC hampers the ability to tap into new external knowledge sources and impedes an organization’s ability to interpret and combine new external knowledge. It decreases the probability of

ﬁnding novel solutions and leads to the acceptance of the status quo with a high level of conformance (Ashforth & Saks, 1996) now allowing a new shared understanding to evolve within the organization. Therefore, exploitative learning process and routines in the internal stage of AC act as core rigidities to dealing with new knowledge for the purposes of radical innovation. Hence, the following hypotheses are proposed:

Hypothesis 2a (H2a): Transformative absorptive capacity characterized by the interaction between

assimilate and transform is positively associated with radical innovation performance.

Hypothesis 2b (H2b): High levels of exploit in the internal stage of AC will erode the

performance benefits of radical innovation attributed to transformative absorptive capacity.

100

3.4 Absorptive capacity and organizational learning (conceptual model)

The importance of achieving competitive advantage through the pursuit of exploratory and exploitative learning has become one of the most enduring themes in management and innovation research (Afuah & Tucci, 2012; March, 1991; Wang et al., 2017).

Organizational behavioural capacity to simultaneously demonstrate exploratory and exploitative learning across an entire system has been referred to as the contextual ambidexterity (Gibson & Birkinshaw, 2004). In order to balance exploration and exploitation

(Tushman & O'Reilly, 1996) it is important that firms have “the ability to simultaneously pursue both incremental and discontinuous innovation and change results from hosting multiple contradictory structures, processes, and cultures” (Tushman and O’Reilly, 1996, p.

24). However, success of an organization to pursue both exploration and exploitation depends on whether these two types of learning are treated as competing or complementary aspects of organizational decisions. This study recognizes the dichotomy of exploration and exploitation as key drivers of firms’ innovation activities and performance (Gupta et al., 2006;

Sidhu et al., 2007) which rely on contradictory capabilities and routines.

Prior studies (Grimpe & Sofka, 2009; Katila & Ahuja, 2002) acknowledge the importance of firm-level absorptive capacity (AC) in order to drive organizational learning

(Lane et al., 2006). Organisations that create knowledge on an ongoing basis are likely to develop dynamic and unique capabilities that potentially facilitate continuous organisational learning in developing their technological advantages. From an organizational learning lens, firms from advanced economies normally develop technological advantages by relying mostly on cumulative experiential learning (Piperopoulos, Wu, & Wang, 2018). This study argues that the dynamic capability view of AC is an example of OL in relation to external knowledge, and that each dimension of AC is generated by experiential learning processes that are embedded in routines of the organization (Crossan et al., 1999; Sun & Anderson, 2010). Therefore, firms

101

learn in different ways subject to their learning ability, prior experiences and the knowledge base (Barkema & Vermeulen, 1998).

Exploitation refers to learning gained via local search, experiential refinement and improvements in existing components, which build on the current technological trajectory, aimed at improving present product-market domains (Baum et al., 2000; Benner & Tushman,

2002; He & Wong, 2004). In other words, exploitative learning takes place inside a firm’s focal knowledge domain and relates to refinement, choice, efficiency, improvement, and implementation characteristic of incremental innovation (Baum et al., 2000; March, 1991). On the other hand, exploration includes learning gained through processes of concerted variation, planned experimentation, and play, with a shift to a different technological trajectory, aimed at entering new product-market domains (Baum et al., 2000; Benner & Tushman, 2002; He &

Wong, 2004). Thus, explorative learning takes place outside of a firm’s focal knowledge domain and relates to search, variation, play, and experimentation characteristic of radical innovation.

The challenge of managing organizational learning processes can be attributed to firms’ constrained capabilities to acquire, assimilate and transform external knowledge and convert it effectively into relevant innovation outcomes (Lane et al., 2006; Zahra & George,

2002). Absorptive capacity enables the firm’s identification of the value of new external know-how in different contexts (e.g. customers, suppliers, competitors, strategic alliance partners, joint ventures, etc.) and its incorporation that is ‘‘largely a function of the firm’s level of prior related knowledge’’ (Cohen and Levinthal, 1990, p. 128). Therefore, by deploying AC firms can expand their learning by collaborating with external partners and tap into their knowledge base. Higher AC is likely to improve their innovation performance by cognitively processing various sources of knowledge, integrating and applying knowledge to commercial ends. Due to the strong link between OL and AC, several previous studies operationalized the

AC construct using an organizational learning lens (Cohen & Levinthal, 1990; Lane et al.,

102

2006; Zahra & George, 2002). Previous research also implicitly recognizes the multiple levels of learning and the path-dependent nature of AC (Jansen et al., 2005a) underpinned by routines (Lewin et al., 2011).

Absorptive capacity depends on processes and routines (Gavetti & Levinthal, 2000;

Winter, 1964) within the firms that enable the organization to share, communicate, and transfer individual-level learning to the organizational level. The early definition of a routine - a “pattern of behavior that is followed repeatedly, but is subject to change if conditions change” (Winter, 1964, p. 263) reflects the emphasis on the environment’s role in determining organizational learning and change. Furthermore, routines “reflect experiential wisdom in that they are the outcome of trial and error learning and the selection and retention of past behaviors” (Gavetti and Levinthal, 2000, p. 113). The dynamic nature of AC, rooted in organizational routines, ultimately strengthens the firm’s capability to learn and benefit from new knowledge and novel technologies within its environment. For the purpose of guiding exploration and exploitation (March, 1991) understanding the nature of AC sub-process stages and underpinning routines represents a key dynamic capability for firms. This argument does not suggest that firms must adopt the entire bundle of AC routines. Although this may have consequences for innovation performance by not realizing the full potential from exploiting the complementarities among routines, previous research suggests that firms may only adopt some elements of the proposed configuration (Lewin et al., 2011). However, it is unclear which configuration is the most appropriate for driving the exploratory versus exploitative learning outcomes with regards to radical and incremental innovation performance. In other words, what is missing is a clear articulation of specific AC dimensions that facilitate exploration and exploitation (March, 1991).

To address this research problem, the next section proposes two theoretical arguments and hypotheses, which represent the “exploitation-based view of absorptive capacity” and the “exploration-based view of absorptive capacity”. Specifically, the study

103

theorizes that the interaction between internal AC dimensions of knowledge transformation and exploitation is most critical for driving incremental innovation performance from a more passive, exploitative form of learning. The study argues that efforts to cycle through routines of acquire and assimilate characteristic of external AC, and assimilate and transform characteristic of transformative AC, are not required for knowledge that is tacitly and experientially familiar. Conversely, the study asserts that the interaction between assimilate and transform dimensions, which occurs at the interface between internal and external AC

(Lewin et al., 2011) is the most critical for driving radical innovation performance from a more deliberate form of exploratory learning. This is because firms must engage in conscious and substantial efforts to assimilate and transform knowledge that is both tacitly and experientially distant or unfamiliar (Garud et al., 2011). The study argues, however, that such performance benefits may be eroded in the presence of strong exploitation routines that are tethered to knowledge that is tacitly and experientially familiar. The proposed theoretical arguments and hypotheses are illustrated in the conceptual model below (Figure 1.14).

Figure 1.14 Conceptual model (developed for the purpose of the study)

Exploitative learning

Incremental Internal AC H1 + innovation (transform and exploit) performance

Exploratory learning

Transformative AC

(assimilate and transform) H2a + Radical

innovation H2b - performance Transformative AC and exploit

104

Chapter 4 – Methodology

4.1. Research paradigm

The academic research is based on philosophical assumptions about what constitutes a valid research and which methods are suitable for studying the phenomenon. Such assumptions relevant for development of the knowledge in a particular study are referred in literature as a research paradigm (Lincoln & Guba, 2000; Scotland, 2012). It is described as an interpretative framework, guided by a set of beliefs and feelings about the world and how it should be understood and studied (Guba, 1990). There are three categories of those beliefs: 1) ontology that deals with the question of what is real, 2) epistemology that examines the relationship between the inquirer and the known and 3) methodology that deals with the questions related to the techniques used to know the world and gain knowledge of it (Lincoln

& Guba, 2000). In addition to the ontological, epistemological and methodological stance, the research design also includes methods - data collection and analysis tools (Scotland, 2012).

Therefore, the choice of philosophical assumptions is further elaborated with special attention on justification and appropriateness for this study.

4.2. Ontology

This study mainly adopts elements of critical realism that retains an ontological realism – ‘real’ reality but only imperfectly and probabilistically apprehendable (Ahmad, Umar

Bello, Kasim, & Martin, 2014). A critical realist believes that there is a reality independent of our thinking, perceptions and constructions that the science can study (Putnam, 1990).

However, it is different from the ‘naive’ realism that assumes ‘real’ reality which is apprehendable (Ahmad et al., 2014). Critical realism on the other hand rejects a relativism

105

strand – our understanding of this world is inevitably a construction from our own perspectives and standpoint. (Niiniluoto, 1991). The philosophical assumptions pertaining to critical realism in this study are concerned with a structured way of thinking when describing the AC-organizational learning-innovative performance links, considering that causality exists as the potential. On the other hand, as the study includes innovation practitioners it recognizes that social life is both generated by the actions of individuals, and has an external impact on them (Ackroyd & Fleetwood, 2000). Such position is an alternative to both naive realism mentioned before and to radical constructivist views that deny the existence of any reality apart from our own constructions (Ahmad et al., 2014). Taking this perspective in the research has also been given explicit philosophical defenses. The argument is summarized as the following: “empirical science is an enterprise that seeks to develop images and conceptions that can successfully handle and accommodate the resistance offered by the empirical world under study” (Blumer, 1969). In economics and management studies, critical realism points to the key limits of neoclassical economics based on econometrics principles that are reductionist in nature. Therefore, it provides a philosophical and methodological foundation for a broad set of alternative approaches (Maxwell & Mittapalli,

2010).

4.3. Epistemology

Guided by the assumptions of critical realism and detached approach in the research, the driving epistemology behind the study is post positivism (Gasson, 2009). It is therefore less strong version of positivism that characterizes the research due to the fact that the patterns will be identified through conducting a sample survey, rather than undertaking a true experiment. Unlike strong positivists guided by ‘naive’ realism assumptions, the post positivism implications distinguish most contemporary realist approaches from positivists

106

(Baert, 1998). The difference is that the post-positivist critical realist recognizes that all observation is imperfect and has error, that all theory is revisable, being critical of ability to know reality with certainty (Easterby-Smith, Thorpe, & Jackson, 2015). The reason for adopting this approach in the study is to maximize its objectivity and acknowledge reported information from the survey. Such detached approach in the study allowed setting aside personal experiences and perceptions of the researcher, thus minimizing biases in conducting the study. Due to the wider criteria for data acceptability than is the case for positivism, post- positivism is used where survey information is categorized to produce quantitative data to be analyzed using statistical methods (Gasson, 2009). This is appropriate for the study since the data related to the key research concepts such as AC and innovative performance is represented by the variables to analyze correlation among them.

4.4 Research methods

The empirical study is undertaken through the application of quantitative research methods (Bryman, 2004). These methods are considered appropriate as they attempt to maximize objectivity, replicability, and generalizability of findings related to innovation phenomenon (Harwel, 2010). Quantitative research assumes using instruments such as tests or surveys to collect data, and reliance on probability theory to test statistical hypotheses that correspond to research question (Harwel, 2010). The advantage of using this method is maximizing the data reliability and validity by asking the same questions to a large number of innovation practitioners. This was achieved in a brief period of time and in a relatively cost- effective manner, allowing generalizability of the results and objectivity for testing hypothesis

(Ackroyd & Hughes, 1981). Therefore, the nature of this study is deductive and is concerned with testing the hypotheses drawn from the theory. This approach enabled general inferences related to interactions outlined in the conceptual model, explaining correlation amongst

107

variables. The survey is assumed to offer a series of highly reliable and valid measures of an organization’s innovation activities and performance ascribed to exploitative and exploratory forms of learning. It was designed to capture organizational-level constructs relating to firms’ innovation activities and was therefore broadly based on the Community Innovation Survey

(CIS) and was further enriched with AC variables.

As data used for the purposes of this study is obtained from a single survey, it raises issues concerning response and non-response biases. Previous studies noted that the characteristics of late respondents tend to be similar to those of non-respondents (Lankford,

Buxton, Hetzler, & Little, 1995; Ullman & Newcomb, 1998). For this purpose, the wave analysis technique was followed to check for non-response bias, by comparing the responses of early and late respondents in the survey (Rogelberg & Stanton, 2007). In addition, confidentiality was assured to respondents in order to reduce potential social desirability. Any possible ambiguity from the item was removed by pre-testing the questionnaire and making appropriate adjustments to the language, where necessary. To minimise survey error due to satisficing, acquiescence and response order, procedures were followed for the presentation and structure of the items and designing the response scales for the survey questionnaire

(Blasius & Thiessen, 2012). Further procedures for data collection were followed to ensure the quality of survey data and minimise the bias from common method variance (Podsakoff,

MacKenzie, Lee, & Podsakoff, 2003).

4.5 Survey methodology

For the purpose of examining relations among variables, the study adopts a survey methodology (Pinsonneault & Kraemer, 1993). Survey research is a specific type of field study that involves the collection of data from a sample of elements drawn from a well- defined population using a questionnaire for lengthier discussions. Survey is considered a

108

valuable methodology for assessing opinions and trends, appropriate due to the low cost and accessible information (Pinsonneault & Kraemer, 1993). The study follows the best practice of the Community Innovation Survey (CIS) methodology, which has been successful in providing information on innovation of firms in the EU context and beyond. The CIS is a microdata questionnaire that measures organizational-level innovation activities, as described in the Organization for Economic Co-operation and Development’s (OECD) Oslo Manual

(OECD, 2005), and has been used in several academic articles in strategy and innovation management (Cassiman & Valentini, 2016; Laursen & Salter, 2006; Love et al., 2014). It is considered the main data source for measuring innovation in Europe and common methodological approach to innovation research according to the Oslo Manual (OECD,

2005). Taking into account how innovation activities are usually organised, the enterprise

(firm) is in general the most appropriate statistical unit (OECD, 2005) which is adopted in this study as well.

4.6. Sample and Data Collection

The data used for this study were obtained from a primary survey questionnaire distributed to senior practicing innovation managers and innovation executives among a sampling frame of 715 publicly listed firms that were affiliated with individual members of a global professional innovation management association. A link to the survey was distributed through the professional association’s mailing list with the sponsorship and support of the executive director in the summer of 2017 and was self-administered via Qualtrics. In addition to gaining executive sponsorship for survey instrument, a number of further steps were taken to improve response rates, including personal discussions with respondents, presentations at association events leading up to the survey’s launch, and extensive follow-up activity. In the data collection phase, particular attention was paid to checking the reliability and consistency

109

of data and to follow-up or reminder procedures. Additionally, the respondents were motivated to answer as accurately as possible by the option of receiving a benchmark report and the was accompanied survey by an official university cover letter. Non-respondents were sent two email reminders and offered the option to provide their answers over the telephone if preferable to online. Of the 715 firms within the original sampling frame a total of 309 participated in the survey, yielding a response rate of 43%.

To ensure comparability, surveys must specify an observation period for questions on innovation. It is recommended that the length of the observation period for innovation surveys should not exceed three years or be less than one year (OECD, 2005). Similar to the

CIS, all survey questions regarding organizational-level innovation activities in this study refer to the average of a two-year period spanning 2016 and 2017, while measures relating to innovation performance were only ascribed to 2017 as to account for the time lag that exists between innovation activities and outcomes (Hess & Rothaermel, 2011). In other words, the innovation outcomes were required to be evaluated only for the last year of that period. By doing so, the study attempts to temporally separate the independent and the dependent variables in order to account for time lags between innovation activities and innovation outcomes. This also adds to overcoming potential common method bias concerns.

4.7. Measures

The questionnaire is developed using standardized measures (scales and subscales) from the literature, also including the CIS items. CIS measures are gathering information on innovation activity in a “subject-oriented” way (Laursen, 2011:717) by asking respondents about the innovation output of their firm. This is seen as complementary way to traditional measures to elicit information about innovation performance. For example, besides the fact that not all innovations are necessarily patented, patent data analysis rather offers a proxy for

110

firms’ appropriability strategy than for their innovation performance (Laursen and Salter,

2006). In contrast, CIS measures indicate whether firms have actually been able to benefit from their innovation activities. Firms are asked to disclose information on the share of revenue that they ascribe to innovations with different degrees of novelty.

The measures used in this study are partly based on published scales and partly on newly developed items. Therefore, the measures used in this study are predominantly derived from existing variables and scales relating to innovation activities and performance, and partly on items that are theoretically and conceptually verified in published academic studies for the separate dimensions of AC. The survey was enriched with measures of individual dimensions of AC gleaned from existing measures and conceptualizations of the construct from the extant literature. These measures of AC dimensions were empirically verified through a confirmatory factor analysis (CFA). Before sending out the survey, the questionnaire was pre- tested with the help of a group of managers from innovation management backgrounds.

Thus, both AC measures and the whole survey were comprehensively and repeatedly pre- tested. The feedback was used to adapt and improve the questionnaire before its final distribution to firms.

Dependent Variables

A large number of typologies have been proposed to examine the degree of novelty of innovation in products or processes (Garcia & Calantone, 2002), which causes specific measurement problems (Danneels & Kleinschmidtb, 2001). Laursen and Salter (2006) measured innovative performance as a proportion of turnover related to new products, which is only one aspect of innovative performance. Furthermore, Atuahene-Gima (2005) argued that the concept of innovative performance should include a number of different aspects such as: the number of new product innovations introduced by the firm, the percentage of sales of new product innovations, and the relative frequency of introducing innovations compared

111

with competitors (Atuahene-Gima, 2005). Consistent with typologies for innovation practice

(Garcia & Calantone, 2002) both “new to the world” measures and “new to the ﬁeld” measures are suggested as new measures of innovation performance

To test the competing hypotheses in this study relating to variations in the role of AC for exploitative and exploratory learning, two measures of innovative performance are operationalized to distinguish between performance attributed to incremental innovation

(exploitative learning), and performance attributed to radical innovation (exploratory learning). To account for a firm’s innovation success, the study relies on self-reported measures and use proxies that aim at reflecting the different degrees of novelty of a firm’s product innovations, as they are also used in the CIS instruments. All these questions refer to the fraction of a firm’s turnover stemming from the respective innovation type. Therefore, following prior studies that leverage the CIS (Laursen & Salter, 2006; Leiponen & Helfat,

2010) radical innovation performance was measured as the proportion of organizational turnover attributed to new to world product sales and incremental innovation performance as the proportion of organizational turnover attributed to new to firm product sales. Given the skewed nature of both these variables, a logarithmic transformation was performed for subsequent analysis.

Independent Variables

To measure the four separate dimensions of the AC construct pertaining to acquire, assimilate, transform, and exploit, the study built on a combination of existing measures

(Distel, 2017; Jansen et al., 2005b) and conceptualizations from the extant literature

(Todorova & Durisin, 2007; Zahra & George, 2002), and then adapted them in order to represent observable, organizational routine-based operationalization (Lewin et al., 2011).

Absorptive capacity (independent variable) dimensions are subjected to confirmatory factor analyses (CFA). Therefore, a four-factor model of the independent variables was assessed as

112

well as each of them separately consisting of 4 items each based on conceptual routine-based model (Lewin et al., 2011).

AC was operationalized with the four proposed dimensions (Zahra & George, 2002) relying on existing items (routines) for these dimensions (Lewin et al., 2011) and in line with conceptual discussions in AC research (Lane et al., 2006; Todorova & Durisin, 2007). First, the four-item scale for acquire addresses a firm’s efforts to acquire new knowledge from external sources; second, the four-item scale for assimilate gauges a firm’s ability in analysing and understanding new external knowledge; then, the four-item scale for transform reflects the extent to which a firm is able to combine existing knowledge with new knowledge; and finally, the four-item scale for exploit assesses a firm’s proficiency in exploiting new knowledge and applying it in new products and technologies.

Control Variables

Based on prior empirical research on innovation performance in the management literature, the study includes a series of organizational-level control variables that have been found to affect the innovation performance of firms. To control for potential differences in productivity by country that could affect organizational innovative performance, the study operationalized a categorical variable that accounts for country of origin. Similarly, as the data is cross-sectional, the study controlled for industry related firm characteristics that influence innovation performance, such as a higher propensity to innovate in different industries. For this purpose, the study controls for possible differences in innovative performance owing to industry by including a categorical variable that distinguishes for whether a firm is or is not a member of that industry. These were identified based on the NACE classification codes.

Finally, the study controls for the potential effect of firm size as larger firms may be able to devote more resources to innovation by including a variable that accounts for the firm’s total number of employees. Besides the cost benefits associated with that, larger firms have better

113

financial capabilities and larger customer bases and are therefore more likely to derive high profits from innovation activities. Moreover, in previous studies firm size has been found to affect a firm’s search behaviour (Argyres & Silverman, 2004; Katila & Ahuja, 2002) as well as the propensity of firms to maintain inter-organizational linkages (Knoben & Oerlemans,

2010).

4.8. Data analysis

The process of data analysis for testing the proposed hypotheses of the study followed the quantitative techniques described below.

1) Hierarchical linear regression

In hierarchical linear regression, a series of linear regression analyses is performed to determine the extent to which a given predictor variable uniquely accounts for individual differences in the dependent variable. Hierarchical linear regression is used to determine the extent to which the influence of an exogenous variable on a dependent variable. Interaction effects are typically evaluated by testing the significance of a multiplicative term consisting of the product between two or more predictor variables controlling for associated lower order main effects (e.g., Cohen, 1978). When a significant interaction is found, it is common to further decompose this effect to better understand the structure of the relation (Aiken, West, & Reno, 1991). Specifically, hierarchical regression refers to the process of adding or removing predictor variables from the regression model in steps. In this study it was performed by examining the effects of AC variables on innovative performance using the function which was computed in STATA.

114

2) Tobit regression

Second, to account for the left-censored observations on innovation performance dependents that are measured as the proportion of turnover, the study performs a Tobit regression as a robustness check. The Tobit analysis procedure is based on the assumption of a normal distribution of the dependent variables. However, the measures for innovation performance in this study are all skewed to the left, which compromises the underlying assumption of normally distributed residuals in the Tobit model. When the dependent variable is outside of a known bound, but the exact value of the variable is unknown the sample is censored. Censored regression models commonly arise in cases where the variable of interest is only observable under certain conditions (in this study: innovative performance subject to R&D/innovation activities). The censored regression or Tobit model is appropriate when the dependent variable is censored at some upper or lower bound as an artefact of how the data are collected (Maddala, 1983; Tobin, 1958). Authors in previous studies (Laursen &

Salter, 2006; Leiponen & Helfat, 2010)often rely on limited dependent variable models, namely a Tobit type I regression (Amemiya, 1985; Tobin, 1958) because they recognize the non-negativity of sales with new products. Similarly, this study runs regression with innovation expenditures or sales with new products as dependent variable. For those companies that do not invest at all in R&D/innovation activities and therefore do not come up with new products/services, the study observes zeros for their innovative performance and categorizes these firms as left-censored.

