Master Thesis
The moderating role of business group affiliation on the relationship between
innovation and firm performance: Empirical evidence from the Indian drug and
pharmaceutical industry
Burcu Yedikapu Student number: 2851423
University of Groningen – Faculty of Economics and Business
First Supervisor: Dr. S. R. Gubbi Co-assessor: Dr. A. A. J. van Hoorn
The moderating role of BG-affiliation on the relationship between innovation and firm performance: Empirical evidence from the Indian drug and pharmaceutical industry
Burcu Yedikapu, Faculty of Economics and Business, University of Groningen January 2016
Abstract
In this study, I examine whether business group affiliation impacts the relationship between innovation and firm performance. Using firm level data for the time frame between 2005-2011 from a total of 145 companies - both business groups and independent firms - operating within the Indian drug and pharmaceutical industry, I find support for the hypothesis that business group-affiliation indeed positively influences the relationship between innovation and firm performance. The findings provide implications for further research and the role of business groups in innovational contexts.
Table of content
1 INTRODUCTION ...6
2 LITERATURE REVIEW ... 11
2.1 Innovations within emerging markets ... 13
2.2 Business Groups... 14
2.2.1 Business Group diversity ... 16
2.3 Internationalization ... 17
2.3.1 Learning by exporting ... 18
2.3.2 Reducing costs of innovation ... 18
3 CONCEPTUAL MODELS ... 19
3.1 Model (1) Business group vs. non-business group-affiliated firms ... 19
3.2 Model (2) Business groups only ... 19
4 METHODOLOGY ... 20
4.1 Empirical Context: India ... 20
4.2 Data sources and sample ... 21
4.3 Measurement... 23
4.3.1 Dependent variable ... 23
4.3.2 Independent variable ... 23
4.3.3 Moderators and controls ... 24
4.3.4 Firm and year dummies ... 25
4.4 Model and Estimation ... 26
5 RESULTS ... 29
5.1 Business group vs. non-business group model ... 29
5.1.1 Controls ... 29
5.1.2 Independent variable effects ... 30
5.1.3 Interaction effects ... 30
5.2 Business group model only ... 31
Table overview
Table 1: Descriptive Statistiscs and Correlations - (1) Business group vs. non-business
group-affiliated firms ... 28
Table 2: Descriptive Statistiscs and Correlations - (2) Business groups only ... 28
Table 3: (1) Business group vs. non-business group-affiliated firms... 33
Table 4: (2) Business groups only ... 33 Table 5: Collinearity statistics - (1) Business group vs. non-business group-affiliated firms ... H Table 6: Collinearity statistics - (2) Business group model only ... H Table 7: Announcement examples made on the BSE: New product launch & Final product
approval ... J Table 8: Overview of tested hypotheses ... K
Figure overview
List of abbreviations
BG = Business groups
BSE = Bombay Stock Exchange
CMIE = Center for Monitoring Indian Economy CRAMS = Contract & Research Manufacturing Services
MNCs = Multinational companies
ROS = Return on Sales
ROTA = Return on Total Assets
1 INTRODUCTION
Rapid changes in technology, revolution in information systems and increased globalization of business activities has led to intensified competition among countries and markets worldwide. With the emergence of an ever more competitive environment innovation within firms has gained increased attention and recognition. Subsequently, innovation and the effective implementation of innovative ideas is being considered as crucial for maintaining competitive advantage and surviving in today’s knowledge intensive business environment as well as achieving superior levels of firm performance (Kostopoulos, Papalexandris, Papachroni & Ioannou, 2011; Gunday, Ulusoy, Kilic & Alpkan, 2011).
In line with this development the global economy has moved from developed markets towards emerging market economies. This trend has been accompanied by the rising number of innovations developed in emerging markets (Economist, 2010). Long only seen as a source for cheap labor and location of production, emerging market countries are nowadays competing with developed economies for business innovations (Govindarajan & Ramamurti, 2011). Petrick & Juntiwasarakij (2011:24) go even further and describe emerging markets as “hotbeds of innovation”, while the number and quality of innovations, also described as “frugal or reverse innovations” (Govindarajan & Ramamurti, 2011; Zeschky, Widenmayer & Gassmann, 2011) conducted by businesses in emerging markets is dramatically increasing. The reason behind this development is based on two specific factors, namely the changes in demand and supply of innovation. In fact, the demand for innovations has grown immensely during the past decades bred by the accelerating growth coupled with the growing needs of the middle class consumers of emerging market economies. On the other hand and the supply side of innovation, economic liberalization and technological change has resulted in the development of numerous organizations that are now able to both utilize local and global resources to innovate (Govindarajan & Ramamurti, 2011).