3) Confirmatory factor analysis (CFA)

This study deals with routine-based AC construct that is already conceptually established in the theory, thus it applies confirmatory factor analysis (CFA) as the appropriate technique to confirm validity. CFA is a confirmatory technique driven by the theoretical

115

relationships among the observed and unobserved variables (Schreiber, Nora, Stage, Barlow, &

King, 2006). It is a special form of factor analysis, used to test whether measures of a construct are consistent with our understanding of the nature of that construct or factor. As such, the objective of confirmatory factor analysis is to test whether the data fit a hypothesized measurement model. This hypothesized model is based on theory or previous analytic research and it was first developed by Jöreskog and has built upon older methods (Jöreskog & Wold,

1982). Model fit measures are then be obtained to assess how well the proposed model captured the covariance between all the items or measures in the model (Kline, 2015).

CFA plays the role of validating and finding the reliability of measurement. It as an effective in examining convergent validity: an indicator converges if it has a factor loading value that is high and significant, and a standardized factor loading estimate is greater than 0.5. While

CFA assesses convergent validity effectively, the scholars argue it is not the best technique to assess discriminant validity (Luo, Kannan, & Ratchford, 2008; Noriega & Blair, 2008; Orth &

Malkewitz, 2008). Discriminant validity shows how much variance is in the indicators that are able to explain variance in the construct. Therefore, they call for calculating average variance extracted (AVE) which allows for the performance of the discriminant validity test as a more stringent measure of discriminant validity (Farrell & Rudd, 2009).

Measurement invariance is usually tested using multigroup CFA which examines the change in the goodness-of-fit index (GFI) when cross-group constraints are imposed on a measurement model. Because of the substantial influence of sample size, researchers have developed a variety of goodness-of-fit indices that they claim to be unaffected by sample size.

Though several varying opinions exist, Kline (2010) recommends reporting the Chi-squared test, the Root mean square error of approximation (RMSEA), the comparative fit index (CFI), and the standardised root mean square residual (SRMR). Hu and Bentler (1999) also provide rules of thumb for deciding which statistics to report and choosing cut-off values for declaring significance. When RMSEA values are close to .06 or below and CFI and TLI are close to .95

116

or greater, for example, the model may have a reasonably good fit. Therefore, it is recommended to report not only χ2 but RMSEA and CFI/TLI (Hu & Bentler, 1999). The summary of goodness of fit indices and recommended cut-offs are presented below (Table 1.5).

Table 1.5 Goodness of fit indices

Measure Name Description Cut-off for good fit

2 Χ Model Chi- Square Assess overall fit and the discrepancy between the sample p-value> 0.05 and fitted covariance matrices.

Sensitive to sample size. H0: The model fits perfectly.

(A)GFI (Adjusted) GFI is the proportion of variance accounted for by the GFI ≥ 0.95 Goodness of Fit estimated population covariance. Analogous to R2. AGFI AGFI ≥0.90 favors parsimony. (N)NFI (Non) Normed- Fit NFI of .95, indicates the model of interest improves the fit NFI ≥ 0.95 TLI Index by 95% relative to the null model. NNFI is preferable for smaller samples. Sometimes the NNFI is called the Tucker NNFI ≥ 0.95 Tucker Lewis index Lewis index (TLI)

CFI Comparative Fit A revised form of NFI. Not very sensitive to sample size. CFI ≥.90 Index Compares the fit of a target model to the fit of an independent, or null model. RMSEA Root Mean Square A parsimony-adjusted index. Values closer to 0 represent a RMSEA < Error of good fit. 0.08 Approximation (S)RMR (Standardized) Root The square-root of the difference between the residuals of SRMR <0.08 Mean Square the sample covariance matrix and the hypothesized model. Residual If items vary in range (i.e. some items are 1-5, others 1-7) then RMR is hard to interpret, better to use SRMR.

AVE (CFA Average Value The average of the R2 for items within a factor AVE >.5 only) Explained Source: Fit Statistics commonly reported for CFA and SEM (Parry, 2017)

Despite the progress in studying GOF indices there is a considerable controversy about the proposed cut-offs. The key reason for their popularity to assess the fit of models in covariance structure analyses is their elusive promise of golden rules - absolute cut-off values that allow researchers to decide whether a model adequately fits the data, with a wide

117

generalizability across different conditions and sample sizes (Hu & Bentler, 1999). However, some researchers do not believe that fit indices add value to the analysis (Barrett, 2007) and only the chi square should be interpreted. Others argue that cut-offs for fit indices can actually be misleading and subject to misuse (Hayduk, Cummings, Boadu, Pazderka-

Robinson, & Boulianne, 2007). Most scholars believe in the value of fit indices, but caution against strict reliance on cut-offs (Kenny, 2014) has been suggested. In particular, a caution around the idea of linking covariance fit to the worth of a model must be nuanced when brought into the SEM context, which represent specific theory-based causal connections between latent variables and between those latents and relevant indicator variables (Hayduk et al., 2007). Therefore, an overemphasis on determining the fit of a model to the data comes at risk of missing the opportunity to see that theoretical intent can be encapsulated in, and solid description provided by SEM for testing the underlying theory.

There are important logical problems underlying the rationale of a hypothesis-testing approach to setting cut-off values for fit indices. While Hu and Bentler (1999) offered cautions about the use of GOF indexes, subsequent research seems to have incorporated their guidelines without enough attention to the limitations noted by Hu and Bentler (1999).

Therefore, model fit can no longer be claimed via recourse to some published ‘threshold-level recommendation’, as there are “no recommendations anymore which stand serious scrutiny.’’

(Barrett, 2007, p. 821). For example, a combination of intuition and experience led researchers to suggest that RMSEA less than 0.05 is indicative of a “close fit,” and that values up to 0.08 represent reasonable errors of approximation. However, as emphasized by McDonald and

Marsh (1990), the traditional cut-off values (e.g., incremental fit indexes > .90) amount to little more than rules of thumb based largely on intuition and have little statistical justification

(McDonald & Marsh, 1990). Ultimately, data interpretations and their justification are a subjective undertaking that requires researchers to immerse themselves in their data (Jöreskog

& Sörbom, 2004).

118

The intent of the GOF indices has been to provide an alternative to traditional hypothesis-testing approaches based on traditional test statistics (e.g. maximum likelihood chi-square). However, Hu and Bentler (1999) specifically evaluated the GOF indices in relation to a traditional hypothesis-testing paradigm in which the maximum likelihood chi- square test outperformed all the indices. Indeed, “the performance of fit indices is complex and additional research with a wider class of models and conditions is needed” (Hu and

Bentler, p. 446). Iit is difficult to designate a specific cut-off value for each fit index because it does not work equally well with various types of fit indices, sample sizes, estimators, or distributions” (p. 449). They found that a designated cut-off value may not work equally well with various types of fit indexes, sample sizes, estimators, or distributions” (1999, p. 16). In conclusion, research should avoid the temptation to treat currently available “rules of thumb” as if they were golden rules. Instead, researchers must contend with rules of thumb that provide, at best, preliminary interpretations that must be pursued in relation to the specific details of their research (Hu & Bentler, 1999).

4) Post estimation marginal effects

Empirical research using limited dependent variable models often give little attention to the coefficient estimates, which cannot be interpreted as straightforwardly as OLS coefficients, and instead focus on predictions based on these coefficients. The reason is that the marginal effects and predicted quantities are the keys to understanding the relationships of interest in the population. For example, in a logit or probit model, either the marginal effect of a change in the independent variable of interest on the probability of success or the discrete difference in the probability of success due to a change in the independent variable of interest is more informative. Since these models are nonlinear and inherently interactive in all of the variables, the size of the effect of a change in the independent variable of interest

119

depends on the values of the other independent variables (Hanmer & Ozan Kalkan, 2013). As a result, the estimated effects, and thus the conclusions one can draw regarding the substantive significance of the variables of interest, are sensitive to which values for the other variables are chosen. According to Ai and Norton (2003), the magnitude of interaction effects in non-linear models cannot simply be read from the computed coefficient values.

In presenting predictions for a change in an independent variable of interest in limited dependent variable models, there are two general approaches for dealing with the other independent variables. The first involves creating an example case by selecting a set of specific values for the other variables and calculating the relevant predicted probabilities or marginal effect for that case. The second approach involves holding each of the other independent variables at the observed values for each case in the sample, calculating the relevant predicted probabilities or marginal effect for each case, and then averaging over all of the cases, referred as the “observed value” approach (Hanmer & Ozan Kalkan, 2013). Due to both theoretical and methodological reasons, researchers using limited dependent variable models should report predicted probabilities and marginal effects estimated via the observed-value approach.

120

Chapter 5 – Analysis and Results

This chapter presents the results of empirical analysis based on cross-sectional data in two parts in line with the conceptual framework. Commencing with the analysis of AC construct as a process, the first part of the chapter operationalizes and tests the sub-process stages/dimensions of AC and the corresponding routines. It deploys organizational routines as the conceptual foundation of differing AC dimensions that allows for empirical testing of such observable routines within firms (Lewin et al., 2011). The first part further advances research on AC by extending and empirically validating the conceptual distinction between the four dimensions through a number of steps, including confirmatory factor analysis (CFA).

The second part of the chapter is concerned with applying a series of empirical tests in the regression analysis to determine the relationship between independent variables (AC dimensions) and dependent variables (radical and incremental innovation performance).

5.1 Absorptive Capacity – operationalizing and pre-testing measures

The conceptualization of AC as constituted by routines aims at overcoming some limitations of the extant literature, which focus on indirect proxy measures e.g., R&D expenditures or patents (Lewin et al., 2011). Focusing attention on the routines is a step forward to the operationalization of the construct (Lewin et al., 2011) using valid latent measures for acquire, assimilate, transform, and exploit dimensions of the AC construct.

According to Lewin et al. (2011), the configurations of practiced routines that support these separate dimensions addresses earlier issues of construct abstraction and uniformity hampering its measurement and empirical progress (Lane et al., 2006). Furthermore, the type of capability that the firms deploy is influenced by how the firms develop routines and resources in response to environmental dynamism (Felin & Foss, 2011). As routines are

121

objectively observable at the firm level and have been theorized in the context of organizational learning, the study proceeds by validating designed-for-purpose scales for differing AC dimensions based on existing conceptual framework (Lewin et al., 2011).

DeVellis (2012, p. 12) defines a scale as consisting of “effect indicators’ – that is, items whose values are caused by an underlying construct (or latent variable).”

The measurement strategy in this study mirrors the methodology from similar studies in the field that sought to develop scales for analysing dynamic learning capabilities and second-order competences (Danneels, 2008; Verreynne, Hine, Coote, & Parker, 2016). The only exception is the fact that this study did not rely on the qualitative design to undertake considerable number of interviews with participants as this would deviate the quantitative research methodology and philosophical assumptions of the post-positivism. Instead, the study draws extensively and builds on exploratory research performed in the earlier study which identified the key underpinning AC routines (Lewin et al., 2011). Thus, drawing on routine-based model of AC (Lewin et al., 2011), which forms the basis of new appropriately derived items that capture elements specific to AC, the study operationalized valid latent measures for four separate dimensions of the construct. The items derived from this process are related to operationalization of AC through observable routine-based activities that theoretically and conceptually mapped to the separate latent dimensions of the AC construct.

In validating meaningful scales to test four separate AC dimensions the study applied the approach to develop scales through a number of steps (DeVellis, 2016; Verreynne et al., 2016) described in more detail below.

The study first articulated the measurement strategy using the existing literature to identify theoretically valid items. Based on the conceptual work on AC (Lewin et al., 2011) the configuration of the routines were identified as building blocks that underpin each AC dimension. A large pool of items was created based on the existing literature and mainly derived from the conceptual routine-based model (Lewin et al., 2011). In total 35 items were

122

generated to address the four AC dimensions and such multi-item scales have a potential to reduce measurement error and provide a robust measure of complex variables (Verreynne et al., 2016). Such items served as a starting point for the item pool, with the aim to focus on redundancy to maximise the content validity of the scale (Hogan, Soutar, McColl-Kennedy, &

Sweeney, 2011). Second, the expert opinion was sought from academic experts (n=6) to help refine the scale and eliminate any of the researcher’s perceptual distortions. Therefore, for each of the AC dimensions, the study created a consolidated pool of items with the assistance of a panel of experts by drawing on existing measures and pre-tested measures of AC.

Experts were provided with AC conceptual framework and a list of items pertaining to each

AC dimension. The study applied the established rating method (Zaichkowsky, 1985) which suggests retaining items for which at least 80% of experts rated as at least somewhat representative. The feedback of experts resulted in refining wording of some items, as well as the elimination of 9 items, leaving 26 items for further analysis.

Third, to ensure the external validity and relevance of items derived from conceptual framework, AC measures were pre-tested with innovation practitioners (n=6) from different firms/industries in a trial questionnaire, which narrowed down the items for further analysis.

The innovation practitioners were asked to complete the questionnaire and indicate any ambiguity regarding the phrasing of the items. Each of the questions was structured in such a manner that innovation practitioners could express the extent to which they agreed with positively phrased statement. Such feedback enabled the researcher to gain an early insight into innovation practice in order to bridge it with the theoretical framework, thus helping to reveal and understand the key routines that underpin AC dimensions. The results were used to adapt the instrument and thus establish a better understanding of the survey and increase the content validity. This process also led to refinement of item wording and substantial reduction of the number of items (from 26 to remaining 16, or 4 per dimension). Based on the operationalization of AC, the following routines (items) (Table 1.6) were used as the

123

relevant measures for further testing through Confirmatory Factor Analysis (CFA), which proved their reliability and validity.

Table 1.6 Absorptive capacity dimensions and items (developed for the purpose of the study)

Acquire Our firm uses formal or informal gatekeepers to monitor the external environment. Our firm uses formal environmental probing and scanning mechanisms. Our firm accesses knowledge through market research, end-user surveys, and other forms of primary research. Our firm accesses knowledge through patent records, trade magazines, and other forms of secondary research. Assimilate Our firm uses inter-firm communities and steering groups to analyse and interpret knowledge. Our firm seeks to understand new knowledge through trial and error learning Our firm keeps pace with external partners and other external constituents to interpret the environment. Our firm uses formal training programs to understand changing market demands. Transform Our firm communicates and distributes external information and knowledge to internal stakeholders and departments. Our firm uses a formal IT system or knowledge system to codify, store, and manage knowledge. Our firm engages in knowledge integration practices across divisions and departments. Our firm uses cross-functional teams as a means to pivot on existing and new knowledge. Exploit Our firm often reflects on and updates products, technologies, and processes as a result of external search. Our firm is good at applying new knowledge and insights to guide decisions. Our firm is good at determining and allocating resources to commercial projects. Our firm utilizes knowledge to develop and implement new technologies and products to the market.

5.2 Confirmatory Factor Analysis (CFA)

In a study of potential and realized absorptive capacity, Jansen, Van den Bosch, and

Volberda (2005) show that acquisition, assimilation, transformation, and exploitation represent four empirically distinct dimensions of absorptive capacity. Confirmatory factor analysis demonstrates that a four-factor model in which acquisition, assimilation, transformation, and exploitation are separate dimensions is clearly superior to a two-factor model in which acquisition and assimilation are combined into a potential absorptive capacity factor and transformation and exploitation are combined into realized absorptive capacity factor. Based on this, the study applies four factor scale which confirms the above argument about its superiority. First, the four-item scale for acquisition addresses a firm’s efforts to acquire new knowledge from external sources. Second, the four-item scale for assimilation

124

gauges a firm’s proficiency in analysing and understanding new external information. Then, the four-item scale for transformation reflects the extent to which a firm is able to combine existing knowledge with new information and interpret existing knowledge in a new way.

Finally, the four-item scale for exploitation assesses a firm’s proficiency in exploiting new knowledge and applying technologies in new products.

Absorptive capacity was therefore operationalized with the four proposed dimensions

(Zahra & George, 2002) relying on existing conceptual items and pretested measures for these dimensions (Jansen et al., 2005b) and, in line with conceptual discussions in AC research

(Todorova & Durisin, 2007; Zahra & George, 2002). For operationalizing and empirically testing AC construct, this study adopts items that are theoretically and conceptually verified in published academic studies for the separate dimensions of AC. After scales were presented in a trial questionnaire and pretested with six academic experts and six practitioners, relevant items were subsequently included for further analysis. The study conducts confirmatory factor analysis (CFA) of the items pertaining to four AC dimensions of in order to check for independence of these constructs and reports goodness of fit indices. All the items for each dimension of AC separately were measured on five-point Likert-type scales, with scale scores calculated as the summation of individual item scores. Each item was allowed to load only on the factor for which it was a proposed indicator.

An important indicator related to psychometric properties (Farrell & Rudd, 2009) is the construct reliability in order to determine how well the data actually measures the construct of interest. Therefore, the study provides Cronbach’s alpha to assess construct reliability for each

AC dimension. All values resulted higher than 0.6, which as acceptable (Hair, 2009), so it can be concluded that the data measures the constructs of interest quite well. In addition, path diagrams (Byrne, 2013) in Stata were used as the graphic representation of the models to check how well they fit the observed data. Regarding convergent validity, the standardized factor loadings were all significant and above the minimum value of 0.50, and the factors’ average

125

variances extracted all exceeded the recommended cut-off of 0.50 (Hair et al., 2010). Since the correlations among the constructs are low to moderate discriminant validity did not seem to be a problem. Moreover, each factor’s average variance extracted was larger than the squared value of the correlations of this factor to other factors (Fornell & Larcker, 1981).

To assess the validity of the four-dimension AC construct, the study performs a CFA

(Table 1.7) with a robust maximum-likelihood estimator in Stata. The four-item scale for acquisition encompasses a variety of routines deployed by the firm that assist in the identification and acquisition of new knowledge from the external environment and broadly embodies activities that facilitate external knowledge search, such as market research, environmental probing and scanning, and patent searches. The reliability of this scale is  =

0.698 and the average variance explained (AVE) = 0.668. The four-item scale for the assimilate dimension of AC encompasses the routines deployed by the firm that assist in the analysis, interpretation, and understanding of externally acquired knowledge. The reliability of this scale is  = 0.653 and the AVE = 0.508. The four-item scale for the transform dimension of AC encompasses the routines deployed by the firm that enable the extension or renewal of existing knowledge through the combination of externally acquired and assimilated knowledge with existing stocks of knowledge. This involves routines relating to the storing and retrieval of knowledge and enablement of variation from current understandings such as engaging in cross-functional communication and brainstorming. The reliability of this scale is

 = 0.731 and the AVE = 0.675. The four-item scale for the exploit dimension of AC encompasses the routines deployed by the firm for the commercial application of knowledge and its incorporation into operations for exploitation. The reliability of this scale is  = 0.667 and the AVE = 0.741.

All standardized factor loadings were significant and ranged between 0.50 and 0.78 and the factors’ AVE ranged between 0.51 and 0.74, with between dimension correlations being

126

low to moderate and well below 0.90 (Kline, 2015). Furthermore, the comparative fit index

(CFI), Tucker-Lewis index (TLI), root mean square error of approximation (RMSEA), and standardized root mean square residuals (SRMR) model fit indices were: 0.856, 0.824, 0.079, and 0.982, respectively (2 = 246.584, df = 98, p < 0.00), which indicated a decent degree of fit. Together, these fit indices and factor loadings evidence convergent and discriminant validity (Hair Jr, Black, Babin, Anderson, & Tatham, 2010). A reliability analysis was also manually performed which showed that the deletion of any single item in the four-item scales would decrease the . The analysis of AC construct advances our understanding of four separate capabilities through conceptually identifying and empirically examining common routines of acquisition assimilation, transformation and exploitation. According to the CFA undertaken, the four dimensions underlying absorptive capacity are not only theoretically, but also empirically distinguishable. Though the specified capabilities may differ in detail, the study draws attention to important routine-based commonalities as identified and examined in the study (Eisenhardt & Martin, 2000).

127

Table 1.7 Results of Confirmatory Factor Analysis (developed for the purpose of the study)

b/(SE) z/(P) Alpha () AVE Acquire Our firm uses formal or informal gatekeepers to 0.771 (0.047) 16.53 (0.000) monitor the external environment. Our firm uses formal environmental probing 0.647 (0.051) 12.80 (0.000) and scanning mechanisms. Our firm accesses knowledge through market 0.517 (0.060) 8.55 (0.000) 0.698 0.668 research, end-user surveys, and other forms of primary research. Our firm accesses knowledge through patent 0.522 (0.058) 9.04 (0.000) records, trade magazines, and other forms of secondary research. Assimilate Our firm uses inter-firm communities and 0.527 (0.062) 8.48 (0.000) steering groups to analyse and interpret knowledge. Our firm seeks to understand new knowledge 0.500 (0.064) 7.77 (0.000) through trial and error learning (learning from good and bad experience). 0.653 0.508 Our firm keeps pace with external partners and 0.615 (0.052) 11.83 (0.000) other external constituents to interpret the environment. Our firm uses formal training programs to 0.679 (0.054) 12.66 (0.000) understanding changing market demands. Transform Our firm communicates and distributes external 0.641 (0.048) 13.24 (0.000) information and knowledge to internal stakeholders and departments. Our firm uses a formal IT system or knowledge 0.591 (0.052) 11.42 (0.000) system to codify, store, and manage 0.731 0.675 knowledge. Our firm engages in knowledge integration 0.590 (0.051) 11.49 (0.000) practices across divisions and departments. Our firm uses cross-functional teams as a means 0.726 (0.043) 17.00 (0.000) to pivot on existing and new knowledge. Exploit Our firm often reflects on and updates products, 0.576 (0.054) 10.76 (0.000) technologies, and processes as a result of external search. Our firm is good at applying new knowledge and 0.701 (0.049) 14.31 (0.000) insights to guide decisions. Our firm is good at determining and allocating 0.570 (0.054) 10.51 (0.000) resources to commercial projects. Our firm utilizes knowledge to develop and 0.508 (0.059) 8.60 (0.000) implement new technologies and products to the market. Note: b = factor loading with standard errors reported in parentheses (SE). z = inferential statistic with exact p- values reported in parentheses (P). AVE = factor’s average variance extracted.

128

Comparison of AC models - chi-square difference test

The fit indices of our four-factor model indicate partially satisfactory benchmark for parsimonious, absolute and incremental fit. Following its advantage, the study applies chi- square difference test to compare the fit of the four-factor model to the fit of other established two-factor (potential and realized) and single factor AC models. First, the single factor and two-factor AC models were compared for which LR chi-square = 49.03, with df =

1 and associated p < 0.001, indicating that the two-factor model fits significantly better.

Therefore, the assumption that the four underlying dimensions of absorptive capacity converged best on one common factor was unambiguously rejected. Second, two-factor model was compared to the four-factor model and the result LR chi-square = 59.48, df = 5, p

< 0.001 shows that the four-factor model fits the data significantly better than the plausible rival two-factor model. Accordingly, the four dimensions underlying absorptive capacity are not only theoretically, but also empirically distinguishable.

5.3 Part II – Data analysis

The empirical analysis of the second part of the chapter is concerned with regressing four AC dimensions on different types of organizational learning outcomes as measures of incremental and radical innovative performance. This part of empirical analysis encompasses relevant steps in applying quantitative techniques, presents the results within the context of the provided dataset generated by the survey. As a part of the analysis, data handling and analysis steps are explained. Finally, the results of testing two hypotheses representing exploration-based and exploitation-based views of absorptive capacity are presented.