institutions and resources (Ricart, Enright, Ghemawat, Hart & Khanna, 2004 in Wang, Yi, Kafouros & Yan, 2015). The majority of businesses in emerging market economies suffer from a lacking institutional framework, the so called “institutional voids” and major resource constraints that limit their possibilities for successful operation within the market (Khanna & Palepu, 1997). In other words, innovation conducted in a traditional sense within developed markets cannot easily be extended to emerging market economies.
One example to understand the different approach of innovation in emerging versus developed markets may be the case of the Tata Ace, which was a small four-wheeled commercial vehicle that proved to be a resounding success, after being launched in 2005 by Tata Motors in India. Previous to the launch of the vehicle, the Indian transportation market was suffering from major issues emanating from underdeveloped roads in both urban and rural areas as well as governmental restrictions that banned the usage of trucks in cities and the three-wheelers on newly constructed highways, which represent the most common transportation method in India (Palepu & Srinivasan, 2008). Thus, consumers were in need for a commercial vehicle that would be allowed on every road and could travel anywhere. But unlike conditions in developing markets where innovations are based on unlimited availability of resources and the existence of institutions as well as lower levels of price consciousness, Tata Motors had to deal with major resource constraints. Based on high fuel prices and low income levels customers within India were very cost conscious with their transportation decision. Consequently, a majority of the potential customers were only willing to spend as much as for a three-wheeler (Palepu & Srinivasan, 2008). Despite the issues concerning the limited budget for the development of a new vehicle that meets both the demanded price level of potential customers and the challenging environmental conditions within India, Tata Motors was able to successfully launch the products by innovatively combining existing knowledge and technologies such as usage of existent facilities, creative design based on transportation issues and outsourcing of production activities to save costs. By meeting the unique needs of the Indian transportation sector, Tata Ace can be characterized as a novel niche vehicle that achieved rapid success within the Indian market.
economies (Yiu, Lau & Brutan, 2007; Guillén, 2000 in Gubbi, Aulakh & Ray, 2015). Even though literature provides several explanations for the role of business groups in emerging market economies, the majority of the literature agrees on the notion that they can be characterized as an organizational form that has primarily emerged due to institutional conditions in specific economies (Khanna & Yafeh, 2007 in Gubbi et al., 2015). Since absent or weak market institutions within emerging market economies enhance existing social inequalities, businesses strongly rely on local institutional agreements that govern opportunities and access to markets. Consequently, business groups are of great importance and account for a notable share of the private sector within multiple emerging market economies by serving as a substitute for missing institutions to ensure market function in the event of failures and guaranteeing transactions between unaffiliated firms (Carney et al., 2011; Khanna, Palepu & Sinha, 2005; Khanna & Rivkin, 2001).
Linking business groups with innovativeness in emerging market economies, multiple scholars argue that business group affiliation results in greater innovative success based on economies of scale and group internal knowledge spillovers granting firms with better access to complementary resources (Khanna & Yafeh, 2005; Mahmood & Mitchell, 2004; Belentzon & Berkovitz, 2010; Khanna & Palepu, 1997). Consequently, it is argued that business groups have a superior position within emerging market economies and are likely to maintain actions and strategies without endangering their survivability within markets when engaged in innovative practices (Kim, Kim & Hoskisson, 2010).
business group affiliation, the majority of the literature has investigated the relationship of business group affiliation with either innovation or firm performance levels (Belenzon & Berkovitz, 2010; Ghosh, 2010; Carney et al., 2011; Khanna & Palepu, 2007; Kim et al., 2010; Wang et al., 2015).
However, none of these studies have specifically investigated the combined effects of business group affiliation and businesses conducting innovation on firm performance levels. Thus, the field of research is rather fragmented, leaving numerous gaps behind. Especially the role of business group affiliation and its impact on the relationship between innovation and firm performance remains rather unexplored and requires a deeper investigation in this field.