129

5.3.1 Data cleaning

From the original sampling frame of 715 firms, a total of 309 firms participated in the survey. During the data cleaning phase, however, 78 observations were subsequently removed owing to incomplete or erroneous entries in core variables of interest, which resulted in a final database of 241 firms. The fact that a number of firms did not reply to the survey leaves questions about the representativeness of sample for the original population and a possible non-response bias. This was examined by a comparison of variables which are known for the base population and the sample as well as a comparison of variables between late and early respondents, respectively. After generating dummy variables for sample versus non-sample and late versus early respondent firms, the study compared firms by regressing the dummy variables on certain demographic and model variables. Firms in the sample and firms from the initial population were compared with regard to their age, and turnover. Additionally, the study examined differences of late and early responding firms with regard to certain model variables. By and large, no appreciable differences were detectable. The comparison did not reveal any significant differences (p<.05), thereby there are no major concerns about the representativeness and nonresponse bias. In conclusion, tests performed to account for potential nonresponse and late-response bias showed no significant results.

Having gone through the raw data, it became apparent that the level of measurement of variables in the analysis varied (nominal vs. scale). In order to produce the mean and standard deviation of variables, the STATA Descriptives option was performed. This step enabled the researcher to get the useful summary of the data and helped in reducing problems associated with multicollinearity prior to analysis (Tabachnick & Fidell, 2001). In addition, no particular missing values have been found (except minor ones for the control variables), so the missing data condition (under 10%) was met (Cohen, Cohen, West, & Aiken, 2013). As the dataset might be affected by outliers (number of extremes that are deemed to be out-of-

130

range) (Tabachnick & Fidell, 2001) the regression technique can be sensitive. For this purpose, descriptive statistics and Nick Cox’s extremes command in STATA was used to identify the cases with the most extreme high and low values. In addition, the evaluation of other relevant assumptions such as normality, homoscedasticity and homogeneity of residual variances (Tabachnick & Fidell, 2001) was also performed.

The importance of normal distribution is an underlying assumption of many statistical tests and once it is violated the interpretation and inferences may not be reliable or valid

(Razali & Wah, 2011). The three common procedures in assessing normal distribution are: graphical methods (histograms, boxplots), numerical methods (skewness and kurtosis indices) and formal normality tests (Razali & Wah, 2011). For the purpose of the study Shapiro-Wilk test (Field, 2013; Shapiro & Wilk, 1965) was used, as it is able to detect departures from normality due to either skewness or kurtosis, and has become the preferred test because of its good power properties. The results of the normality test for AC dimensions indicate that the p =.001 < 0.05 (not statistically significant), so it can be concluded that the AC variables are normally distributed. However, in case of the results in examining dependent variable

(innovative performance) normality assumption is violated as variables are skewed.

5.3.2 Hierarchical linear regression and Tobit regression results

The study performs a hierarchical linear regression to examine the main effects of each

AC dimension and their interactions at external, transformative, and internal stages on incremental and radical innovation performance dependents. Therefore, the study regresses the two dependent variables of innovation performance on four AC dimensions (acquire, assimilate, transform and exploit) as well as their intersections, controlling for country, industry and size. The reported coefficients do not suggest multicollinearity as an issue and a check of variance-inflation-factors performed after the main hierarchical linear regression

131

confirm this with all values well under the threshold of 10. Descriptive statistics and pairwise correlations are presented in Table 1.8.

As the assumption of normality of residuals in the standard Tobit model is not satisfied, alternative specifications of the Tobit model have been formulated that account for departures of the distributions from normality (Greene, 2000). The variables reflecting the innovative performance of firms are highly skewed and, accordingly, the pattern observed in the empirical distribution is better represented by lognormal distributions. Other studies, facing similar problems in terms of similar characteristics of skewness and departure from normality, have proposed a log-transformation of the Tobit model with a multiplicative exponential error term (Filippucci, Drudi, & Papalia, 1996; Papalia & Di Iorio, 2001). Data is available on AC dimensions (independent variable) and our model estimates the relationship between AC dimensions and innovative performance. However, such estimates undertaken using OLS regression might be biased by the fact that for firms that decide not to innovate

(left-censored) it is not possible to observe the innovative performance they would have reached had they engaged in R&D/innovation activities.

The scores of innovation performance measures, which reflect the percentage of sales from products with different degrees of novelty, range between 0 and 100. Thus, the dependent variables are double censored (having an upper censoring point at the score 100 and a lower censoring point at scores that equal 0; the former is also referred to as “left censoring” while the latter is referred to as “right censoring”). Analysing such data using ordinary least squares (OLS) regression might lead to biased coefficient estimates (Tobin,

1958). The most appropriate method then to adjust the coefficient estimates and account for the fact that the dependent variable is censored between 0 and 100 is a Tobit analysis

(Greene, 2003). To account for this departure from the normality assumption, a logarithmic transformation of dependent variables was performed assuming a lognormal distribution for

132

the residuals of the Tobit model. This model introduces two latent variables, as a logarithmic transformation of the observed measures where INN* = ln (1 + INN).

1) Log transformed variable of radical innovation performance that represents

percentage of total sales derived from new or products that are considered new to the

market

2) Log transformed variable of incremental innovation performance that represents

percentage of total sales derived from improved products that are considered new to

the firm.

It is then assumed that the latent variable of innovative performance of a firm i is a function of a number of explicative variables. When observations are selected so that they are not independent of the outcome variables in a study, sample selection leads to biased inferences.

In order to address such bias, the study performs additional robustness checks.

133

Table 1.8 Descriptive Statistics and Correlation Coefficients

Mean SD 1 2 3 4 5 6 7 8 1. Incremental innovation performance 10.28 8.28 2. Radical innovation performance 8.92 8.06 0.77 3. Country 9.24 13.33 -0.05 -0.02 4. Industry 10.51 5.62 -0.10 -0.05 -0.02 5. Size 2.07 1.15 0.30 0.17 -0.05 -0.25 6. Acquire 7.49 3.26 0.16 0.19 0.02 -0.15 0.29 7. Assimilate 7.24 2.83 0.18 0.27 -0.02 0.03 0.19 0.54 8. Transform 9.16 3.14 0.24 0.27 -0.01 -0.12 0.29 0.42 0.51 9. Exploit 9.04 3.20 0.14 0.16 0.07 -0.09 0.15 0.45 0.42 0.55

134

Based on further analysis the results of the main hierarchical linear and Tobit regression analysis for the exploitation-based view of AC are presented in Table 1.9. Model 1 is the baseline model that includes only control variables and separate dimensions of AC.

Synonymous with the hierarchical approach, Models 2 to 4 sequentially and separately include the interaction terms between acquire and assimilate, assimilate and transform, and transform and exploit dimensions, characteristic of external, transformative, and internal stages of AC respectively. The study theorized in H1 that only the internal stage of AC embodied by the interaction between transform and exploit routines would be positively related to incremental innovation performance. The results provide strong support for this assertion, as both the coefficients for external AC (b = 0.02, p = 0.767) in Model 2 and transformative AC (b =

0.06, p = 0.261) in Model 3 are non-significant; while the coefficient for internal AC (b =

0.09, p = 0.038) in Model 4 is positive and significant. Furthermore, the analysis of model fit through each increment from the baseline model (Model 1) indicates that only Model 4 significantly improves fit (2 = 4.47, df = 1, p < 0.035), leading us to reject all other specifications. In Model 5, Model 4 is re-estimated using a Tobit estimation as a robustness check to account for the potential effects of left-censoring in the dependent variable of incremental innovation performance. The result confirms the main findings and in fact demonstrates a stronger effect of internal AC (b = 0.13, p = 0.038) on driving incremental innovation performance.

135

Table 1.9 Regression Analysis for Exploitation-Based View of Absorptive Capacity

Model 1 Model 2 Model 3 Model 4 Model 5 b/(SE) t/(P) b/(SE) t/(P) b/(SE) t/(P) b/(SE) t/(P) b/(SE) t/(P) Country -0.02 -0.52 -0.02 -0.55 -0.03 -0.66 -0.02 -0.57 -0.04 -0.67 (0.04) (0.603) (0.04) (0.582) (0.04) (0.508) (0.04) (0.570) (0.05) (0.503) Industry -0.03 -0.35 -0.04 -0.38 -0.04 -0.46 -0.04 -0.45 -0.08 -0.65 (0.10) (0.725) (0.10) (0.704) (0.10) (0.645) (0.09) (0.653) (0.13) (0.518) Size 1.77 (0.48) 3.66 1.77 (0.49) 3.65 1.75 (0.48) 3.61 1.86 (0.48) 3.86 2.38 (0.68) 3.49 (0.000) (0.000) (0.000) (0.000) (0.001) Acquire -0.03 -0.16 -0.14 -0.34 -0.05 -0.25 -0.08 -0.42 -0.29 -0.69 (0.20) (0.876) (0.42) (0.74) (0.20) (0.806) (0.20) (0.678) (0.67) (0.494) Assimilate 0.19 (0.24) 0.80 0.08 (0.44) 0.18 -0.34 -0.64 0.19 (0.24) 0.79 0.25 (0.33) 0.75 (0.423) (0.855) (0.53) (0.520) (0.433) (0.452) Transform 0.37 (0.21) 1.73 0.37 (0.21) 1.73 (0.09) 0.00 (0.39) 0.00 -0.37 -0.89 -0.55 -0.96 (0.086) (0.996) (0.41) (0.374) (0.57) (0.340) Exploit 0.37 (0.21) 0.11 0.02 (0.20) 0.10 0.01 (0.20) 0.04 -0.76 -1.79 -1.04 -1.76 (0.910) (0.922) (0.971) (0.43) (0.075) (0.59) (0.080) Acquire  assimilate 0.02 (0.05) 0.30 (0.767) Assimilate  transform 0.06 (0.05) 1.13 (0.261) Transform  exploit 0.09 (0.04) 2.08 0.13 (0.06) 2.09 (0.038) (0.038) Constant 2.45 (2.32) 1.05 3.22 (3.50) 0.92 5.90 (3.84) 1.54 8.68 (3.78) 2.30 7.74 (5.30) 1.46 (0.293) (0.358) (0.126) (0.022) (0.145) Chi-square 4.65 (0.000) 4.06 (0.000) 4.23 (0.000) 4.67 (0.000) 29.15 (0.000) N. 241 241 241 241 241 Note: b = regression coefficient with standard errors reported in parentheses (SE). t = inferential statistic with exact p-values reported in parentheses (P). Models 1-4 report main hierarchical linear regression results and Model 5 reports Tobit regression results to account for censoring (74 out of 241 observations are left-censored)

136

5.3.3 Average marginal effects

Adjusted predictions and marginal effects can make the results from many analyses much more intuitive and easier to interpret. Marginal effects are popular means by which the effects of variables can be made more intuitively meaningful. Older Stata commands generally default to using the means for variables whose values have not been otherwise specified, that is, they estimate marginal effects at the means (MEMs). However, at least two other approaches are also possible with the margins command: average marginal effects (AMEs) and marginal effects at representative values (MERs). Rather than use the means when computing predicted values, scholars argue it is best to use the actual observed values for the variables whose values are not otherwise fixed (Williams, 2012). As Cameron and Trivedi

(2010, p. 343) note, “a marginal effect (ME), or partial effect, most often measures the effect on the conditional mean of y of a change in one of the regressors, say, xj. In the linear regression model, the ME equals the relevant slope coefficient, greatly simplifying analysis.

For nonlinear models, this is no longer the case, leading to remarkably many different methods for calculating MEs.” Most scholars seem to prefer AMEs over MEMs (Bartus,

2005; Cameron & Trivedi, 2010), though the output from the margins command can be difficult to read due to space constraints. Therefore, the margins plot command in Stata makes it easy to create a visual display of results (Williams, 2012).

In order to unpack the hypothesized interaction effects at different stages of AC, the study performs a post estimation marginal effects analysis on the more stringent Tobit specification of the main analysis. Doing so is consistent with recognizing that the core of the scientific endeavour is to “infer beyond the immediate data to something broader that is not directly observed” (King, Keohane, and Verba 1994, 8). The observed-value approach helps associate the theoretical framework and data-collection activities with the properties of limited dependent variable models (King, Keohane, & Verba, 1994).

137

Therefore, to aid interpretation of the interaction effects of internal AC on incremental innovation performance, the study analyses and graphs the average marginal effects (AME) of a change in transform, between low (-2 SD) and high (+2 SD) levels, at low (-2 SD) and high

(+2 SD) levels of exploit, while holding all other variables constant, and vice-versa. Thus, the

AME illustrate how a change in association between transform and exploit at low and high levels of each form of routine, all else being equal, influence incremental innovation performance. Figure 1.15 shows that high levels of exploit are critical for driving incremental innovation performance gains at mean to high levels of transform, while low levels of exploit induce slightly negative effects on incremental innovation performance beyond low levels of transform. A similar pattern is observed in Figure 1.16, albeit more pronounced, as low levels of transform result in increasingly negative effects on incremental innovation performance as exploit increases; while the opposite is true for high-levels of transform. These results suggest that the cycling through transform and exploit routines during the internal stage of AC is critical for driving incremental innovation performance and provide strong support for H1.

Figure 1.15 Average Marginal Effects for Exploitation-Based View (developed for this study)

AME of Transform Conditional on Exploit

138

Figure 1.16 Average Marginal Effects for Exploitation-Based View (developed for this study)

AME of Exploit Conditional on Transform

Table 1.10 presents the results of the main hierarchical linear and Tobit regression analysis for the exploration-based view of AC. Model 1 is the baseline model that includes only control variables and separate dimensions of AC. Again, Models 2 to 4 sequentially and separately include the interaction terms for each stage of AC. According to H2a that the transformative stage embodied by the interaction between assimilate and transform routines would be positively related to radical innovation performance. The results provide support for this assertion, as the coefficient for external AC (b = 0.07, p = 0.150) in Model 2 is non- significant and the coefficient for internal AC (b = 0.08, p = 0.078) in Model 4 is only marginally significant at the 10 percent level; while the coefficient for transformative AC (b =

0.12, p = 0.020) in Model 3 is strongly positive and significant at the 5 percent level.

Furthermore, for jointly considered transformative and internal AC in Model 5, it is only the coefficient for transformative AC that remains significant (b = 0.10, p = 0.078). An analysis of model fit from the baseline also indicates that Model 3 (2 = 5.64, df = 1, p < 0.016),

139

which only includes the interaction between assimilate and transform, provides a superior fit to all others. As a final robustness check, Model 3 was re-estimated using a Tobit estimation, which again confirms results and shows a stronger positive effect of transformative AC (b =

0.19, p = 0.018) on radical innovation performance.

140

Table 1.10 Regression Analysis for Exploration-Based View of Absorptive Capacity

Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 b/(SE) t/(P) b/(SE) t/(P) b/(SE) t/(P) b/(SE) t/(P) b/(SE) t/(P) b/(SE) t/(P) Country -0.01 -0.21 -0.01 -0.38 -0.02 -0.51 -0.01 -0.24 -0.02 -0.48 -0.03 -0.46 (0.04) (0.837) (0.04) (0.708) (0.04) (0.611) (0.04) (0.807) (0.04) (0.633) (0.06) (0.644) Industry -0.02 -0.19 -0.03 -0.35 -0.04 -0.43 -0.03 -0.28 -0.04 -0.43 -0.08 -0.59 (0.09) (0.846) (0.09) (0.729) (0.09) (0.670) (0.09) (0.783) (0.09) (0.666) (0.14) (0.557) Size 0.64 1.34 0.63 1.33 0.59 1.26 0.72 1.50 0.64 1.35 0.50 0.68 (0.48) (0.181) (0.47) (0.184) (0.47) (0.210) (0.48) (0.134) (0.48) (0.177) (0.73) (0.498) Acquire 0.02 0.09 -0.50 -1.22 -0.02 -0.10 -0.03 -0.13 -0.04 -0.19 -0.10 -0.31 (0.20) (0.930) (0.41) (0.222) (0.20) (0.923) (0.20) (0.895) (0.20) (0.852) (0.30) (0.753) Assimilate 0.51 2.19 -0.01 -0.03 -0.56 -1.09 0.51 2.18 -0.38 -0.69 -0.86 -1.10 (0.23) (0.029) (0.43) (0.977) (0.51) (0.275) (0.23) (0.030) (0.55) (0.493) (0.77) (0.271) Transform 0.42 2.01 0.43 2.07 -0.32 -0.85 -0.19 -0.48 -0.53 -1.19 -0.56 -0.96 (0.21) (0.046) (0.21) (0.040) (0.38) (0.398) (0.40) (0.634) (0.45) (0.235) (0.58) (0.337) Exploit -0.05 -0.26 -0.06 -0.33 -0.08 -0.42 -0.71 -1.69 -0.44 -0.98 -0.11 -0.37 (0.20) (0.798) (0.20) (0.742) (0.19) (0.677) (0.42) (0.093) (0.44) (0.326) (0.30) (0.712) Acquire  assimilate 0.07 1.44 (0.05) (0.150) Assimilate  transform 0.12 2.34 0.10 1.77 0.19 2.38 (0.05) (0.020) (0.06) (0.078) (0.08) (0.018) Transform  exploit 0.08 1.77 0.04 0.89 (0.04) (0.078) (0.05) (0.374) Constant 0.64 0.28 4.33 1.27 7.64 2.04 5.86 1.58 9.31 2.22 6.16 1.08 (2.28) (0.778) (3.42) (0.207) (3.74) (0.042) (3.72) (0.116) (4.18) (0.027) (5.71) (0.282) Chi-square 3.98 (0.000) 3.76 (0.000) 4.24 (0.000) 3.90 (0.000) 3.85 (0.000) 26.78 (0.001) N. 241 241 241 241 241 241 Note: b = regression coefficient with standard errors reported in parentheses (SE). t = inferential statistic with exact p-values reported in parentheses (P). Models 1-5 report main hierarchical linear regression results and Model 6 reports Tobit regression results to account for censoring (90 out of 241 observations are left-censored).

141

Again, to aid interpretation of the interaction effects of transformative AC on radical innovation performance, the study analyses and graphs the AME of a change in assimilation, between low (-2 SD) and high (+2 SD) levels, at low (-2 SD) and high (+2 SD) levels of exploit, while holding all other variables constant, and vice-versa. Thus, the AME illustrate how a change in association between assimilate and transform at low and high levels, all else being equal, influence radical innovation performance. Figure 1.17 shows that high levels of transform are always critical for driving radical innovation performance gains at both low and high levels of assimilate, while low levels of transform induce negative effects. The interpretation is that firms with low levels of transform capability are unable to internalize or benefit from the distant knowledge that is externally assimilated for the purposes of exploratory learning. A similar pattern is observed in Figure 1.18, as low levels of assimilate have no effect on radical innovation performance at both low and high levels of transform; while the opposite is true for high-levels of assimilate. The interpretation is that low levels of external knowledge assimilation nullify the firms’ capacity to transform such knowledge for radical innovation. These results suggest that the cycling through assimilate and transform routines during the transformative stage of AC is critical for driving radical innovation performance and provide strong support for H2a.

142

Figure 1.17 Average Marginal Effects for Exploration-Based View (developed for this study)

AME of Assimilate Conditional on Transform

Figure 1.18 Average Marginal Effects for Exploration-Based View (developed for this study)

AME of Transform Conditional on Assimilate

143

In H2b the study theorized that high levels of exploit in the internal stage of AC would negate any performance benefits attributed to transformative AC. To test this hypothesis, the study estimated a model that compares sample firms with low to moderate levels of exploit stage of internal AC that fall below the 75th percentile, against sample firms with high levels of internal AC above the 75th percentile. The results are provided in Table 1.11. Model 1 shows a positive, marginally significant effect of transformative AC (b = 0.13, p = 0.061) on radical innovation performance at low to moderate levels of exploit stage of internal AC. In Model 2, however, a negative effect is observed at high levels of exploit stage of internal AC, albeit non-significant (b = -0.153, p = 0.682). Taken together these results provide partial support for H2b. Again, robustness check was performed using a Tobit estimation in Models 3 and 4 which confirm this result.

144

Table 1.11 Sample Split Regression Analysis for Exploration-Based View of Absorptive Capacity

Model 1 Model 2 Model 3 Model 4 b/(SE) t/(P) b/(SE) t/(P) b/(SE) t/(P) b/(SE) t/(P) Country 0.00 (0.05) 0.05 (0.959) -0.06 (0.08) -0.79 (0.432) 0.00 (0.07) 0.04 (0.965) -0.06 (0.10) -0.57 (0.572) Industry -0.04 (0.11) -0.41 (0.681) 0.03 (0.23) 0.14 (0.885) -0.09 (0.17) -0.56 (0.578) 0.03 (0.29) 0.09 (0.929) Size 0.62 (0.52) 1.19 (0.236) 1.29 (1.34) 0.97 (0.338) 0.55 (0.82) 0.66 (0.507) 1.07 (1.76) 0.61 (0.545) Acquire 0.01 (0.22) -0.04 (0.970) 0.24 (0.50) 0.49 (0.627) -0.08 (0.35) -0.23 (0.817) 0.26 (0.67) 0.39 (0.700) Assimilate -0.51 (0.58) -0.88 (0.379) 2.27 (4.45) 0.51 (0.613) -0.90 (0.90) -1.00 (0.320) 4.69 (6.28) 0.75 (0.459) Transform -0.46 (0.45) -1.00 (0.317) 3.14 (3.28) 0.96 (0.343) -0.77 (0.71) -1.08 (0.282) 5.24 (4.64) 1.13 (0.264) Exploit -0.11 (0.23) -0.48 (0.634) 0.15 (0.73) 0.20 (0.843) -0.07 (0.36) -0.21 (0.836) 0.07 (0.97) 0.08 (0.940) Assimilate  transform 0.13 (0.07) 1.88 (0.061) -0.153 (0.37) -0.41 (0.682) 0.21 (0.11) 1.97 (0.051) -0.31 (0.52) -0.61 (0.545) Constant 8.00 (4.00) 2.00 (0.047) -37.10 (41.63) -0.89 (0.377) 6.41 (6.29) 1.02 (0.310) -67.38 (59.03) -1.14 (0.259) Chi-square 1.95 (0.055) 1.24 (0.294) 12.53 (0.129) 10.44 (0.235) N. 180 61 180 61 Note: b = regression coefficient with standard errors reported in parentheses (SE). t = inferential statistic with exact p-values reported in parentheses (P). Models 1 and 2 report linear regression results for sample firms that fall below (Model 1) and above (Model 2) the 75th percentile of internal AC. Models 3 and 4 report the Tobit regression results to account for censoring.

145

Chapter 6 – Findings and discussion

This chapter discusses findings of the study, including theoretical implications and contributions of this research based on obtained results that provide answers to the research questions. Then, it also provides managerial implications for innovation practitioners and concludes with a discussion of limitations and opportunities for further research.

6.1 Discussion of findings

Previous literature suggested that due to the different nature of two types of organizational learning, the development of the organizational capabilities and managerial challenges of exploration and exploitation vary substantially (Lichtenthaler & Lichtenthaler,

2009). The findings of this study further confirm that distinct capabilities drive two types of organizational learning through examining dual nature AC. At any point in time, firms may adopt a mix of learning behaviours constituted by a semi-automatic accumulation of experience and by deliberate investments in knowledge articulation and codification activities.

The application of the routines in diverse contexts generates new information as to the performance implications of the routines employed (Zollo & Winter, 2002). This study sheds light on the routines that underpin different types of AC for radical innovators in high- velocity environments versus incremental innovators in low-velocity markets. Therefore, it answers the research question related to the underpinning constituents of AC dimensions at the firm level by empirically validating the routine-based model (Lewin et al., 2011).

Firms learn systematic ways to shape their routines by adopting an opportune mix of behavioural and cognitive processes, and by learning how to articulate and codify knowledge.

Exploration activities, characteristic of the variation and selection phase, involve high levels of

146

cognitive activities. The “learning investment” is likely to be higher when the organization (or the relevant unit) relies on knowledge articulation processes to attempt to master or improve a certain activity (Zollo & Winter, 2002). Exploitation by contrast is viewed as primarily driven by the diffusion and enactment of the novel insights, requiring relatively lower levels of deliberate cognitive effort. The level of investment in developing capabilities will be lower as firms count on the experience accumulation process, as the learning happens in an essentially semi-automatic fashion on the basis of the adaptations that individuals enact in reaction to unsatisfactory performance. Such arguments are clearly in line with the findings of this study through the lens of “exploitation view of absorptive capacity” and “exploration view of absorptive capacity”. Therefore, the study confirms that exploration focused firms are most likely to deploy capabilities and routines by combining diverse knowledge pieces to secure newness, while exploitation focused firms more likely to deploy capabilities and routines that secure integration of knowledge in the existing processes to make innovation occur.

The study offers further insights into navigating successfully the transformative interface of AC relevant for radical innovation performance. It suggests firms must consolidate their in-house knowledge resources with newly assimilated knowledge resources

(redeploy resources) (Capron, Dussauge, & Mitchell, 1998) and renew capabilities for recombining existing knowledge with the complementary new knowledge to generate new options for dealing with emerging unusual experiences (Garud et al., 2011). While the benefits of external knowledge may seem to be particularly strong, assimilating such knowledge poses stark knowledge management challenges due to its “stickiness” (Szulanski, 1996). Some knowledge might also require further refinement through processes of codiﬁcation of knowledge and past experience (Levitt & March, 1988), retrospective sense making, or the imitation of competitors’ superior practices (Szulanski, 1996). Accordingly, firms often have to go through a period of trial-and-error to learn how to deal with and benefit from external knowledge sources.

147

Once assimilated this knowledge can also be viewed as a resource for firms to extend and nurture their knowledge base. In addition, cross-functional interfaces are beneficial to integrating diverse knowledge components and to creating a desirable amount of redundancy within units (Cohen & Levinthal, 1990). They promote non-routine and reciprocal information processing and contribute to a unit's ability to overcome differences, interpret issues, and build understanding about new external knowledge (Daft & Lengel, 1986). This process changes the character of knowledge through bisociation, which occurs when a situation or idea is perceived in two self-consistent but incompatible frames of reference

(Koestler, 1966). Thus, the capability of firms to assimilate seemingly incongruous sets of information and then combine them to arrive at a new solution represents unique difficult-to- imitate capability relevant for radical innovation. Deploying such capability early allows firms to achieve first-mover advantages and, as a result, high profits (Bamiatzi et al., 2016) reflected in radical innovation performance. This can be achieved through routines focused on engaging with inter-firm communities and steering groups to analyse and interpret knowledge, facilitating knowledge integration practices focused on problem-solving, as well as facilitating valuable linkages and cross-team knowledge exchange to break silos (Freeman, 1987; Lessard

& Zaheer, 1996).

On the other hand, navigating internal AC relevant for incremental innovation performance requires different knowledge mechanisms and routines, as prior knowledge makes their innovation process more predictable and reliable by reducing the likelihood of errors and false starts for developing innovations (Ahuja & Katila, 2001). The role of human capital and organizational cognitive structures (Minbaeva et al., 2003) are crucial in this process as organizations go through trial-and-error learning (Rerup & Feldman, 2011) before utilizing the new knowledge. In this process it is more likely that firms expand the application scope of their existing knowledge, allowing access to, and efficient use of the organizational memory (Volberda et al., 2010). By embedding knowledge into organizational memory such

148

as storing it in electronic repositories or documenting best practices, firms continuously improve their responses to known categories of experiences (Garud et al., 2011). Such categories characterize a given situation that is comprehended explicitly in words through sensemaking process (Weick, Sutcliffe, & Obstfeld, 2005). This process requires firms to retrieve the knowledge that has already been harvested and incorporate it into their operations. The outcomes of managing internal AC are the persistent creation of goods, systems, processes, knowledge or organisational forms (Spender, 1996) that are new to the firm. Based on validation of routine-based AC measures and confirmation of hypothesis, as well as following discussion of the key findings of the study, the exploitation based view of

AC (Figure 1.19) and exploration view of AC (Figure 1.20) are illustrated below

Figure 1.19 Exploitation based view of AC (developed for this study)

Internal AC

Transform

Communicating and distributing external information and knowledge to internal stakeholders and departments

Using formal IT system or knowledge system to codify, store, and manage knowledge

Engaging in knowledge integration practices across divisions and departments

Using cross-functional teams as a means to pivot on existing and new knowledge Incremental innovation Exploit performance

Reflecting on and updates products, technologies and processes as a result of external search

Applying new knowledge and insights to guide decisions

Determining and allocating resources to commercial projects

Utilizing knowledge to develop and implement new technologies and products to the market.

149

Figure 1.20 Exploration based view of AC (developed for this study)

Transformative AC

Assimilate

Use of inter-firm communities and steering groups to analyse and interpret knowledge

Understanding new knowledge through trial and error learning

Keeping pace with external partners and other external constituents to interpret the environment

Using formal training programs to understand changing market demands. Radical innovation

Transform performance Communicating and distributing external information and knowledge to internal stakeholders and departments

Using formal IT system or knowledge system to codify, store, and manage knowledge

Engaging in knowledge integration practices across divisions and departments

Using cross-functional teams as a means to pivot on existing and new knowledge

Even though the study highlights differences in capabilities for pursuing exploratory versus exploitative learning, it is worth noting some potential similarities between the two processes that are mostly evident in the transform AC dimension. For example, the routine related to cross-functional interfaces enables radical innovators to bring together various sources of knowledge and increase lateral interaction between areas of functional or component knowledge. Units use cross-functional interfaces such as liaison personnel, task forces, and teams to enable knowledge exchange (Gupta & Govindarajan, 2000) and deepen knowledge flows across functional boundaries, thus reconfiguring the knowledge relevant for radical innovation. However, on the other hand such transformative routine may be beneficial

150

for incremental innovators as well since cross-functional interfaces provide an effective way of generating commitment and facilitating the implementation of decisions towards commercializing knowledge (Bahrami & Evans, 1987). Thus, cross-functional routines have potential to increase the transformation interface and internal AC and have value for both radical and incremental innovation performance. The same might hold true for other routines that shape AC transform dimension. While this research has not empirically examined such relationships, it can be a fruitful avenue for further research to map separate AC transform routines and investigates their impact on final performance outcomes. This would be particularly useful for research that attempts to explain the importance of ambidexterity by balancing exploration and exploitation, thereby deploying the AC transformative routines could be a fruitful avenue.

This study has not theorized each dimension of AC (i.e. acquire in the external stage) to be applicable to driving radical or incremental innovation performance. This is because previous research (Laursen & Salter, 2006; Leiponen & Helfat, 2010) suggested that acquisition of external knowledge is not always beneficial for innovation success (Felin &

Zenger, 2014; Foss, Husted, & Michailova, 2010). In other words, firms that acquire new external knowledge are more likely to be innovative, though “the benefits to openness are subject to decreasing returns” (Laursen and Salter 2006, p. 132), thus additional search becomes unproductive after a certain point. In particular, the reasons for insignificant role of acquire dimension may be found in the cognitive limitations of managers according to the attention-based theory of the firm (Ocasio, 1997). This theory suggests that managerial attention is a limited resource and that managers therefore need to concentrate their efforts and energy. According to this theory, managers need to “concentrate their energy, effort and mindfulness on a limited number of issues in order to achieve sustained strategic performance

(Ocasio, 1997, p. 203). Consequently, with increasing exposure to external knowledge sources, firms face an attention and resource allocation problem due to information overflow. Due to

151

the scarceness of attention, managers can only focus on a limited number of knowledge sources/channels and at the same time deal with them to sustain an acceptable level of performance (Ocasio, 1997). In such circumstances, acquiring new knowledge might lead to knowledge increasingly being underutilized, while the acquisition efforts and cost rise at the same time.

6.2 Theoretical Implications

This study is rooted in organizational learning theory which emphasizes that organizational processes are subject to significant causal ambiguity with respect to their performance implications (Lippman & Rumelt, 1982) and particularly so in rapidly changing environmental contexts (Zollo & Winter, 2002). When individual and group learning become institutionalized, organizational learning occurs and knowledge is embedded in non-human repositories such as routines (Nelson & Winter, 1982). In order to penetrate the veil of ambiguity and facilitate the adjustment of the routines, which is not a random process, firms and their managers need to manage knowledge processes and develop capabilities to facilitate dual organizational learning (Minbaeva et al., 2003). While exploitation provides higher certainty, and thus higher probability of innovation (March, 1991) exploration provides greater potential for higher innovativeness of innovation performance. Exploitation can prime exploration, and that this is more likely when it addresses diverse contexts. As exploration and exploitation emerge from contradictory knowledge-processing capabilities

(Floyd & Lane, 2000), there are different underlying logics, routines, activities, and processes that exploration and exploitation entail (March, 1991; Tushman & O'Reilly, 1996). Therefore, this study examined the impact of one of the key capabilities – AC in driving performance benefits from exploratory and exploitative learning. By conceptualizing and empirically testing dual perspective of AC - “exploitation-based view of absorptive capacity” and the

152

“exploration-based view of absorptive capacity”, the study examined the dual role of AC and thus has direct theoretical implications on AC construct.

The findings of this study suggest the existence of two dual pathways of the construct for unlocking the performance benefits from exploitative and exploratory learning. In the context of exploitative learning, the results show that only internal AC capabilities that are derived from the interaction between transform and exploit routines are required in order to benefit from external knowledge that is tacitly and experientially familiar. In the context of exploratory learning, however, the results suggest that transformative AC capabilities that are derived from the interaction between assimilate and transform are required to benefit from external knowledge that is tacitly and experientially unfamiliar. Yet, such benefits are eroded in the presence of strong exploitation routines that are tethered to old frames of reference that reverse the bisociation of knowledge that emerges from the bricolage of old and new frames of reference (Zahra & George, 2002) during the transformative stage of AC.

The study also contributes to experiential organizational learning theory (Zollo &

Winter, 2002) and the contradictions highlighted between exploitative and exploratory approaches to learning (Argote, 2011; March, 1991) by identifying the exploitation routines of

AC as the potential source of organizational tension. Specifically, the study suggests that strong exploitation routines that are developed over time through the tacit accumulation of familiar knowledge (Zollo & Winter, 2002) become core rigidities (Leonard‐Barton, 1992) that provide little flexibility to accommodate the commercial application of new, unfamiliar knowledge. In contrast, weaker exploitation routines provide more flexibility to commercially apply unfamiliar knowledge, but simultaneously hamper a firm’s ability to harvest benefits from familiar knowledge. Consequently, this poses a conundrum for organization that must decide between pursuing strong or weak exploitation routines to support either exploitative or exploratory learning outcomes respectively. It is also worth noting that despite including

153

different control variable the study found no notable differences between the industries, sizes and countries of firms.

Recent conceptualization (Lewin et al., 2011) shows that organizational routines are crucial underpinning mechanisms of AC. Therefore, this study contributes to empirical operationalization and validation of the routine-based AC construct by demonstrating the superior fit of four-factor AC model and valuable evidence on the importance of individual dimensions of AC, as well as the interactions of separate dimensions with regards to incremental and radical innovation performance. Conversely, using a unidimensional scale may hide these interactions and lead to the conclusion that such interactions do not exist. The study has laid the groundwork for the continued development of AC as the multidimensional construct by effective measurement of its dimensions, thus overcoming empirical limitations with regards to using R&D proxies. Even though the introduction of any new variable into established research area can be contentious (Laursen & Salter, 2006), the empirical deployment of the firm-level AC as four separate variables does enable researchers to better examine the relationship between AC and innovation performance. Therefore, such approach should enrich AC empirical studies in examining whether AC dimensions operate individually or whether they can operate in combination (Ambrosini & Bowman, 2009). This study particularly draws attention for empirical research to examine dynamics occurring at the intersection between different AC dimensions.

Beyond AC and organizational learning lens, the study has indirect implication on open innovation literature (Chesbrough et al., 2006) by providing a more nuanced approach to AC for dealing with external knowledge sources. At the centre of the open innovation model (Chesbrough, 2006) is how firms use knowledge of external actors in their innovation processes, thus AC routine-perspective provides new insights for dealing with such knowledge along the exploitation and exploration pathways (March, 1991). In line with a few previous studies (Foss et al., 2011; Foss et al., 2013) it draws attention to the importance of

154

organizational mechanisms in capturing value form external knowledge sources. Hence, openness to external knowledge might not necessarily translate to increasing performance benefits (Laursen & Salter, 2006) unless firms have well-established routines and achieve synergies between AC dimensions. Such mechanisms allow firms to adapt to changing environments through the integration and reshaping of both internal and external knowledge sources.

The study also has theoretical implications on knowledge management (KM) as firm- level AC rooted in one knowledge domain is not helpful in managing knowledge from another domain on incremental and radical innovation pathways. According to knowledge- based view (KBV), the firms may generate useful knowledge by combining their existing knowledge and expertise, in order to effectively create new knowledge as the crucial source of competitive advantage (Grant, 1996). Without knowledge being shared and transferred, it will remain unusable and exert only limited impact on firm effectiveness (Inkpen, 1996). As knowledge is dynamic, constantly changing, and evolving, knowledge management systems must also be robust and flexible to take frequent updates (Hoegl & Schulze, 2005a) to ensure effective deployment of different types of AC. This is particularly important for transformative AC focused on integrating new, complex and systematic sets of explicit knowledge (McInerney, 2002).

Transformative AC is similar to the concept of combinative capability (Kogut &

Zander, 1992), the integration concept (Grant, 1996) and the conﬁguration concept

(Henderson & Clark, 1990), which are all focused on synthesizing knowledge. It is also worth noting that pieces of knowledge might move backward and forward between assimilation and transformation stages of AC before integrated into the knowledge structures (Todorova &

Durisin, 2007). This process of is known as knowledge retention (Lichtenthaler &

Lichtenthaler, 2009) which may over time lead to an infinite spiral of developing capabilities to renew capabilities (Zahra et al., 2006). Therefore, the transformative stage of AC is also

155

focused on knowledge retention, thus dealing with knowledge that has been embedded in the organizational memory, which is impacted by experience, processes and knowledge repositories (Darr, Argote, & Epple, 1995). Knowledge repositories are of key significance as they are intentional remedies to increase retention and include the organization's rules and routines, altered by the processes of routine development and routine modification (Cohen &

Bacdayan, 1994). Overall, for successfully realizing the opportunities of AC for innovation, firms should establish an integrated KM approach, which includes retention of the firm’s knowledge as well as the knowledge transactions with other firms (Lichtenthaler & Ernst,

2006).

Finally, the study has implications on understanding further that dynamic capabilities

(Teece et al., 1997) are built over time through organizational routines (Nelson & Winter,

1982), and as a result are embedded in the firm. (Eisenhardt & Martin, 2000). The influential theory of dynamic capabilities (Teece et al., 1997) has been particularly useful in explaining the role of AC as the firm’s ability to deploy the knowledge necessary to facilitate organizational learning (Jansen et al., 2005b; Lane et al., 2006). This research associated a

“sensing” capability with exploration and a “seizing” capability with exploitation side of organizational learning (Birkinshaw et al., 2016). Accordingly, sensing and seizing have been considered direct counterparts to the notions of exploration and exploitation (Birkinshaw et al., 2016). In contrast to such argument, this study found that “reconfiguring” expressed through transformative AC is actually associated with exploration, rather than “sensing” expressed through external AC. On the other hand, “seizing” expressed through internal AC has been certainly found significant for exploitation, as previously stated (Birkinshaw et al.,

2016).

This study draws attention to the fact that AC as dynamic capability is dependent on different organizational learning processes where learning from experience leads to building a substantial stock of knowledge (Verreynne et al., 2016). However, such knowledge needs to

156

be developed further in order to remain dynamic and keep pace with environmental dynamics, otherwise it can become redundant (Winter, 2003) and cause inertia (Szulanski,

1996), thus deteriorating innovation performance. This is particularly important in case of radical innovation that requires changes in existing knowledge base and organizational schemas. By examining the dual role of AC, the study confirmed path dependence as a relevant aspect of dynamic capabilities that is not only influenced by learning but also by how firms develop routines in response to novelty of innovation (radical versus incremental).

Furthermore, by providing measuring instrument of AC as higher-order capability focused on learning and underpinned by organizational routines, the study demystified the essence of capability deployment on two different pathways of radical and incremental innovation.

Finally, due to implications on different theoretical streams and integrative AC-organizational learning perspective, this study can be useful for addressing the boundaries and overlaps between the DC and KM research concepts (Easterby‐Smith & Prieto, 2008).

6.3 Managerial Implications

Beyond the theoretical contributions, the results of this study also bear important insights for managerial practice in innovation management and in particular for the management of organizational learning and capability related activities. Developing valued capabilities requires a depth of experience and cumulative knowledge that cannot be quickly or easily nurtured internally. Such capabilities emerge in primitive forms (Winter, 2003) which indicates that system complexity might be low in the initial phase of capability development, but process complexity could be high. Therefore, managers may be confronted by causal ambiguity as they have little understanding of the direction in which a process might evolve or how market uncertainties might be resolved (Pandza et al., 2003). Therefore, the study provides recommendations for managers with regards to key capabilities and routines that

157

have value for organizational learning and innovation performance. Managers can benefit from them substantially as they have to commit to capabilities ahead of when they might be needed and at a time when their future value is still uncertain (Pandza et al., 2003).

The results suggest approaches to align AC dimensions with firms’ attempts to increase their innovation performance by means of considering different learning objectives and pathways (exploration versus exploitation). More specifically, from a practical perspective, the study has important implications for managers engaged in the pursuit of incremental and radical innovation projects regarding the choice of knowledge routines to unlock superior performance benefits. It also sensitizes them to the trade-off inherent in exploitation routines of AC when pursuing incremental and radical innovation simultaneously. Specifically, the study suggests that particular configurations of capabilities and routines may be better suited to different environmental contexts that vary in their dynamism and rates of change.

Therefore, before managers undertake innovation projects (radical vs incremental) they need to understand the current environmental contexts and in particular required organizational routines to help them accomplish those. What is beneficial for the projects that require a high degree of novelty and breakthrough solutions may not be beneficial for those that require only incremental improvements and vice versa.

In high-velocity environments, for example, which are characterized by rapid and discontinuous change in demand, competitors, technology and/or regulation (Bourgeois III

& Eisenhardt, 1988) managers will not benefit from the development of strong internal AC as the opportunities to benefit from tacit accumulation of experience are limited. Rather, in such conditions, managers would benefit from a heightened focus on transformative AC with weak exploitation routines as a means to be more receptive and reactive to radical change.

Managers in these contexts need to be alert to reconfiguring their processes and combing knowledge from different sources by deploying routines focused on sharing knowledge and superior practices across the firm (Lewin et al., 2011). Such routines at the transformative AC

158

interface strengthen the communication channels that are a fundamental part of the organisational sense-making process, enabling a more comprehensive picture of the firm knowledge base, and in turn identifying the fragments of knowledge that are still missing.

(Jansen et al., 2005a; Lenox & King, 2004).

Exploration process supports unit members in rethinking the systematic nature of existing products and services and revisit the ways in which components are integrated. Firms are limited in their ability to learn from past acquisitions unless they consciously develop deliberate learning routines (Zollo & Winter, 2002). Therefore, managing the transformative

AC interface requires managers to invest time and resources to identify the complementarity between the newly assimilated and existing knowledge. This can be accomplished through established steering groups, communities of practice and keeping the pace with counterparts from other companies with regards to the latest technologies. While some companies take opportunity of exchanging ideas with their peers (e.g. interpreting market trends), others use formal training programs to understand changing market demands. As the assimilation rate increases, the set of possible knowledge integration opportunities increases exponentially due the number of available knowledge elements. For this purpose, managers could deliberately facilitate interaction across business units for information dissemination and liaison between corporate groups to share knowledge obtained externally. Finally, it is recommended for managers in such contexts to build and engage cross-functional teams to transform knowledge (e.g. during the NPD process) before their competitors to gain first-mover advantage.

The opposite is true for managers that operate in low-velocity environments where tacit accumulation and benefits from internal AC are more likely. Managers in these firms should recognize the important roles of their existing innovation processes and knowledge to leverage it for incremental improvements. They need to actively manage internal selection

159

regimes (Lewin et al., 2011) as the utilization of new knowledge depends on the processes they put in place to select various projects to invest in and decisions on how to allocate resources among them. The use of external knowledge sources for exploitation often require considerable delegation of decision-making rights (Foss et al., 2013). It enables employees to interact with external sources of knowledge and understand better the nature of that knowledge for exploiting opportunities (Foss et al., 2013). Firms with high exploitative orientation often go through trial-and-error difficulties in learning how to best leverage the knowledge in their local search domain. Therefore, they constantly need to evaluate their routines and processes to decide whether to commercialize their existing knowledge or to invest further efforts to recombine it. The flexibility of exploitation focused firms would enhance their ability to apply knowledge and embed it into business activities that support superior innovation performance.

Exploitation oriented managers can benefit from internal AC by establishing routines focused on prioritizing and allocating resources to those commercial projects that have the best revenue prospects. In that process, they are likely to continually reflect on experience relevant for updating technologies and fine-tuning the products before launching improved versions on the market. As the product matures and the market expands, the number of actors with specific knowledge of various aspects of the technology increases. Learning tends to be local, and incorporating proximate knowledge does not require changes in a firm’s organizational structure (Zahra et al., 2006). Managers with exploitation focus should be alert to following market signals and feedback of customers for guiding decisions for improvement of products can be an effective mechanism to help managers on the pathway of commercializing knowledge. They should also focus on knowledge integration practices through problem-solving activities, such as developing practice amongst the firm members of asking for solutions and integrating knowledge using IT systems. Finally, a number of routines for internal knowledge sharing can be relevant for them, including regular staff

160

meetings, email digests of internal and external developments and frequently updated bulletin boards. However, even though such established routines denote stability, there is a risk of getting locked into an established system, which prevents experimentation and is therefore likely to exhibit a high degree of inertia, making it challenging for firms to respond to change

(Potts, 2000).

In order to innovate effectively, managers need to allocate their attention to a number of external knowledge sources while focusing on combining and utilizing knowledge internally in order to produce value for the firm. This study provides a nuanced framework on the basis of which managers can purposefully decide on the allocation of resources and deployment of organizational routines depending on the nature of innovation projects (radical versus incremental). It might require, however, extensive time and effort to build up an understanding of the norms, habits, routines of different external knowledge channels

(Laursen & Salter, 2006). Innovation managers should also recognize that dealing with external knowledge requires a change in the way the firm thinks about it. In this process, managers may need to minimize attitudes within the firm which may lead to a reduction of its potential to benefit from external knowledge. This attitudinal trait has been referred to as the

“not-invented-here” (NIH) syndrome (Katz & Allen, 1982; Lichtenthaler & Ernst, 2006).

Therefore, firms may suffer from tendencies to reject or to penalize knowledge if it comes from outside the firm, just for the very reason that it comes from outside the firm boundaries.

Accordingly, the NIH syndrome may induce a substitution relationship (Laursen & Salter,

2006) between the use of transformative and internal AC and cause a learning myopia

(Levinthal & March, 1993), indicating that managers may overemphasize exploitation instead of engaging in exploration, thus missing considerable opportunities for radical innovation.

While configuration of AC routines may be unique and idiosyncratic, emerging from path-dependent histories of individual firms (Teece et al., 1997), this study validated common

161

features associated with their effective deployment across firms. Therefore, managers could apply the recommended routine-based model of AC in their firms, but also bear in mind the possibility of developing AC from many starting points and along different paths (Eisenhardt

& Martin, 2000). It is important to note that the exact AC is unlikely to be identified at the beginning of the process as knowledge of organisational processes and markets evolve over them (Pandza et al., 2003) subject to changes in the external environment, which managers should be able to respond to. In addition, changes in routines can be caused by external pressures to improve performance (Feldman & Pentland, 2003), so managers may employ AC routines as a means to generate change and enhance the preferred pathway of organizational learning (March, 1991). However, incentives may be necessary to motivate people to create a general attitude toward knowledge sharing and transfer (Minbaeva et al., 2003), make use of the knowledge they absorb, develop a culture of asking for solutions and help with problem solving. This is part of developing innovation culture, which should also facilitate the acquisition of knowledge, leading to capabilities that drive firm performance (Knight &

Cavusgil, 2004). While this process, especially in case of radical innovation, requires significant commitment on the part of the members of the organization, such efforts can produce an improved understanding of the new and changing action-performance links, and therefore result in adaptive adjustments to the existing sets of routines.

6.4 Limitations and Future research

The study has a number of limitations that may serve as motivation for future research. First, although the study and insights are derived from a rich set of quantitative data and hold in light of a number of post-hoc analyses and robustness checks via alternative specifications, the cross-sectional design utilized remains a questionable approach for making causal inferences. Notwithstanding the clear set of findings from this study and strength of

162

theoretical explanations that support them, alternative interpretations may exist, such as the presence of a complementary management innovation (Birkinshaw, Hamel, & Mol, 2008) to develop new routines and capabilities that support firms’ innovation performance or expenditures on R&D that support innovation activities, which the data of this study is unable to rule out or account for. Future research can adopt a longitudinal or panel research design to account for these potential endogeneity issues owing to omitted variables and causally verify the results presented in the paper by employing natural instruments.

Second, despite the processual nature of the AC construct, the present study only considered the effects of separate AC stages and dimensions on outcomes of performance and verified only a subset of stages in driving performance gains in incremental and radical innovation. Therefore, not all the AC dimensions and stages may be tethered to performance outcomes and may instead be linked to intermediary outcomes that exist between separate stages of the overall AC process that precede performance, such as stimulating new R&D opportunities. In other words, AC is a process-based construct that has a temporal aspect to it that this study cannot by its design account for as it does not measure changes in AC and innovative performance that might occur over a longer period of time. Again, this is a downfall of the cross-sectional design that can be addressed by future research that could adopt a multi-level perspective of the phenomenon with a wider set of variables that will allow for more fine-grained insights into the mechanisms underlying exploitative and exploratory learning. It is also worth noting the limitation of the study with regards to focusing on distinction between incremental and radical innovation only, hence future research could examine the effects of AC dimensions on other types of innovation.

Furthermore, reasons for differences with regards to innovation performance of firms may extend beyond the deployment of AC. While AC is one of the key drivers of it, other variables may influence changes in levels of innovation performance. Therefore, it is

163

acknowledged, that reasons may be found in the different market conditions of the respective industries as well as in different underlying technology bases and their complexity or innovation proneness, or the alignment of external search with internal resources. Yet, reasons may also be found in differences that industries and firms exhibit with regard to their organization of innovation, knowledge search-related activities or organizational culture.

Examining the role of culture in such context would be an interesting avenue for further research to enrich our understanding with regards to “not-invented-here” syndrome mentioned earlier. Since radical and incremental innovation pathways entail contradictory organizational routines, firms should be able to develop the appropriate organizational culture and social processes that facilitate their integration (Gibson & Birkinshaw, 2004).

Organizational culture influences knowledge management, and as a result knowledge transfer, as it creates the context within which the firms shape assumptions about knowledge, and constitutes the context for the social interactions required for the exchange of knowledge (De

Long & Fahey, 2000).

The researcher also acknowledges that the sample and setting of this study focused on organizations affiliated with a professional innovation management association may have idiosyncratic characteristics that effect the generalizability of results. Future research should focus on exploring and replicating the research through a large ‘N’ study that can account for generalizations to other samples and settings where alternative characteristics are present, explaining possible differences in the mechanism across different size and industry categories.

Although, it is expected that results will hold true in a wider context given the diversity of industrial contexts accounted for in this dataset. Furthermore, as items for AC are empirically examined for inherently difficult to measure constructs/dimensions, it would be useful for future studies to relate objective measures of AC to scales used in this study. Future research could therefore apply rigorous steps by further confirming or amending the AC measurement

164

scales to ensure reliability and predictive testing, relating them to different performance outcomes.

It important to note that the data in this study incorporates self-reported assessments of innovation managers and although several steps were undertaken in both the design and testing phases to limit concerns regarding single-informant data, the issues of key informant bias and common method bias cannot be completely eliminated. However, strong interrater agreement and interrater reliability, together with the confidentiality that was assured for respondents reduced concerns that participants artificially inflated their responses. Common method bias would have produced consistent effects of the same variables on both types of innovative performance, yet differential effects of several AC variables on final outcomes was found. In addition, there are research limitations with regards to goodness of fit (GOF) indices for absorptive capacity measures that do not entirely meet the stringent thresholds prescribed by the previous studies (Hair Jr et al., 2010; Jöreskog & Sörbom, 1993). Despite this limitation, chi-square difference test showed clear advantage of the proposed four-factor model of absorptive capacity in comparison to alternative models. Future research could examine further how the proposed routine-based (four-factor) model of absorptive capacity fits the data in different research contexts and report any discrepancies to our findings.

Previous literature suggests that ambidextrous organizations should manage actively the knowledge processes, in particular knowledge retention (Lichtenthaler & Lichtenthaler,

2009) which may over time lead to an ‘infinite spiral’ of developing capabilities to renew capabilities (Zahra et al., 2006). Therefore, a particularly fruitful area for future research in this regard would be to examine the role of AC in the simultaneous pursuit of exploitative and exploratory learning as part of the wider ambidexterity debate. By drawing on the AC routine- based building blocks that unlock each organizational learning mechanism examined in this study, future research could provide recommendations for balancing tensions between them, thereby advancing a capability perspective of the ambidextrous organization. Future research

165

may also investigate how contextual ambidextrous units within the firms combine contradictory elements (Gibson & Birkinshaw, 2004) and increase their levels of different types of AC simultaneously. In addition, the locus of organizational knowledge search with regard to the search intention and the kind of knowledge searched for in association to exploration and exploitation, or ambidexterity could be an interesting area for future research.

Future studies may also incorporate multiple levels of analysis and examine moderating effects of organizational innovation to further enhance understanding of how AC contributes to innovative performance. It is also worth noting that even though the study hasn’t found that acquire dimension of AC is associated with innovation performance, it might still determine some interim outcomes. Therefore, future research could examine the relationship between AC dimensions and interim variables such as R&D intensity and management innovation. In addition, it could investigate the role of acquire and other AC dimensions on the curvilinear (inverted U-shape) relationship between search patterns and innovative performance found earlier (Laursen & Salter, 2006). Such analysis could inform research about the potential of AC dimensions in moving the tipping point (Laursen & Salter, 2006) to eventually mitigate negative over-search tendencies (Katila & Ahuja, 2002).

Further research could also link the routine-based model of AC to the inbound and outbound open innovation activities that reside in separate functional silos of the firms with internal and external focus (Schroll & Mild, 2011). Such analysis could be beneficial in diversifying the open search strategy (Leiponen & Helfat, 2010) and navigating how different dimensions of AC can yield the best results for various knowledge sources (i.e. suppliers, clients, etc.). According to Katila and Ahuja (2002) search scope (local and distant) and search depth provide extension of March’s (1991) distinction between exploitation and exploration, so empirical deployment of routine-based model of AC can be beneficial in explaining the impact of search patterns on radical and incremental performance. Such analysis would add value and build on knowledge of previous studies that investigated how AC influences firms’

166

engagement in broad or deep search strategies (Grimpe & Sofka, 2009) and the effects of such search strategies on innovation performance (Bahemia & Squire, 2010).

Finally, this study focused on analysis of firms of different sizes across industries in several OECD countries/developed economies, that are members of global innovation association. These firms particularly operate within knowledge-intensive and technology- oriented industries, especially in services and advanced manufacturing. However, future research could examine the phenomenon and report similarities or differences in the context of emerging market enterprises (EME)(Piperopoulos et al., 2018). Since EMEs operate in innovation systems with under-developed institutional landscape that constrains the development of internal capabilities for innovation (Luo & Tung, 2007), caution has been recommended in applying existing organizational learning frameworks to EMEs

(Piperopoulos et al., 2018). International business (IB) research suggested that organizational learning, knowledge development (Carneiro, Bamiatzi, & Cavusgil, 2018) as well as organizational slack (Johanson & Vahlne, 1977) are critical elements of the successful internationalisation of firms (Coviello & Munro, 1995) particularly with the advent of advanced information and communication technologies and rapid globalization. These processes have facilitated the movement of knowledge across national boundaries (Jones &

Coviello, 2005). In such borderless environment for knowledge sharing (Coviello, 2015;

Knight & Cavusgil, 2004; Mathews, 2006), innovation depends on the firm’s capability to learn and integrate diverse knowledge and resources from multiple countries (Hitt, Hoskisson,

& Kim, 1997). Therefore, the analysis of AC and organizational learning on innovation performance, which requires adaptation of the model through IB theoretical lens, could be a fruitful avenue for further research.

167

6.5 Conclusion

This study sought to understand and examine the differences in the deployment of AC dimensions across external, transformative, and internal stages for driving performance benefits of exploitative and exploratory learning in the form of incremental and radical innovation performance. While prior research has largely conceptualized and empirically deployed AC as a uniform construct (e.g., (Lewin et al., 2011) irrespective of the learning outcomes pursued by the organization, this view neglects the possibility for variations in the deployment of the AC construct for learning outcomes that rely on contradictory knowledge capabilities and routines (Benner & Tushman, 2003). The study theorized that for exploitative learning that is focused on the refinement and extension of existing knowledge through incremental innovation, internal AC is most critical for driving performance benefits. In contrast, for exploratory learning that is focused on the variation, replacement, and renewal of existing knowledge through radical innovation, transformative AC coupled with low exploitation capability is the most critical for driving performance benefits. The findings of this study confirmed these theoretical assertions and in doing so provide noteworthy contributions to the study of AC and organizational learning respectively. The findings open new avenues for operationalizing AC through a routine-based model, overcoming the issue of the construct uniformity. While exploration focused firms benefit mostly from transformative

AC to secure diversity and integration of knowledge, the exploitation focused firms are more likely to deploy internal AC to secure operationalization in the existing processes and make the incremental innovation occur.

This study also sought to understand the underpinning constituents of AC dimensions at the firm level in order to advance research on AC by empirically validating the conceptual distinction between its dimensions. In doing so, the study operationalized AC construct and validated meaningful scales to test four separate dimensions, demonstrating a superiority of

168

such routine-based model comparing to the rival single factor and two-factor models. Once again, this study raises concerns with regards to combining all items on one measure of AC, which can “diminish our quest to provide both a more taxonomical and a more functional evidential approach” (Verreynne et al., 2016, p. 4301) to the study of AC. Therefore, the AC measurement in this study provides both a significant empirical validation of the routine- based model and an effective capability development tool for innovation managers.

Examining routines underpinning different AC dimensions not only clarifies how it may be developed for different types of learning, but also reveal possibilities for managing AC stages successfully. The challenge for the future research is to test these scales in other contexts using longitudinal research designs where independent (AC dimensions) and dependent variables (innovation performance) can be tested in different time intervals and possibly include other mediation or moderation models.

In conclusion, this study offers intriguing insights for both researchers and practitioners by acknowledging multifaceted nature of AC with new insights regarding how firms may deploy it to gain increasing levels of innovation performance. Collectively, the findings surface an interesting tension that occurs during the internal stage of AC, which enables incremental performance benefits from exploitative learning but erodes radical performance benefits from exploratory learning. This finding has important implications with regards to our understanding of AC as a flexible rather than uniform construct. Specifically, the results suggest that the organization’s learning orientation is a critical boundary condition that determines how AC stages and dimensions are deployed to unlock performance benefits.

The investigation of firms’ AC and its performance effects adds to theoretical knowledge about firms’ organizational learning activities, and aids managers in their efforts to manage knowledge successfully. In sum, the contribution of the research is evident in investigating the dual nature of AC for firms’ radical and incremental innovation performance and the particular emphasis that research has given to validation of routine-based AC model as

169

discussed previously. The results of the study provide ground for an ongoing and more differentiated examination of the conditions suitable for AC deployment. The study shows that firms have substantial means to increase their potential to benefit from routine-based model of AC that unlocks benefits of radical and innovative performance. Despite the limitations and demands for further research, the study has provided valuable insights for both academic research and management practice.

170

Appendix

Appendix 1: Ethics approval

17 August 2017

BL-EC 37-17 The role of absorptive capacity in facilitating open innovation outcomes

Thank you for submitting the above project for consideration by the Faculty Human Ethics Advisory Group (HEAG). The HEAG recognised that the project complies with the National Statement on Ethical Conduct in Research Involving Humans (2007) and has approved it. You may commence the project upon receipt of this communication. The approval period is for four years. It is your responsibility to contact the Faculty HEAG immediately should any of the following occur:

• Serious or unexpected adverse effects on the participants • Any proposed changes in the protocol, including extensions of time • Any changes to the research team or changes to contact details • Any events which might affect the continuing ethical acceptability of the project • The project is discontinued before the expected date of completion.

You will be required to submit an annual report giving details of the progress of your research. Failure to do so may result in the termination of the project. Once the project is completed, you will be required to submit a final report informing the HEAG of its completion.

Please ensure that the Deakin logo is on the Plain Language Statement and Consent Forms. You should also ensure that the project ID is inserted in the complaints clause on the Plain Language Statement, and be reminded that the project number must always be quoted in any communication with the HEAG to avoid delays. All communication should be directed to [email protected]

The Faculty HEAG and/or Deakin University Human Research Ethics Committee (HREC) may need to audit this project as part of the requirements for monitoring set out in the National Statement on Ethical Conduct in Research Involving Humans (2007).

If you have any queries in the future, please do not hesitate to contact me.

We wish you well with your research.

Kind regards,

BL-HEAG Secretariat

171

Appendix 2: Survey questionnaire

You have been invited to participate in a research project that seeks to understand the firm-level practices and contracting mechanisms used to support different types of innovative activities undertaken within your organisation. The survey will ask you as an informed respondent to reflect on and provide information with regards to the firm's general innovative activities, observed practices and routines, and the types of contracting arrangements used during the two years (2016 and 2017).

The study is designed as part of collaboration with Deakin Business School, Melbourne, Australia. It aims to provide clarity on how firms and managers should approach different types of innovative activities in terms of resources and capabilities, search patterns, and the complementary contracting arrangements for driving performance.

Participation is strictly voluntary and the survey is designed in a simplistic and intuitive way. It will take no longer than 15 minutes to complete. Please provide answers to the best of your knowledge. The study will provide significantly new insights that will improve professional practice and the way individuals and organizations should approach projects and other internal innovation activities. Results will be made available to all respondents and a follow up discussion will be organised (if demand is there) to provide you with the opportunity to listen and ask questions about the study and what it means for you.

If you have any other questions regarding the study, please contact Boris Radickovic at [email protected]. By clicking the consent button below, you are indicating that you agree to participate and will be directed to the survey instrument. In the case of any complaints you also may contact: The Manager, Ethics and Biosafety, Deakin University, [email protected]. Please quote project number BL-EC 37-17.

Thank you in advance for your support.

Boris Radickovic (PhD candidate); Dr Matt Mount; Dr Heather Round

Consent

I agree to participate in the study

Part I The firm's search activities during the two years (2016 and 2017)

1. In the two-year time period has the firm engaged in any local* search activities for accessing new information and knowledge?

*Local search refers to search activities performed within the existing or a very familiar knowledge domain that aligns with that currently possessed within the organisation

Yes

172

2. Please identify the intensity of local search activity across the following different types of external sources (0=none, 1=low to 7=high)

0 1 2 3 4 5 6 7

Suppliers of equipment, materials, components, or software

Clients or customers

Competitors

Consultants

Commercial laboratories/R&D enterprises

Universities or higher education institutes

Government research organizations

173

0 1 2 3 4 5 6 7

Other public sector, e.g. business links, government offices

Private research institutes

Trade associations

3. In the two-year time period has the firm engaged in any distant* search activities for accessing new information and knowledge?

*Distant search refers to search activities performed outside of an existing or familiar knowledge domain of the organisation and refers to knowledge that is unfamiliar

Yes

4. Please identify the intensity of distant search activity across the following different types of external sources (0=none, 1=low to 7=high)

0 1 2 3 4 5 6 7

Suppliers of equipment, materials, components, or software

Clients or customers

Competitors

174

0 1 2 3 4 5 6 7

Consultants

Commercial laboratories/R&D enterprises

Universities or other higher education institutes

Government research organizations

Other public sector, e.g. business links, government offices

Private research institutes

Trade associations

Part II Outcomes of innovative activities (product innovation) during the two years (2016 and 2017)

Product innovation refers to a good or service that is either substantially new, or an improved version of previous goods or services.

5. In the two-year time period has the firm introduced any of the following?

175

Yes No

Goods innovations: Substantially new or improved goods

Service innovations: Substantially new or improved services

6. Who has developed these product innovations? Tick all that apply

Your organisation by adapting or modifying goods or Your services organisation originally Your together with developed by organisation other other Other by itself organisations organisations organisations

Goods innovations Service innovations

7. Please estimate the percentage (choices must total 100%) of the firm's total turnover in the previous year (2017 only) from:

0 Substantially new products that were new to your market 0 Improved products that were only new to your organisation

Products that were unchanged or only marginally modified (include the resale of new 0 products purchased from other organisations)

Total 0

8. Have any of the product innovations been:

Yes No

A first in your country 176

Yes No

A world first

9. Please estimate the percentage of your total turnover in 2017 from world first product innovations? (This should be a subset of your new-to-market turnover share in question 7)

0% to less than 1%

1% to less than 5%

5% to less than 10%

10% to less than 25%

25% or more

Part III Process and organisational innovation during the two years (2016 and 2017)

A process innovation is the implementation of a new or improved production process, distribution or supporting activity.

An organisational innovation is a new or improved method of working and includes new business practices, structural configuration, workplace organisation, or external relations that has not been previously used by the firm.

10. In the two-year time period has the firm introduced any of the following?

New or improved methods of manufacturing for producing goods or services

New or improved logistics, delivery or distribution methods for your inputs, goods or services

New or improved supporting activities for your processes, such as maintenance systems operations for purchasing, accounting, or computing

New or improved business practices for organising procedures (i.e. first time use of sup p management, business re-engineering, knowledge management, quality management, e

New or improved methods of organising work responsibilities and decision making (i.e. t decentralisation, integration or de-integration of departments, education/training systems

New or improved methods of organising external relations with other firms or public organizations (i.e. first time use of alliances, partnerships, outsourcing, etc.)

11. From question 10, please indicate the most prominent process and organisational innovation(s) introduced.

Tick all that apply

177

New supply chain management procedure

Business process re-engineering project

New knowledge management procedure

New TQM procedure

New technological systems

Team working structure

Decentralised structure

Centralised structure

Strategic alliances

Innovation partnerships

Outsourcing

Joint venture

Part IV R&D activities and expenditures during the two years (2016 and 2017)

12. Has your firm engaged in the following R&D activities during the two years 2016 to 2017?

Yes No

In-house R&D (Creative work undertaken within your organisation to increase the stock of knowledge for developing new and improved products, including in-house software development

External R&D (Same activities as above, but performed by other firms or public/private research organisations and purchased by your organisation)

Acquisition of machinery, equipment and software (acquisition of advanced machinery, equipment and computer hardware or software to produce new or improved products)

Acquisition of external knowledge (Purchase or licensing of patents and non-patented inventions, know-how and other types of knowledge from other firms for the development of new or improved products)

13. Please estimate the amount of expenditure (in USD) for each of the following R&D activities in 2017 only (include personnel and related costs)

In-house R&D ______Purchase of external R&D ______

178

Part V Organizational routines for knowledge management and innovation during the two years (2016 and 2017)

14. Please identify the most common practices used to identify and acquire external knowledge over the last two years?

Not applicable Rarely Moderately Regularly

Using of formal or informal gatekeepers to monitor the external environment

Using formal environmental probing and scanning mechanisms

Market research, end-user surveys and other forms of primary research

Using patent records, trade magazines and other forms of secondary research

Establishing co-development relationships with external partners

Consulting with important end users or customers

Using on and offline systems or networks for collaboration

Using open sourcing methods - innovation communities, crowdsourcing etc.

Using leadership positions and memberships to complementary organisations and professional bodies

Attending conferences, trade shows, and other events

15. Please identify the most common practices used to interpret and understand external knowledge over the last two years?

Not applicable Rarely Moderately Regularly

Using inter-firm communities and steering groups

External secondments and sabbaticals

179

Not applicable Rarely Moderately Regularly Keeping pace with the partners with regards to the rate of knowledge creation

Reflecting on and updating products, professional bodies technologies, and processes that occur as a result of external search

Learning from good and bad experience - trial and error

Learning from managing alliances

Formal training programs (in-house and external)

Replicate and imitate external practices (copy exact principle)

16. Please identify the most common practices used to combine and integrate external knowledge with internal knowledge over the last two years?

Not applicable Rarely Moderately Regularly

Communicating and distributing external information and knowledge to internal stakeholders

Using formal IT system or knowledge system to codify, store and manage knowledge

Visiting other enterprise's divisions for integrating diverse knowledge components

Forming cross-functional teams

Solicitation of scientists and engineers to propose and pursue innovative ideas

Fostering informal interactions among employees (e.g. open office plan, open town halls, etc.)

Encouraging knowledge sharing inside the enterprise to eventually lead to cross- business innovations (e.g. Technology Forum, Technical Council)

Use of brainstorming sessions and sandpits to bring together diverse organisational actors

180

17. Please identify the most common practices used for applying and exploiting new knowledge for commercial ends over the last two years?

Not Applicable Rarely Moderately Regularly

Shared sense in determining projects to be funded and allocation of resources among them

Seeking market signals for guiding decisions

Developing prototypes that perform at least as well as what is available on the market

Autonomy of middle management to support and allocate resources outside CEO's vision

Enhancing internal rate of change (e.g. creating goals and expectations that stimulate change)

Imposing "stretch goals" or conducting best in class benchmarks comparisons

Part VI Contracting for managing external knowledge and innovation during the two years (2016 and 2017)

18. Has the firm built a contractual relationship for innovation with other firms during the last two years?

Yes

19. Please specify the type of contractual relationship and the estimate percentage (choices must total 100%)

Informal 0

Formal 0

Total 0

20. Please specify the type of contract and the estimate percentage (choices must total 100%)

Relational 0

Transactional 0

Total 0

181

Part VII General information about the firm

21. What country are you based in?

Please select

22. What is the main industry in which your enterprise operates?

Please select

23. What is the size* of your enterprise?

*European Commission classification of enterprises by size is used.

Micro (less than 10 employees)

Small (10-49 employees)

Medium (50-249 employees)

Large (250 or more employees)

24. Please estimate your organisation's total turnover (in USD) for 2016 and

2017. *Turnover is defined as the market sales of goods and services (include all taxes except VAT)

0 2016

0 2017

25. Would you like to be invited to a follow up discussion for receiving insights and discussing findings of this research project?

Yes

182

References

Ackroyd, S., & Fleetwood, S. 2000. Realist Perspectives on Management and Organisation Studies. London and New York: Routledge. Ackroyd, S., & Hughes, J. A. 1981. Data Collection in Context: Longman. Adams, R., Bessant, J., & Phelps, R. 2006. Innovation management measurement: A review. International Journal of Management Reviews, 8(1): 21-47. Adler, P. S., Goldoftas, B., & Levine, D. I. 1999. Flexibility versus efficiency? A case study of model changeovers in the Toyota production system. Organization Science, 10(1): 43-68. Adner, R., & Helfat, C. E. 2003. Corporate effects and dynamic managerial capabilities. Strategic Management Journal, 24(10): 1011-1025. Adner, R., & Kapoor, R. 2010. Value creation in innovation ecosystems: How the structure of technological interdependence affects firm performance in new technology generations. Strategic Management Journal, 31(3): 306-333. Afuah, A., & Tucci, C. L. 2012. Crowdsourcing as a solution to distant search. Academy of Management Review, 37(3): 355-375. Ahmad, A., Umar Bello, M., Kasim, R., & Martin, D. 2014. Positivist and Non-Positivist Paradigm in Social Science Research: Conflicting Paradigms or Perfect Partners? Journal of Management and Sustainability, 4(3): 79-95. Ahuja, G. 2000. Collaboration networks, structural holes, and innovation: A longitudinal study. Administrative Science Quarterly, 45(3): 425-455. Ahuja, G., & Katila, R. 2001. Technological acquisitions and the innovation performance of acquiring firms: A longitudinal study. Strategic Management Journal, 22(3): 197-220. Ahuja, G., & Morris Lampert, C. 2001. Entrepreneurship in the large corporation: A longitudinal study of how established firms create breakthrough inventions. Strategic Management Journal, 22(6/7): 521-543. Aiken, L. S., West, S. G., & Reno, R. R. 1991. Multiple regression: Testing and interpreting interactions: Sage. Alavi, M., & Leidner, D. E. 2001. Knowledge management and knowledge management systems: Conceptual foundations and research issues. MIS Quarterly, 25(1): 107-136. Ambrosini, V., & Bowman, C. 2009. What are dynamic capabilities and are they a useful construct in strategic management? International Journal of Management Reviews, 11(1): 29-49. Amemiya, T. 1985. Advanced econometrics: Harvard university press. Anand, A., & Singh, M. 2000. Understanding knowledge management. International Journal of Engineering Science and Technology, 3(2): 926-939. Andriopoulos, C., & Lewis, M. W. 2009. Exploitation-exploration tensions and organizational ambidexterity: Managing paradoxes of innovation. Organization Science, 20(4): 696-717. Argote, L. 2011. Organizational learning research: Past, present and future. Management Learning, 42(4): 439-446. Argote, L. 2012. Organizational learning: Creating, retaining and transferring knowledge: Springer Science & Business Media. Argote, L. 2015. An opportunity for mutual learning between organizational learning and global strategy researchers: transactive memory systems. Global Strategy Journal, 5(2): 198- 203. Argote, L., & Ingram, P. 2000. Knowledge transfer: A basis for competitive advantage in firms. Organizational Behavior and Human Decision Processes, 82(1): 150-169. Argote, L., McEvily, B., & Reagans, R. 2003. Managing knowledge in organizations: An integrative framework and review of emerging themes. Management Science, 49(4): 571-582.

183

Argyres, N. S., & Silverman, B. S. 2004. R&D, organization structure, and the development of corporate technological knowledge. Strategic Management Journal, 25(8/9): 929-958. Argyris, C., & Schon, D. 1978. Organizational learning: A theory of action approach. Argyris, C., & Schon, D. A. 1996. Organiational learning II: Addison Wesley. Argyris, C., & Schön, D. A. 1997. Organizational learning: A theory of action perspective. Reis,(77/78): 345-348. Ashforth, B. K., & Saks, A. M. 1996. Socialization tactics: Longitudinal effects on newcomer adjustment. Academy of Management Journal, 39(1): 149-178. Atuahene-Gima, K. 2005. Resolving the capability—rigidity paradox in new product innovation. The Journal of Marketing, 69(4): 61-83. Baert, P. 1998. Social theory in the twentieth century. Washington Square, NY: New York University Press. Bahemia, H., & Squire, B. 2010. A contingent perspective of open innovation in new product development projects. International Journal of Innovation Management, 14(04): 603- 627. Bahrami, H., & Evans, S. 1987. Stratocracy in high-technology firms. California Management Review, 30(1): 51-66. Bamiatzi, V., Bozos, K., Cavusgil, S. T., & Hult, G. T. M. 2016. Revisiting the firm, industry, and country effects on profitability under recessionary and expansion periods: A multilevel analysis. Strategic Management Journal, 37(7): 1448-1471. Baregheh, A., Rowley, J., & Sambrook, S. 1996. Towards a multidisciplinary definition of Innovation, Management Decision. Australia Business Innovation Analysis, 47(8 ): 1323- 1339. Barkema, H. G., & Vermeulen, F. 1998. International expansion through start-up or acquisition: A learning perspective. Academy of Management Journal, 41(1): 7-26. Barney, J. 1991. Firm resources and sustained competitive advantage. Journal of Management, 17(1): 99-120. Barney, J., Wright, M., & Ketchen, D. J. 2001. The resource-based view of the firm: Ten years after 1991. Journal of Management, 27(6): 625-641. Barney, J. B. 1995. Looking inside for competitive advantage. The Academy of Management Executive, 9(4): 49-61. Barrett, P. 2007. Structural equation modelling: Adjudging model fit. Personality and Individual Differences, 42(5): 815-824. Bartus, T. 2005. Estimation of marginal effects using margeff. Stata Journal, 5(3): 309-329. Baum, J. A., Li, S. X., & Usher, J. M. 2000. Making the next move: How experiential and vicarious learning shape the locations of chains' acquisitions. Administrative Science Quarterly, 45(4): 766-801. Becker, M. C., & Knudsen, T. 2005. The role of routines in reducing pervasive uncertainty. Journal of Business Research, 58(6): 746-757. Becker, M. C., Lazaric, N., Nelson, R. R., & Winter, S. G. 2005. Applying organizational routines in understanding organizational change. Industrial and Corporate Change, 14(5): 775-791. Belderbos, R., Carree, M., & Lokshin, B. 2004. Cooperative R&D and firm performance. Research Policy, 33(10): 1477-1492. Belderbos, R., Cassiman, B., Faems, D., Leten, B., & Van Looy, B. 2014. Co-ownership of intellectual property: Exploring the value-appropriation and value-creation implications of co-patenting with different partners. Research Policy, 43(5): 841-852. Benner, M. J., & Tushman, M. 2002. Process management and technological innovation: A longitudinal study of the photography and paint industries. Administrative Science Quarterly, 47(4): 676-707.

184

Benner, M. J., & Tushman, M. L. 2003. Exploitation, exploration, and process management: The productivity dilemma revisited. Academy of Management Review, 28(2): 238-256. Bianchi, M., Cavaliere, A., Chiaroni, D., Frattini, F., & Chiesa, V. 2011. Organisational modes for Open Innovation in the bio-pharmaceutical industry: An exploratory analysis. Technovation, 31(1): 22-33. Biedenbach, T., & Müller, R. 2012. Absorptive, innovative and adaptive capabilities and their impact on project and project portfolio performance. International Journal of Project Management, 30(5): 621-635. Birkinshaw, J., Hamel, G., & Mol, M. J. 2008. Management innovation. Academy of Management Review, 33(4): 825-845. Birkinshaw, J., Zimmermann, A., & Raisch, S. 2016. How do firms adapt to discontinuous change? California Management Review, 58(4): 36-58. Blackler, F. 1995. Knowledge, knowledge work and organizations: An overview and interpretation. Organization Studies, 16(6): 1021-1046. Blasius, J., & Thiessen, V. 2012. Assessing the quality of survey data: Sage. Blomqvist, K., & Levy, J. 2006. Collaboration capability–a focal concept in knowledge creation and collaborative innovation in networks. International Journal of Management Concepts and Philosophy, 2(1): 31-48. Blumer, H. S. 1969. Symbolic Interactionism: Perspectives and Method. Englewood Cliffs, NJ: Prentice Hall Bontis, N., Crossan, M. M., & Hulland, J. 2002. Managing an organizational learning system by aligning stocks and flows. Journal of Management Studies, 39(4): 437-469. Boudreau, K. 2010. Open platform strategies and innovation: Granting access vs. devolving control. Management Science, 56(10): 1849-1872. Bouncken, R. B., & Kraus, S. 2013. Innovation in knowledge-intensive industries: The double- edged sword of coopetition. Journal of Business Research, 66(10): 2060-2070. Bourgeois III, L. J., & Eisenhardt, K. M. 1988. Strategic decision processes in high velocity environments: Four cases in the microcomputer industry. Management Science, 34(7): 816-835. Bowker, G., & Star, S. L. 1999. Sorting things out. Brandenburger, A. M., & Nalebuff, B. J. 1996. Co-opetition: A revolutionary mindset that combines competition and cooperation in the marketplace, Harvard Business School Press, Boston. Brouwer, E., Kleinknecht, A., & Reijnen, J. 1993. Employment growth and innovation at the firm level. Journal of Evolutionary Economics, 3(2): 153-159. Brown, S. L., & Eisenhardt, K. M. 1997. The art of continuous change: Linking complexity theory and time-paced evolution in relentlessly shifting organizations. Administrative Science Quarterly, 42(1): 1-34. Brunswicker, S., & Van de Vrande, V. 2014. Exploring open innovation in small and medium-sized enterprises. Bryman, A. (Ed.). 2004. Social research methods (2nd ed.). Oxford, UK: Oxford University Press. Brynjolfsson, E., & Hitt, L. M. 2000. Beyond computation: Information technology, organizational transformation and business performance. The Journal of Economic Perspectives, 14(4): 23-48. Bush, A. A., & Tiwana, A. 2005. Designing sticky knowledge networks. Communications of the ACM, 48(5): 66-71. Byrne, B. M. 2013. Structural equation modeling with AMOS: Basic concepts, applications, and programming: Routledge. Cameron, A. C., & Trivedi, P. K. 2010. Microeconometrics using stata: Stata press College Station, TX. 185

Capron, L., Dussauge, P., & Mitchell, W. 1998. Resource redeployment following horizontal acquisitions in Europe and North America, 1988–1992. Strategic Management Journal, 19(7): 631-661. Capron, L., & Hulland, J. 1999. Redeployment of brands, sales forces, and general marketing management expertise following horizontal acquisitions: A resource-based view. The Journal of Marketing, 63(2): 41-54. Carneiro, J., Bamiatzi, V., & Cavusgil, S. T. 2018. Organizational slack as an enabler of internationalization: The case of large Brazilian firms. International Business Review, 27(5): 1057-1064. Cassiman, B., & Valentini, G. 2016. Open innovation: Are inbound and outbound knowledge flows really complementary? Strategic Management Journal, 37(6): 1034-1046. Cassiman, B., & Veugelers, R. 2006. In search of complementarity in innovation strategy: Internal R&D and external knowledge acquisition. Management Science, 52(1): 68-82. Cavusgil, S. T., & Knight, G. 2015. The born global firm: An entrepreneurial and capabilities perspective on early and rapid internationalization. Journal of International Business Studies, 46(1): 3-16. Cepeda, G., & Vera, D. 2007. Dynamic capabilities and operational capabilities: A knowledge management perspective. Journal of Business Research, 60(5): 426-437. Chadee, D., & Raman, R. 2012. External knowledge and performance of offshore IT service providers in India: the mediating role of talent management. Asia Pacific Journal of Human Resources, 50(4): 459-482. Chandler, A. D. 1962. Strategy and structure: Chapters in the history of the American enterprise, Massachusetts Institute of Technology Cambridge, Vol. 4: 125-137. Chandy, R. K., & Tellis, G. J. 2000. The incumbent’s curse? Incumbency, size, and radical product innovation. The Journal of Marketing, 64(3): 1-17. Chesbrough, H. 2003. Open innovation: the new imperative for creating and profiting from technology. Boston, MA: Harvard Business School Press. Chesbrough, H. 2007. The market for innovation: implications for corporate strategy. California Management Review, 49(3): 45-66. Chesbrough, H., & Crowther, A. K. 2006. Beyond high tech: early adopters of open innovation in other industries. R&D Management, 36(3): 229-236. Chesbrough, H., Vanhaverbeke, W., & West, J. (Eds.). 2006. Open innovation: Researching a new paradigm. Oxford, UK: Oxford University Press Chesbrough, H. W. 2006. Open Innovation: A New Paradigm for Understanding Industrial Innovation. Oxford: Oxford University Press. Christensen, C. 2013. The innovator's dilemma: when new technologies cause great firms to fail. Boston, MA: Harvard Business Review Press. Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. 2013. Applied multiple regression/correlation analysis for the behavioral sciences: Routledge. Cohen, M. D., & Bacdayan, P. 1994. Organizational routines are stored as procedural memory: Evidence from a laboratory study. Organization Science, 5(4): 554-568. Cohen, W. M., Goto, A., Nagata, A., Nelson, R. R., & Walsh, J. P. 2002. R&D spillovers, patents and the incentives to innovate in Japan and the United States. Research Policy, 31(8-9): 1349- 1367. Cohen, W. M., & Levinthal, D. A. 1989. Innovation and learning: the two faces of R & D. The Economic Journal, 99(397): 569-596. Cohen, W. M., & Levinthal, D. A. 1990. Absorptive Capacity: A New Perspective on Learning and Innovation. Administrative Science Quarterly, 35(1): 128-152.

186

Cook, S. D., & Brown, J. S. 1999. Bridging epistemologies: The generative dance between organizational knowledge and organizational knowing. Organization Science, 10(4): 381- 400. Cornell University, I., and WIPO. 2016. The Global Innovation Index 2016: Winning with Global Innovation. Ithaca, Fontainebleau, and Geneva. Cottrell, T., & Nault, B. R. 2004. Product variety and firm survival in the microcomputer software industry. Strategic Management Journal, 25(10): 1005-1025. Coviello, N. 2015. Re-thinking research on born globals. Journal of International Business Studies, 46(1): 17-26. Coviello, N. E., & Munro, H. J. 1995. Growing the entrepreneurial firm: networking for international market development. European journal of marketing, 29(7): 49-61. Cowan, R., & Jonard, N. 2009. Knowledge portfolios and the organization of innovation networks. Academy of Management Review, 34(2): 320-342. Crépon, B., Duguet, E., & Mairessec, J. 1998. Research, Innovation And Productivity: An Econometric Analysis At The Firm Level. Economics of Innovation and new Technology, 7(2): 115-158. Criscuolo, P. 2005. On the road again: Researcher mobility inside the R&D network. Research Policy, 34(9): 1350-1365. Crossan, M. M., & Berdrow, I. 2003. Organizational learning and strategic renewal. Strategic Management Journal, 24(11): 1087-1105. Crossan, M. M., Lane, H. W., & White, R. E. 1999. An organizational learning framework: From intuition to institution. Academy of Management Review, 24(3): 522-537. Crossan, M. M., Lane, H. W., White, R. E., & Djurfeldt, L. 1995. Organizational learning: Dimensions for a theory. The International Journal of Organizational Analysis, 3(4): 337- 360. Cyert, R. M., & March, J. G. 1963. A behavioral theory of the firm. D'Este, P. 2002. The distinctive patterns of capabilities accumulation and inter‐firm heterogeneity: the case of the Spanish pharmaceutical industry. Industrial and Corporate Change, 11(4): 847-874. Daft, R. L., & Lengel, R. H. 1986. Organizational information requirements, media richness and structural design. Management Science, 32(5): 554-571. Daft, R. L., & Weick, K. E. 1984. Toward a model of organizations as interpretation systems. Academy of Management Review, 9(2): 284-295. Dahlander, L., & Gann, D. M. 2010. How open is innovation? Research policy, 39(6): 699-709. Dahlander, L., O'Mahony, S., & Gann, D. M. 2016. One foot in, one foot out: how does individuals' external search breadth affect innovation outcomes? Strategic Management Journal, 37(2): 280-302. Damanpour, F. 1991. Organizational innovation: A meta-analysis of effects of determinants and moderators. Academy of management journal, 34(3): 555-590. Danneels, E. 2008. Organizational antecedents of second‐order competences. Strategic Management Journal, 29(5): 519-543. Danneels, E. 2011. Trying to become a different type of company: dynamic capability at Smith Corona. Strategic Management Journal, 32(1): 1-31. Danneels, E., & Kleinschmidtb, E. J. 2001. Product innovativeness from the firm's perspective: its dimensions and their relation with project selection and performance. Journal of Product Innovation Management: An International Publication of the Product Development & Management Association, 18(6): 357-373. Darr, E. D., Argote, L., & Epple, D. 1995. The acquisition, transfer, and depreciation of knowledge in service organizations: Productivity in franchises. Management Science, 41(11): 1750- 1762. 187

Darroch, J., & McNaughton, R. 2003. Beyond market orientation: Knowledge management and the innovativeness of New Zealand firms. European Journal of Marketing, 37(3/4): 572- 593. Davenport, T. H., & Prusak, L. 1998. Working knowledge: How organizations manage what they know: Harvard Business Press. De Long, D. W., & Fahey, L. 2000. Diagnosing cultural barriers to knowledge management. Academy of Management Perspectives, 14(4): 113-127. Department of Industry, I., & Science. 2014. Australian innovation system report Dess, G. G., & Picken, J. C. 2001. Changing roles: Leadership in the 21st century. Organizational Dynamics, 28(3): 18-34. DeVellis, R. F. 2016. Scale development: Theory and applications: Sage publications. DIIS. 2015. The Australian Innovation System Report. In I. S. Department of Industry (Ed.). Canberra: Commonwealth of Australia. Distel, A. P. 2017. Unveiling the Microfoundations of Absorptive Capacity: A Study of Coleman’s Bathtub Model. Journal of Management: 0149206317741963. Dosi, G., In, G., Dosi, C., Freeman, R., Nelson, G., & London, N. Y. 1988. The nature of the innovative process. Drucker, P. 2002. The discipline of Innovation, Harvard Business Review, Vol. 80: 95-104. Duguet, E., & Lelarge, C. 2006. Does patenting incresse the private incentives to innovate. A microeconometric analysis. Institut National de la Statistique et des Etudes Economiques (S erie du Document du Travail du CREST. No. 2006-09). Dunning, J. H., & Lundan, S. M. 2008. Multinational enterprises and the global economy: Edward Elgar Publishing. Dutton, J. M., & Thomas, A. 1984. Treating progress functions as a managerial opportunity. Academy of Management Review, 9(2): 235-247. Dyer, J., & Singh, H. 1998. The Relational view: Cooperative strategy and sources of interorganizational competitive advantage. Academy of Management Review, 23(4): 660-679. Easterby-Smith, M., & Prieto, I. M. 2008. Dynamic capabilities and knowledge management: an integrative role for learning? British Journal of Management, 19(3): 235-249. Easterby-Smith, M., Thorpe, R., & Jackson, P. R. 2015. Management and Business Research: SAGE. Easterby‐Smith, M., & Prieto, I. M. 2008. Dynamic capabilities and knowledge management: an integrative role for learning? British Journal of Management, 19(3): 235-249. Eggers, J. P., All Experience is Not Created Equal: Learning, Adapting and Focusing in Product Portfolio Management. Strategic Management Journal (2012), 33(3): pp. 315–335. Eisenhardt, K. M. 1989. Agency theory: An assessment and review. Academy of Management Review, 14(1): 57-74. Eisenhardt, K. M., & Martin, J. A. 2000. Dynamic capabilities: what are they? Strategic Management Journal 21(10/11): 1105-1121. Eisenhardt, K. M., & Schoonhoven, C. B. 1996. Resource-based view of strategic alliance formation: Strategic and social effects in entrepreneurial firms. Organization Science, 7(2): 136-150. Ellis, H. C. (1965). The transfer of learning. Oxford, England: Macmillan. Elkins, T., & Keller, R. T. 2003. Leadership in research and development organizations: A literature review and conceptual framework. The Leadership Quarterly, 14(4): 587-606. Enkel, E., Gassmann, O., & Chesbrough, H. 2009. Open R&D and open innovation: exploring the phenomenon. R&D Management, 39(4): 311-316. Escribano, A., Fosfuri, A., & Tribó, J. A. 2009. Managing external knowledge flows: The moderating role of absorptive capacity. Research Policy, 38(1): 96-105. 188

Ettlie, J. E., Bridges, W. P., & O'keefe, R. D. 1984. Organization strategy and structural differences for radical versus incremental innovation. Management Science, 30(6): 682-695. Faems, D., De Visser, M., Andries, P., & Van Looy, B. 2010. Technology alliance portfolios and financial performance: value‐enhancing and cost‐increasing effects of open innovation. Journal of Product Innovation Management, 27(6): 785-796. Fagerberg, J. 1988 International competitiveness, Economic Journal, 98(391): 355-374. Farrell, A. M., & Rudd, J. M. 2009. Factor analysis and discriminant validity: A brief review of some practical issues. Feldman, M. S., & Pentland, B. T. 2003. Reconceptualizing organizational routines as a source of flexibility and change. Administrative Science Quarterly, 48(1): 94-118. Feldman, M. S., & Rafaeli, A. 2002. Organizational routines as sources of connections and understandings. Journal of Management Studies, 39(3): 309-331. Felin, T., & Foss, N. J. 2011. The endogenous origins of experience, routines, and organizational capabilities: the poverty of stimulus. Journal of Institutional Economics, 7(2): 231-256. Felin, T., Foss, N. J., & Ployhart, R. E. 2015. The microfoundations movement in strategy and organization theory. Academy of Management Annals, 9(1): 575-632. Felin, T., & Hesterly, W. S. 2007. The knowledge-based view, nested heterogeneity, and new value creation: Philosophical considerations on the locus of knowledge. Academy of Management Review, 32(1): 195-218. Felin, T., & Zenger, T. R. 2014. Closed or open innovation? Problem solving and the governance choice. Research Policy, 43(5): 914-925. Field, A. 2013. Discovering statistics using IBM SPSS statistics: Sage. Filippucci, C., Drudi, I., & Papalia, R. B. 1996. Testing the relevance of Tobin's approach for modelling consumption, Economic Notes - Siena: 225-248. Fiol, C. M., & Lyles, M. A. 1985. Organizational learning. Academy of Management Review, 10(4): 803-813. Flatten, T. C., Engelen, A., Zahra, S. A., & Brettel, M. 2011. A measure of absorptive capacity: Scale development and validation. European Management Journal, 29(2): 98-116. Floyd, S. W., & Lane, P. J. 2000. Strategizing throughout the organization: Managing role conflict in strategic renewal. Academy of Management Review, 25(1): 154-177. Fornell, C., & Larcker, D. F. 1981. Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research, 18(1): 39-50. Foss, N. J. 2007. The emerging knowledge governance approach: Challenges and characteristics. Organization, 14(1): 29-52. Foss, N. J., Husted, K., & Michailova, S. 2010. Governing knowledge sharing in organizations: Levels of analysis, governance mechanisms, and research directions. Journal of Management Studies, 47(3): 455-482. Foss, N. J., Laursen, K., & Pedersen, T. 2011. Linking customer interaction and innovation: The mediating role of new organizational practices. Organization Science, 22(4): 980-999. Foss, N. J., Lyngsie, J., & Zahra, S. A. 2013. The role of external knowledge sources and organizational design in the process of opportunity exploitation. Strategic Management Journal, 34(12): 1453-1471. Fredrickson, J. W. 1984. The comprehensiveness of strategic decision processes: Extension, observations, future directions. Academy of Management Journal, 27(3): 445-466. Freeman, C. 1982. The economics of industrial innovation, 2nd edn. Frances Pinter, London. Freeman, C. 1987. Technology policy and economic performance: Lessons from Japan. London: Pinter Publishers. Galunic, C., & Rodan, S. 1998. Resource recombination in the firm: Knowledge, structures and the potential for Schumpeterian innovation. Strategic Management Journal, 19 (12): 1193- 1201. 189

Gambardella, A., & Panico, C. 2014. On the management of open innovation. Research Policy, 43(5): 903-913. Garcia, R., & Calantone, R. 2002. A critical look at technological innovation typology and innovativeness terminology: a literature review. Journal of Product Innovation Management, 19(2): 110-132. Garud, R., Dunbar, R. L., & Bartel, C. A. 2011. Dealing with unusual experiences: A narrative perspective on organizational learning. Organization Science, 22(3): 587-601. Gassmann, O. 2006. Opening up the innovation process: towards an agenda. R&D Management, 36(3): 223-228. Gasson, S. 2009. Handbook of Research on Contemporary Theoretical Models in Information Systems. Gavetti, G., & Levinthal, D. 2000. Looking forward and looking backward: Cognitive and experiential search. Administrative Science Quarterly, 45(1): 113-137. George, G. 2005. Learning to be capable: patenting and licensing at the Wisconsin Alumni Research Foundation 1925–2002. Industrial and Corporate Change, 14(1): 119-151. Gertler, M. S. 2003. Tacit knowledge and the economic geography of context, or the undefinable tacitness of being (there). Journal of Economic Geography, 3(1): 75-99. Ghoshal, S., & Bartlett, C. A. 1994. Linking organizational context and managerial action: The dimensions of quality of management. Strategic Management Journal, 15(S2): 91-112. Gibson, C. B., & Birkinshaw, J. 2004. The antecedents, consequences, and mediating role of organizational ambidexterity. Academy of Management Journal, 47(2): 209-226. Gold, A. H., Malhotra, A., & Segars, A. H. 2001. Knowledge management: An organizational capabilities perspective. Journal of Management Information Systems, 18(1): 185-214. Gosain, S. 2007. Mobilizing software expertise in personal knowledge exchanges. The Journal of Strategic Information Systems, 16(3): 254-277. Grant, R. M. 1996. Prospering in dynamically-competitive environments: Organizational capability as knowledge integration. Organization Science, 7(4): 375-387. Grant, R. M., & Baden-Fuller, C. 2004. A knowledge accessing theory of strategic alliances. Journal of Management Studies, 41(1): 61-84. Greene, W. H. 2000. Econometric analysis (International edition). Greene, W. H. 2003. Econometric analysis: Pearson Education India. Grimpe, C., & Sofka, W. 2009. Search patterns and absorptive capacity: Low-and high-technology sectors in European countries. Research Policy, 38(3): 495-506. Grindley, P. C., & Teece, D. J. 1997. Managing intellectual capital: licensing and cross-licensing in semiconductors and electronics. California Management Review, 39(2): 8-41. Guba, E. G. (Ed.). 1990. The paradigm dialogue. Newbury Park: CA: Sage. Gupta, A. K., & Govindarajan, V. 2000. Knowledge flows within multinational corporations. Strategic Management Journal, 21(4): 473-496. Gupta, A. K., Smith, K. G., & Shalley, C. E. 2006. The interplay between exploration and exploitation. Academy of Management Journal, 49(4): 693-706. Hackman, J. R., & Wageman, R. 2005. A theory of team coaching. Academy of Management Review, 30(2): 269-287. Hage, J. T. 1999. Organizational innovation and organizational change. Annual Review of Sociology, 25(1): 597-622. Hair, J. F. 2009. Multivariate data analysis. Hair Jr, J., Black, W., Babin, B., Anderson, R., & Tatham, R. 2010. SEM: An introduction. Hanmer, M. J., & Ozan Kalkan, K. 2013. Behind the curve: Clarifying the best approach to calculating predicted probabilities and marginal effects from limited dependent variable models. American Journal of Political Science, 57(1): 263-277.

190

Hannan, M. T., & Freeman, J. 1984. Structural inertia and organizational change. American Sociological Review, 49(2): 149-164. Harlow, H. 2008. The effect of tacit knowledge on firm performance. Journal of Knowledge Management, 12(1): 148-163. Harris, S. G. 1994. Organizational culture and individual sensemaking: A schema-based perspective. Organization Science, 5(3): 309-321. Harwel, R. M. 2010. Research Design in Qualitative/Quantitative/Mixed Methods. In SAGE (Ed.): University of Minnesota. Hayduk, L., Cummings, G., Boadu, K., Pazderka-Robinson, H., & Boulianne, S. 2007. Testing! testing! one, two, three–Testing the theory in structural equation models! Personality and Individual Differences, 42(5): 841-850. He, Z.-L., & Wong, P.-K. 2004. Exploration vs. exploitation: An empirical test of the ambidexterity hypothesis. Organization Science, 15(4): 481-494. Hedberg, B. 1981. How organizations learn and unlearn: Oxford University Press Helfat, C., Fynkelstein, S., Mitchell, W., Singh, H., Teece, D., & Winter, S. 2007. Dynamic capabilities: strategic change in organizations: Blackwell, Oxford, UK. Henderson, R. C., & Clark, K. B. 1990. Architectural innovation: The reconfiguration of existing product technologies and the failure of established firms. Administrative Science Quarterly, 35(1): 9-30. Hess, A. M., & Rothaermel, F. T. 2011. When are assets complementary? Star scientists, strategic alliances, and innovation in the pharmaceutical industry. Strategic Management Journal, 32(8): 895-909. Hill, C. W., & Rothaermel, F. T. 2003. The performance of incumbent firms in the face of radical technological innovation. Academy of Management Review, 28(2): 257-274. Hitt, M. A., Hoskisson, R. E., & Kim, H. 1997. International diversification: Effects on innovation and firm performance in product-diversified firms. Academy of Management Journal, 40(4): 767-798. Hoegl, M., & Schulze, A. 2005a. How to Support Knowledge Creation in New Product Development: An Investigation of Knowledge Management Methods. European Management Journal, 23(3): 263-273. Hoegl, M., & Schulze, A. 2005b. How to Support Knowledge Creation in New Product Development:: An Investigation of Knowledge Management Methods. European Management Journal, 23(3): 263-273. Hogan, S. J., Soutar, G. N., McColl-Kennedy, J. R., & Sweeney, J. C. 2011. Reconceptualizing professional service firm innovation capability: Scale development. Industrial Marketing Management, 40(8): 1264-1273. Holmqvist, M. 2004. Experiential learning processes of exploitation and exploration within and between organizations: An empirical study of product development. Organization science, 15(1): 70-81. Hoskisson, R. E., & Hitt, M. A. 1988. Strategic control systems and relative R&D investment in large multiproduct firms. Strategic Management Journal, 9(6): 605-621. Hu, L. T., & Bentler, P. M. 1999. Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6(1): 1-55. Huang, F., & Rice, J. 2009. The role of absorptive capacity in facilitating" Open innovation" outcomes: A study of Australian SMEs in the manufacturing sector. International Journal of Innovation Management, 13(02): 201-220. Huang, J.-W., & Li, Y.-H. 2012. Slack resources in team learning and project performance. Journal of Business Research, 65(3): 381-388.

191

Huber, G. P. 1991. Organizational learning: The contributing processes and the literatures. Organization Science, 2(1): 88-115. Inkpen, A. C. 1996. Creating knowledge through collaboration. California Management Review, 39(1): 123-140. Inkpen, A. C., & Crossan, M. M. 1995. Believing is seeing: Joint ventures and organization learning. Journal of Management Studies, 32(5): 595-618. Jansen, J. J., Van Den Bosch, F. A., & Volberda, H. W. 2005a. Managing potential and realized absorptive capacity: how do organizational antecedents matter? Academy of Management Journal, 48(6): 999-1015. Jansen, J. J., Volberda, H. W., & Van Den Bosch, F. A. 2005b. Exploratory innovation, exploitative innovation, and ambidexterity: The impact of environmental and organizational antecedents. Schmalenbach Business Review, 57(4): pp.351-363. Jantunen, A. 2005. Knowledge-processing capabilities and innovative performance: an empirical study. European Journal of Innovation Management, 8(3): 336-349. Jaworski, B. J., & Kohli, A. K. 1996. Market orientation: review, refinement, and roadmap. Journal of Market-Focused Management, 1(2): 119-135. Jelinek, M. 1979. Institutionalizing innovation: A study of organizational learning systems: Praeger Publishers. Jeppesen, L. B., & Lakhani, K. R. 2010. Marginality and problem-solving effectiveness in broadcast search. Organization Science, 21(5): 1016-1033. Jiménez-Barrionuevo, M. M., García-Morales, V. J., & Molina, L. M. 2011. Validation of an instrument to measure absorptive capacity. Technovation, 31(5): 190-202. Johanson, J., & Vahlne, J.-E. 1977. The internationalization process of the firm - a model of knowledge development and increasing foreign market commitments. Journal of International Business Studies, 8(1): 23-32. Jones, M. V., & Coviello, N. E. 2005. Internationalisation: conceptualising an entrepreneurial process of behaviour in time. Journal of International Business Studies, 36(3): 284-303. Jöreskog, K., & Sörbom, D. 2004. Interactive Lisrel. Jöreskog, K. G., & Sörbom, D. 1993. LISREL 8: Structural equation modeling with the SIMPLIS command language: Scientific Software International. Jöreskog, K. G., & Wold, H. O. 1982. Systems under indirect observation: Causality, structure, prediction: North Holland. Kanigel, R. 1997. The One Best Way: Frederick W. Taylor and the Enigma of Efficiency: 376-378. Kanter, R. M. 1983. The change masters: Binnovation and entrepreneturship in the American corporation: Touchstone Book. Karanikolas, E. 2013. Industry Policy in an open economy. In Parliament of Australia (Ed.): Parliamentary Library Briefing Book: Key Issues for the 44th Parliament. Katila, R., & Ahuja, G. 2002. Something old, something new: A longitudinal study of search behavior and new product introduction. Academy of Management Journal, 45(6): 1183- 1194. Katz, R., & Allen, T. J. 1982. Investigating the Not Invented Here (NIH) syndrome: A look at the performance, tenure, and communication patterns of 50 R & D Project Groups. R&D Management, 12(1): 7-20. Kenny, D. A. 2014. Measuring model fit. Kilmann, R., & Mitroff, I. 1976. On organization stories: An approach to the design and analysis of organizations through myths and stories. Kim, Y. A., Akbar, H., Tzokas, N., & Al-Dajani, H. 2014. Systems thinking and absorptive capacity in high-tech in high-tech small and medium-sized enterprises from South Korea. International Small Business Journal, 32(8): 876-896.

192

Kindström, D., Kowalkowski, C., & Sandberg, E. 2013. Enabling service innovation: A dynamic capabilities approach. Journal of Business Research, 66(8): 1063-1073. King, G., Keohane, R. O., & Verba, S. 1994. Designing social inquiry: Scientific inference in qualitative research: Princeton university press. Kline, R. B. 2015. Principles and practice of structural equation modeling: Guilford publications. Knight, G. A., & Cavusgil, S. T. 2004. Innovation, organizational capabilities, and the born-global firm. Journal of International Business Studies, 35(2): 124-141. Knoben, J., & Oerlemans, L. 2010. The importance of external knowledge sources for the newness of innovations of South African firms. International Journal of Innovation and Regional Development, 2(3): 165-181. Knorr-Cetina, K. D. 1981. The micro-sociological challenge of macro-sociology: towards a reconstruction of social theory and methodology. Koestler, A. 1966. The Act of Creation London: Hutchinson. Kogut, B., & Zander, U. 1992. Knowledge of the firm, combinative capabilities, and the replication of technology. Organization Science, 3(3): 383-397. Kogut, B., & Zander, U. 1996. What firms do? Coordination, identity, and learning. Organization Science, 7(5): 502-518. Kohli, A. K., & Jaworski, B. J. 1990. Market orientation: the construct, research propositions, and managerial implications. The Journal of Marketing: 1-18. Kong, E., & Farrell, M. 2012. Facilitating knowledge and learning capabilities through neuro- linguistic programming. International Journal of Learning, 18(3). Kostopoulos, K., Papalexandris, A., Papachroni, M., & Ioannou, G. 2011. Absorptive capacity, innovation, and financial performance. Journal of Business Research, 64(12): 1335-1343. Koza, M. P., & Lewin, A. Y. 1998. The co-evolution of strategic alliances. Organization Science, 9(3): 255-264. Kuhn, T. S. 1970. Logic of discovery or psychology of research. Kumar, V., Mudambi, R., & Gray, S. 2013. Internationalization, Innovation and Institutions: The 3 I's underpinning the competitiveness of emerging market firms. Journal of International Management, 19(3): 203-206. Laaksonen, O., & Peltoniemi, M. 2018. The essence of dynamic capabilities and their measurement. International Journal of Management Reviews, 20(2): 184-205. Lane, P. J., Koka, B. R., & Pathak, S. 2006. The reification of absorptive capacity: A critical review and rejuvenation of the construct. Academy of Management Review, 31(4): 833-863. Lane, P. J., & Lubatkin, M. 1998. Relative absorptive capacity and interorganizational learning. Strategic Management Journal, 19(5): 461-477. Lane, P. J., Salk, J. E., & Lyles, M. A. 2001. Absorptive capacity, learning, and performance in international joint ventures. Strategic Management Journal, 22(12): 1139-1161. Lankford, S. V., Buxton, B. P., Hetzler, R., & Little, J. R. 1995. Response bias and wave analysis of mailed questionnaires in tourism impact assessments. Journal of Travel Research, 33(4): 8-13. Lau, A. K., Tang, E., & Yam, R. C. 2010. Effects of supplier and customer integration on product innovation and performance: empirical evidence in Hong Kong manufacturers. Journal of Product Innovation Management, 27(5): 761-777. Laursen, K., & Salter, A. 2006. Open for innovation: the role of openness in explaining innovation performance among UK manufacturing firms. Strategic Management Journal, 27(2): 131- 150. Laursen, K., & Salter, A. J. 2014. The paradox of openness: Appropriability, external search and collaboration. Research Policy, 43(5): 867-878. Lavie, D., & Rosenkopf, L. 2006. Balancing exploration and exploitation in alliance formation. Academy of Management Journal, 49(4): 797-818. 193

Lawson, B., & Samson, D. 2001. Developing innovation capability in organisations: a dynamic capabilities approach. International Journal of Innovation Management, 5(03): 377-400. Lehrer, M. 2000. The organizational choice between evolutionary and revolutionary capability regimes: Theory and evidence from European air transport. Industrial and Corporate Change, 9(3): 489-520. Leiponen, A., & Helfat, C. E. 2010. Innovation objectives, knowledge sources, and the benefits of breadth. Strategic Management Journal, 31(2): 224-236. Lenox, M., & King, A. 2004. Prospects for developing absorptive capacity through internal information provision. Strategic Management Journal, 25(4): 331-345. Leonard‐Barton, D. 1992. Core capabilities and core rigidities: A paradox in managing new product development. Strategic Management Journal, 13(S1): 111-125. Leonard, F., & Swap, W. 1999. When sparks fly: Igniting creativity in groups. Boston, MA: Harvard Business School Press. Lessard, D. R., & Zaheer, S. 1996. Breaking the silos: Distributed knowledge and strategic responses to volatile exchange rates. Strategic Management Journal, 17(7): 513-533. Levinthal, D., & March, J. G. 1981. A model of adaptive organizational search. Journal of Economic Behavior & Organization, 2(4): 307-333. Levinthal, D. A., & March, J. G. 1993. The myopia of learning. Strategic Management Journal, 14(S2): 95-112. Levitt, B., & March, J. G. 1988. Organizational learning. Annual Review of Sociology, 14(1): 319- 338. Lewin, A. Y., & Massini, S. 2004. Knowledge creation and organizational capabilities of innovating and imitating firms, Organizations as Knowledge Systems: 209-237: Springer. Lewin, A. Y., Massini, S., & Peeters, C. 2011. Microfoundations of internal and external absorptive capacity routines. Organization Science, 22(1): 81-98. Lewin, A. Y., & Volberda, H. W. 1999. Prolegomena on coevolution: A framework for research on strategy and new organizational forms. Organization Science, 10(5): 519-534.

Lichtenthaler, U. 2011. Open innovation: Past research, current debates, and future directions. The Academy of Management Perspectives, 25(1): 75-93. Lichtenthaler, U., & Ernst, H. 2006. Attitudes to externally organising knowledge management tasks: a review, reconsideration and extension of the NIH syndrome. R&D Management, 36(4): 367-386. Lichtenthaler, U., Ernst, H., & Lichtenthaler, E. 2005. Desorptive capacity: A capabilitiy-based perspective on external knowledge exploitation. Paper presented at the Annual Meeting of the Academy of Management, Honolulu. Lichtenthaler, U., & Lichtenthaler, E. 2009. A capability‐based framework for open innovation: Complementing absorptive capacity. Journal of Management Studies, 46(8): 1315-1338. Lichtenthaler, U., Lichtenthaler, E. R. V., & Ernst, H. 2004. Desorptive Capacity: A New Perspective on the External Commercialization of Knowledge: Wiss. Hochsch. für Unternehmensführung. Lincoln, Y. S., & Guba, E. G. 2000. Paradigmatic controversies, contradictions, and emerging influences. In N. Denzin, & Y. Lincoln (Eds.), Handbook of Qualitative Research, 2nd ed.: 163-188. Thousands Oaks CA: Sage. Lippman, S. A., & Rumelt, R. P. 1982. Uncertain imitability: An analysis of interfirm differences in efficiency under competition. The Bell Journal of Economics, 13(2): 418-438. Love, J. H., Roper, S., & Bryson, J. R. 2011. Openness, knowledge, innovation and growth in UK business services. Research Policy, 40(10): 1438-1452. Love, J. H., Roper, S., & Vahter, P. 2014. Learning from openness: The dynamics of breadth in external innovation linkages. Strategic Management Journal, 35(11): 1703-1716. 194

Luo, L., Kannan, P., & Ratchford, B. T. 2008. Incorporating subjective characteristics in product design and evaluations. Journal of Marketing Research, 45(2): 182-194. Luo, Y., & Tung, R. L. 2007. International expansion of emerging market enterprises: A springboard perspective: Springer. Lüttgens, D., Pollok, P., Antons, D., & Piller, F. 2014. Wisdom of the crowd and capabilities of a few: internal success factors of crowdsourcing for innovation. Journal of Business Economics, 84(3): 339-374. Macpherson, A., Jones, O., & Zhang, M. 2004. Evolution or revolution? Dynamic capabilities in a knowledge‐dependent firm. R&D Management, 34(2): 161-177. Maddala, G. 1983. Methods of estimation for models of markets with bounded price variation. International Economic Review, 24(2): 361-378. Madhavan, R., & Grover, R. 1998. From embedded knowledge to embodied knowledge: new product development as knowledge management. The Journal of Marketing, 62(4): 1-12. Mairesse, J., & Mohnen, P. 2010. Using innovation surveys for econometric analysis, Handbook of the Economics of Innovation, Vol. 2: 1129-1155: Elsevier. Makadok, R. 2001. Toward a synthesis of the resource‐based and dynamic‐capability views of rent creation. Strategic Management Journal, 22(5): 387-401. March, J. G. 1991. Exploration and exploitation in organizational learning. Organization Science, 2(1): 71-87. Markides, C., & Charitou, C. D. 2004. Competing with dual business models: A contingency approach. The Academy of Management Executive, 18(3): 22-36. Marsick, V. J., & Watkins, K. E. 1996. Adult educators and the challenge of the learning organization. Adult Learning, 7(4): 18-20. Martín-de-Castro, G., Delgado-Verde, M., López-Sáez, P., & Navas-López, J. E. 2011. Towards ‘an intellectual capital-based view of the firm’: origins and nature. Journal of Business Ethics, 98(4): 649-662. Martin, M. J. C. 1994. Managing Innovation and Entrepreneurship in Technology-based Firms. New York: Wiley. Massini, S., Lewin, A. Y., & Greve, H. R. 2005. Innovators and imitators: Organizational reference groups and adoption of organizational routines. Research Policy, 34(10): 1550-1569. Mathews, J. A. 2006. Dragon multinationals: New players in 21st century globalization. Asia Pacific journal of management, 23(1): 5-27. Maxwell, J. A., & Mittapalli, K. 2010. Realism and mixed methods. In A. Tashakkori, & C. Teddlie (Eds.), Handbook of Mixed Methods in Social and Behavioral Research, 2nd ed.: 145- 167: SAGE. McDermott, C. M., & O'Connor, G. C. 2002. Managing radical innovation: an overview of emergent strategy issues. Journal of Product Innovation Management: An International Publication of the Product Development & Management Association, 19(6): 424-438. McDonald, R. P., & Marsh, H. W. 1990. Choosing a multivariate model: Noncentrality and goodness of fit. Psychological Bulletin, 107(2): 247. McInerney, C. 2002. Knowledge management and the dynamic nature of knowledge. Journal of the Association for Information Science and Technology, 53(12): 1009-1018. McKinley, W., Latham, S., & Braun, M. 2014. Organizational decline and innovation: Turnarounds and downward spirals. Academy of Management Review, 39(1 ): 88-110. Minbaeva, D., Pedersen, T., Björkman, I., Fey, C. F., & Park, H. J. 2003. MNC knowledge transfer, subsidiary absorptive capacity, and HRM. Journal of International Business Studies, 34(6): 586-599. Mohnen, P., Mairesse, J., & Dagenais, M. 2006. Innovativity: A comparison across seven European countries. Economics of Innovation and New Technology, 15(4-5): 391-413.

195

Mowery, D. C., Oxley, J. E., & Silverman, B. S. 1996. Strategic alliances and interfirm knowledge transfer. Strategic Management Journal, 17(S2): 77-91. Mumford, M. D., Scott, G. M., Gaddis, B., & Strange, J. M. 2002. Leading creative people: Orchestrating expertise and relationships. The Leadership Quarterly, 13(6): 705-750. Nadler, J., Thompson, L., & Boven, L. V. 2003. Learning negotiation skills: Four models of knowledge creation and transfer. Management Science, 49(4): 529-540. Nahapiet, J., & Ghoshal, S. 1998. Social capital, intellectual capital, and the organizational advantage. Academy of Management Review, 23(2): 242-266. Narteh, B. 2008. Knowledge transfer in developed-developing country interfirm collaborations: a conceptual framework. Journal of Knowledge Management, 12(1): 78-91. Neely, A., Filippini, R., Forza, C., Vinelli, A., & Hii, J. 2001. A framework for analysing business performance, firm innovation and related contextual factors: perceptions of managers and policy makers in two European regions. Integrated Manufacturing Systems, 12(2): 114-124. Nelson, R., & Winter, G. 1982. An evolutionary theory of economic change. Cambridge: Harvard Business School Press. Nerkar, A., & Paruchuri, S. 2005. Evolution of R&D capabilities: The role of knowledge networks within a firm. Management Science, 51(5): 771-785. Niiniluoto, I. 1991. Realism, Relativism, and Constructivism. Scientific Knowledge and Scientific Practice, 89(1): 135-162. Nonaka, I. 1991. The Knowledge-Creating Company. Harvard Business Review. Nonaka, I. 1994. A dynamic theory of organizational knowledge creation. Organization Science, 5(1): 14-37. Nonaka, I., & Takeuchi, H. 1995. The knowledge-creating company: How Japanese companies create the dynamics of innovation: Oxford university press. Nonaka, I., & Toyama, R. 2003. The knowledge-creating theory revisited: knowledge creation as a synthesizing process. Knowledge Management Research & Practice, 1(1): 2-10. Nonaka, I., Toyama, R., & Konno, N. 2000. SECI, Ba and leadership: a unified model of dynamic knowledge creation. Long range planning, 33(1): 5-34. Nonaka, I., & Von Krogh, G. 2009. Perspective—Tacit knowledge and knowledge conversion: Controversy and advancement in organizational knowledge creation theory. Organization Science, 20(3): 635-652. Nonaka, I., Von Krogh, G., & Voelpel, S. 2006. Organizational knowledge creation theory: Evolutionary paths and future advances. Organization Studies, 27(8): 1179-1208. Noriega, J., & Blair, E. 2008. Advertising to bilinguals: Does the language of advertising influence the nature of thoughts? Journal of Marketing, 72(5): 69-83. O’Reilly, C. A., & Tushman, M. L. 2008. Ambidexterity as a dynamic capability: Resolving the innovator's dilemma. Research in Organizational Behavior, 28: 185-206. O’Sullivan, D., & Dooley, L. 2009. Applying innovation. Thousand Oakes, CA: SAGE Publications. Ocasio, W. 1997. Towards an attention-based view of the firm. Strategic Management Journal: 187-206. Ocasio, W. 2011. Attention to attention. Organization Science, 22(5): 1286-1296. OECD. 2005. Oslo Manual: Guidelines for Collecting and Interpreting Innovation Data, Third Edition. OECD. 2014. Main science and technology indicators. Orth, U. R., & Malkewitz, K. 2008. Holistic package design and consumer brand impressions. Journal of Marketing, 72(3): 64-81. Othman Idrissia, M., Amaraa, N., & Landrya, R. 2012. SMEs’ degree of openness: the case of manufacturing industries. Journal of Technology Management & Innovation, 7(1): 186- 210. 196

Ozalp, H., & Cennamo, C. 2017. Platform Architecture, Multihoming and Complement Quality. Paper presented at the Academy of Management Proceedings. Ozalp, H., & Kretschmer, T. 2018. Follow the crowd or follow the trailblazer? The differential role of firm experience in product entry decisions in the US video game industry. Journal of Management Studies. Paarup Nielsen, A. 2006. Understanding dynamic capabilities through knowledge management. Journal of Knowledge Management, 10(4): 59-71. Pandza, K., Horsburgh, S., Gorton, K., & Polajnar, A. 2003. A real options approach to managing resources and capabilities. International Journal of Operations & Production Management, 23(9): 1010-1032. Papalia, R. B., & Di Iorio, F. 2001. Alternative error term specifications in the log-Tobit model1, Advances in Classification and Data Analysis: 185-192: Springer. Parry, S. 2017. Fit Statistics commonly reported for CFA and SEM. Cornell Statistical Consulting Unit: Cornell University. Patel, P. C., Kohtamäki, M., Parida, V., & Wincent, J. 2015. Entrepreneurial orientation‐as‐ experimentation and firm performance: The enabling role of absorptive capacity. Strategic Management Journal, 36(11): 1739-1749. Patel, P. C., & Van der Have, R. P. 2010. Enhancing innovation performance through exploiting complementarity in search breadth and depth. Frontiers of Entrepreneurship Research, 30(9): 1. Patterson, W., & Ambrosini, V. 2015. Configuring absorptive capacity as a key process for research intensive firms. Technovation, 36-37: 77-89. Peteraf, M. A. 1993. The cornerstones of competitive advantage: a resource‐based view. Strategic Management Journal, 14(3): 179-191. Piller, F. T., & Walcher, D. 2006. Toolkits for idea competitions: a novel method to integrate users in new product development. R&D Management, 36(3): 307-318. Pinsonneault, A., & Kraemer, K. L. 1993. Research Methodology in Management IS: An Assessment. Journal of Management Information Systems, 10(2): 75-105. Piperopoulos, P., Wu, J., & Wang, C. 2018. Outward FDI, location choices and innovation performance of emerging market enterprises. Research Policy, 47(1): 232-240. Podsakoff, P. M., MacKenzie, S. B., Lee, J.-Y., & Podsakoff, N. P. 2003. Common method biases in behavioral research: a critical review of the literature and recommended remedies. Journal of Applied Psychology, 88(5): 879. Poetz, M. K., & Schreier, M. 2012. The value of crowdsourcing: can users really compete with professionals in generating new product ideas? Journal of Product Innovation Management, 29(2): 245-256. Porras, J. I., & Collins, J. C. 1997. Built to last: HarperBusiness. Potts, J. 2000. The new evolutionary microeconomics. Priem, R. L., & Butler, J. E. 2001. Tautology in the resource-based view and the implications of externally determined resource value: Further comments. Academy of Management Review, 26(1): 57-66. Putnam, H. A. 1990. Defense of Internal Realism. In J. Conant (Ed.), Realism with a Human Face: 30-42. Cambridge: Harvard University Press:. Quintane, E., Casselman, R. M., Reiche, B. S., & Nylund, P. 2011. Innovation as a knowledge-based outcome. Journal of Knowledge Management, 15(6): 928-947. Radas, S., & Božić, L. 2009. The antecedents of SME innovativeness in an emerging transition economy. Technovation, 29(6): 438-450. Raisch, S., Birkinshaw, J., Probst, G., & Tushman, M. L. 2009. Organizational ambidexterity: Balancing exploitation and exploration for sustained performance. Organization Science, 20(4): 685-695. 197

Rakthin, S., Calantone, R. J., & Wang, J. F. 2016. Managing market intelligence: The comparative role of absorptive capacity and market orientation. Journal of Business Research, 69(12): 5569-5577. Razali, N. M., & Wah, Y. B. 2011. Power comparisons of shapiro-wilk, kolmogorov-smirnov, lilliefors and anderson-darling tests. Rerup, C., & Feldman, M. S. 2011. Routines as a source of change in organizational schemata: The role of trial-and-error learning. Academy of Management Journal, 54(3): 577-610. Rezaei-Zadeh, M., & Darwish, T. K. 2016. Antecedents of absorptive capacity: a new model for developing learning processes. The Learning Organization, 23(1): 77-91. Roberts, E. B. 1988. What we’ve learned: Managing invention and innovation. Research Technology Management, 31(1): 11-29. Robey, D., Ross, J. W., & Boudreau, M.-C. 2002. Learning to implement enterprise systems: An exploratory study of the dialectics of change. Journal of Management Information Systems, 19(1): 17-46. Rogelberg, S. G., & Stanton, J. M. 2007. Introduction: Understanding and dealing with organizational survey nonresponse: Sage Publications Sage CA: Los Angeles, CA. Rogoff, B. 1990. Apprenticeship in thinking: Cognitive development in social context: Oxford university press. Rosenkopf, L., & Nerkar, A. 2001. Beyond local search: boundary‐spanning, exploration, and impact in the optical disk industry. Strategic Management Journal, 22(4): 287-306. Rotemberg, J. J., & Saloner, G. 2000. Visionaries, managers, and strategic direction. RAND Journal of Economics, 31(4): 693-716. Rothaermel, F. T., & Alexandre, M. T. 2009. Ambidexterity in technology sourcing: The moderating role of absorptive capacity. Organization Science, 20(4): 759-780. Rothaermel, F. T., & Deeds, D. L. 2004. Exploration and exploitation alliances in biotechnology: A system of new product development. Strategic Management Journal, 25(3): 201-221. Rothaermel, F. T., & York, N. Y. 2013. Strategic management: Concepts & cases. New McGraw- Hill/Irwin. Salvato, C. 2003. The Role of Micro‐Strategies in the Engineering of Firm Evolution. Journal of Management Studies, 40(1): 83-108. Samson, D., & Gloet, M. 2014. Innovation capability in Australian manufacturing organisations: an exploratory study. International Journal of Production Research, 52(21): 6448-6466. Schilke, O. 2014. On the contingent value of dynamic capabilities for competitive advantage: The nonlinear moderating effect of environmental dynamism. Strategic Management Journal, 35(2): 179-203. Schreiber, J. B., Nora, A., Stage, F. K., Barlow, E. A., & King, J. 2006. Reporting Structural Equation Modeling and Confirmatory Factor Analysis Results: A Review. The Journal of Educational Research, 99(6): 323-338. Schroll, A., & Mild, A. 2011. Determinants of open innovation: an empirical study on organisational, market, and human drivers of open innovation adoption across Europe. International Journal of Innovation and Regional Development, 3(5): 465-485. Schumpeter, J. A. 1934. The Theory of Economic Development: Harvard University Press. Schumpeter, J. A. 1950. Capitalism, Socialism and Democracy (3rd edition). London: Allen and Unwin. Scotland, J. 2012. Exploring the Philosophical Underpinnings of Research: Relating Ontology and Epistemology to the Methodology and Methods of the Scientific, Interpretive, and Critical Research Paradigms. English Language Teaching, 5(9): 9-16. Selznick, P. 1957. Leadership in administration: A sociological interpretation. Berkeley: Quid Pro Books.

198

Shapiro, S. S., & Wilk, M. B. 1965. An analysis of variance test for normality (complete samples). Biometrika, 52(3/4): 591-611. Sidhu, J. S., Commandeur, H. R., & Volberda, H. W. 2007. The multifaceted nature of exploration and exploitation: Value of supply, demand, and spatial search for innovation. Organization Science, 18(1): 20-38. Siemieniuch, C., & Sinclair, M. 2004. CLEVER: a process framework for knowledge lifecycle management. International Journal of Operations & Production Management, 24(11): 1104-1125. Simon, H. A. 1969. The sciences of the artiﬁcial. Simonin, B. L. 1999. Ambiguity and the process of knowledge transfer in strategic alliances. Strategic Management Journal, 20(7): 595-623. Simpson, P. M., Siguaw, J. A., & Enz, C. A. 2006. Innovation orientation outcomes: The good and the bad. Journal of Business Research, 59(10): 1133-1141. Soete, L., & Freeman, C. 1997. The economics of industrial innovation. Spender, J. C. 1996. Making knowledge the basis of a dynamic theory of the firm. Strategic Management Journal, 17(S2): 45-62. Spithoven, A., Vanhaverbeke, W., & Roijakkers, N. 2013. Open innovation practices in SMEs and large enterprises. Small Business Economics, 41(3): 537-562. Starbuck, W. H. 1992. Learning by knowledge‐intensive firms. Journal of Management Studies, 29(6): 713-740. Stark, J. 2015. Product lifecycle management, Product Lifecycle Management: 1-29: Springer. Stauss, B., den Hertog, P., van der Aa, W., & de Jong, M. W. 2010. Capabilities for managing service innovation: towards a conceptual framework. Journal of Service Management, 21(4): 490-514. Sterlacchini, A. 1999. Do innovative activities matter to small firms in non-R&D-intensive industries? An application to export performance. Research Policy, 28(8): 819-832. Stuart, T. E., & Podolny, J. M. 1996. Local search and the evolution of technological capabilities. Strategic Management Journal, 17(S1): 21-38. Subramaniam, M., & Youndt, M. A. 2005. The influence of intellectual capital on the types of innovative capabilities. Academy of Management Journal, 48(3): 450-463. Sun, P. Y. T., & Anderson, M. H. 2010. An Examination of the Relationship Between Absorptive Capacity and Organizational Learning, and a Proposed Integration. International Journal of Management Reviews, 12(2): 130-150. Szulanski, G. 1996. Exploring internal stickiness: Impediments to the transfer of best practice within the firm. Strategic Management Journal, 17(S2): 27-43. Szulanski, G., & Jensen, R. J. 2004. Overcoming stickiness: An empirical investigation of the role of the template in the replication of organizational routines. Managerial and Decision Economics, 25(6/7): 347-363. Tabachnick, B. G., & Fidell, L. S. 2001. Using multivariate statistics. Teece, D., & Pisano, G. 1994. The dynamic capabilities of firms: an introduction. Industrial and Corporate Change, 3(3): 537-556. Teece, D. J. 1986. Profiting from technological innovation: Implications for integration, collaboration, licensing and public policy. Research Policy, 15(6): 285-305. Teece, D. J. 2007. Explicating dynamic capabilities: the nature and microfoundations of (sustainable) enterprise performance. Strategic Management Journal, 28(13): 1319- 1350. Teece, D. J. 2016. Dynamic capabilities and entrepreneurial management in large organizations: Toward a theory of the (entrepreneurial) firm. European Economic Review, 86: 202-216.

199

Teece, D. J., Peteraf, M. A., & Leih, S. 2016. Dynamic Capabilities and Organizational Agility: Risk, Uncertainty and Entrepreneurial Management in the Innovation Economy. California Management Review, 58(4): 13-35. Teece, D. J., Pisano, G., & Shuen, A. 1997. Dynamic capabilities and strategic management. Strategic Management Journal, 18(7): 509–533. Teng, B. 2007. Corporate Entrepreneurship Activities through Strategic Alliances: A Resource‐ Based Approach toward Competitive Advantage. Journal of Management Studies, 44(1): 119-142. Terwiesch, C., & Xu, Y. 2008. Innovation contests, open innovation, and multiagent problem solving. Management Science, 54(9): 1529-1543. Tobin, J. 1958. Estimation of relationships for limited dependent variables. Econometrica: Journal of the Econometric Society, 26(1): 24-36. Todorova, G., & Durisin, B. 2007. Absorptive capacity: Valuing a reconceptualization. Academy of Management Review, 32(3): 774-786. Totterdell, P., Leach, D., Birdi, K., Clegg, C., & Wall, T. 2002. An investigation of the contents and consequences of major organizational innovations. International Journal of Innovation Management, 6(04): 343-368. Tripsas, M., & Gavetti, G. 2000. Capabilities, cognition, and inertia: Evidence from digital imaging. Strategic Management Journal, 21(10/11): 1147-1161. Trott, P., & Hartmann, D. 2009. Why'open innovation'is old wine in new bottles. International Journal of Innovation Management, 13(04): 715-736. Tsai, K.-H. 2009. Collaborative networks and product innovation performance: Toward a contingency perspective. Research policy, 38(5): 765-778. Tsai, W. 2001. Knowledge transfer in intraorganizational networks: Effects of network position and absorptive capacity on business unit innovation and performance. Academy of Management Journal, 44(5): 996-1004. Tushman, M. L., & Anderson, P. 1986. Technological discontinuities and organizational environments. Administrative Science Quarterly, 31(3): 439-465. Tushman, M. L., & O'Reilly, C. A. 1996. The ambidextrous organizations: Managing evolutionary and revolutionary change. California Management Review, 38(4): 8-30. Ullman, J. B., & Newcomb, M. D. 1998. Eager, Reluctant, and Nonresponders to a Mailed Longitudinal Survey: Attitudinal and Substance Use Characteristics Differentiate Respondents1. Journal of Applied Social Psychology, 28(4): 357-375. Van de Vrande, V., De Jong, J. P., Vanhaverbeke, W., & De Rochemont, M. 2009. Open innovation in SMEs: Trends, motives and management challenges. Technovation, 29(6): 423-437. Vargo, S. L., Maglio, P. P., & Akaka, M. A. 2008. On value and value co-creation: A service systems and service logic perspective. European Management Journal, 26(3): 145-152. Venkitachalam, K., & Busch, P. 2012. Tacit knowledge: review and possible research directions. Journal of Knowledge Management, 16(2): 357-372. Vera, D., Crossan, M., & Apaydin, M. 2011. A framework for integrating organizational learning, knowledge, capabilities, and absorptive capacity. Verreynne, M.-L., Hine, D., Coote, L., & Parker, R. 2016. Building a scale for dynamic learning capabilities: The role of resources, learning, competitive intent and routine patterning. Journal of Business Research, 69(10): 4287-4303. Veugelers, R. 1997. Internal R&D expenditures and external technology sourcing. Research Policy, 26(3): 303-315. Volberda, H. W. 1996. Toward the flexible form: How to remain vital in hypercompetitive environments. Organization Science, 7(4): 359-374.

200

Volberda, H. W., Foss, N. J., & Lyles, M. A. 2010. Perspective-absorbing the concept of absorptive capacity: how to realize its potential in the organization field. Organization Science, 21(4): 931-951. Von Hippel, E. 1986. Lead users: a source of novel product concepts. Management Science, 32(7): 791-805. Walsh, J. P., & Ungson, G. R. 1991. Organizational memory. Academy of Management Review, 16(1): 57-91. Wang, C. L., & Ahmed, P. K. 2007. Dynamic capabilities: A review and research agenda. International journal of management reviews, 9(1): 31-51. Wang, P., Van De Vrande, V., & Jansen, J. J. 2017. Balancing exploration and exploitation in inventions: Quality of inventions and team composition. Research Policy, 46 (10): 1836- 1850. Wang, Z., Wang, N., & Liang, H. 2014. Knowledge sharing, intellectual capital and firm performance. Management Decision, 52(2): 230-258. Weick, K. E., Sutcliffe, K. M., & Obstfeld, D. 2005. Organizing and the process of sensemaking. Organization Science, 16(4): 409-421. Wenger, E., McDermott, R. A., & Snyder, W. 2002. Cultivating communities of practice: A guide to managing knowledge: Harvard Business Press. West, J., & Bogers, M. 2011. Profiting from external innovation: a review of research on open innovation. West, J., & Bogers, M. 2013. Leveraging External Sources of Innovation: A Review of Research on Open Innovation. Journal of Product Innovation Management, 31(4): 814-831. West, J., & Gallagher, S. 2006. Challenges of open innovation: the paradox of firm investment in open‐source software. R&D Management, 36(3): 319-331. Williams, R. 2012. Using the margins command to estimate and interpret adjusted predictions and marginal effects. Stata Journal, 12(2): 308. Williamson, O. E. 1999. Strategy research: governance and competence perspectives. Strategic Management Journal, 20(12): 1087-1108. Winter, S. G. 1964. Economic" natural selection" and the theory of the firm: Institute of Business and Economic Research, University of California. Winter, S. G. 1984. Schumpeterian competition in alternative technological regimes. Journal of Economic Behavior & Organization, 5(3-4): 287-320. Winter, S. G. 2003. Understanding dynamic capabilities. Strategic management journal, 24(10): 991-995. Winter, S. G., & Nelson, R. R. 1982. An evolutionary theory of economic change. University of Illinois at Urbana-Champaign's Academy for Entrepreneurial Leadership Historical Research Reference in Entrepreneurship. Withey, M., Daft, R. L., & Cooper, W. H. 1983. Measures of Perrow's work unit technology: An empirical assessment and a new scale. Academy of Management Journal, 26(1): 45-63. Woiceshyn, J., & Daellenbach, U. 2005. Integrative capability and technology adoption: evidence from oil firms. Industrial and Corporate Change, 14(2): 307-342. Wolfe, R. A. 1994. Organizational innovation: Review, critique and suggested research directions. Journal of Management studies, 31(3): 405-431. Wu, I.-L., & Chen, J.-L. 2014. Knowledge management driven firm performance: The roles of business process capabilities and organizational learning. Journal of Knowledge Management, 18(6): 1141-1164. Wynarczyk, P., Piperopoulos, P., & McAdam, M. 2013. Open innovation in small and medium- sized enterprises: An overview. International Small Business Journal, 31(3): 240-255.

201

Zahra, S. A., Filatotchev, I., & Wright, M. 2009. How do threshold firms sustain corporate entrepreneurship? The role of boards and absorptive capacity. Journal of Business Venturing, 24(3): 248-260. Zahra, S. A., & George, G. 2002. Absorptive capacity: A review, reconceptualization, and extension. Academy of Management Review, 27(2): 185-203. Zahra, S. A., Sapienza, H. J., & Davidsson, P. 2006. Entrepreneurship and dynamic capabilities: a review, model and research agenda. Journal of Management Studies, 43(4): 917-955. Zaichkowsky, J. L. 1985. Measuring the involvement construct. Journal of Consumer Research, 12(3): 341-352. Zander, U., & Kogut, B. 1995. Knowledge and the speed of the transfer and imitation of organizational capabilities: An empirical test. Organization Science, 6(1): 76-92. Zollo, M., & Winter, S. G. 2002. Deliberate learning and the evolution of dynamic capabilities. Organization Science, 13(3): 339-351.

202