Based on the resource based view (RBV) (Barney, 1991) this study conceptualizes two models by theorizing that in the context of business groups, innovation within firms lead to better firm performance levels when compared to independent firms, as business groups are in advantage due to group specific assets which in turn may benefit the adoption of innovation and ultimately result in superior firm performance (Mahmood, Zhu & Zajac, 2011; Khanna & Palepu, 2007 & Yiu et al., 2007). Additionally, this study also proposes that business group affiliation leads to higher levels of firm performance in the context of domestic innovation than for foreign innovation, since business groups may hold an edge over non-business group affiliated firms within boundaries of their own local market, thus are more likely to succeed when conducting innovation (Mahmood & Mitchell, 2004). As for the second model focusing on business groups only, this study argues that higher levels of business group diversity and business group exports both result in advanced innovation and ultimately better firm performance. Based on George & Kabir (2012) and Smith (2014) this study claims that highly diversified business groups with increased export activity may possess more diversified group capabilities and knowledge, which in turn may foster innovation and simultaneously result in improved firm performance.
The formulated hypotheses were tested by analyzing the firm performance of 147 firms in the Indian drug and pharmaceutical industry within the 2005-2011 period. The Indian drug and pharmaceutical industry was chosen to test the models, as it is composed of numerous business group affiliated and independent firms. On top of that, the industry can be characterized as one of the most thriving markets for innovative products worldwide (Economist, 2012).
This study contributes to the existing field of literature in several ways by extending the growing research on business groups as a distinct organizational form. Drawing on prior topics of research and the lacking empirical evidence, this study attempts to analyze the differences in the impact of both business group-affiliated and non-business group-affiliated firms on the relationship between innovation and firm performance. To fill this gap and to address the issue whether business group affiliated firms achieve higher levels of firm performance compared to standalone firms when engaged in innovative activity, the moderating role of business group affiliation will be investigated. Second, this study will additionally analyze the relationships between innovation and firm performance by only focusing on the business group level. Based on the second model, I will be able to examine whether highly diverse business groups, or business groups that show an intensified export activity, also ultimately achieve improved firm performance when practicing innovation. Third, this study will shed light on the given research gap by focusing on one particular country. Hence, we will focus on the Indian drug and pharmaceutical industry which has become a much sought after destination for innovative activity and is characterized by a notable presence of business groups within the private sector (Jha & Krishnan, 2013).
2 LITERATURE REVIEW
Literature provides numerous perspectives when looking at innovation on an organizational level. Thus, multiple definitions exist resulting from different types of viewpoints (Schumpeter, 1934; Damanpour, 1991; Damanpour & Gopalakrishnan, 1998). According to Schumpeter (1934) and the theory of economic development, innovation can be considered as an essential driver for competitiveness and the center of economic change revolutionizing economic structure. Thus, firms seeking higher levels of profit must innovate. One of the most common definitions of innovation is that it can be described as the adoption of any novel system, policy, program, process, product or service that is new to the firm (Damanpour, 2009; McKinley, Latham & Braun, 2014).
And yet, another perspective describes innovation as a form of organizational learning so as to cope with changes in the business environment. Multiple scholars stress the importance of organizational learning processes in the role of acquiring knowledge and boosting current and future firm performance to cope with external opportunities and threats (Calantone, Cavusgil & Zhao, 2002; Damanpour, 1991; Mavondo, Chimhanzi & Stewart, 2005; Salim & Sulaiman, 2011; Damanpour & Gopalakrishnan, 1998). This approach stands in line with the assumptions of Kostopoulos et al. (2011), stating that firms are forced to innovate due to pressures from intense competition and the need for improved products and services caused by diversified patterns of demand and constant change across markets. When moving this line of thought into the field of emerging market economies, the trend of changing business environments is intensified due to severe market conditions accompanied by disruptive shifts in capital, labor and product markets (Khanna & Palepu, 2007).
impact of process and product innovation on organizational firm performance. Most of these scholars confirm a positive relationship between product and process innovations and firm performance (Fernandes et al., Gunday et al., 2011). However, there are also studies arguing for the existence of a negative relationship or no relationship at all (Capon, Farley & Hoenig, 1990; Subramanian & Nilakanta, 1996).
Nonetheless, most of the previously mentioned studies have focused on multinationals (MNCs) from developed countries such as UK, Spain or Portugal, making it hard to apply and extend the results to businesses in emerging market economies (Fernandes et al., 2013; Kostopoulos et al., 2011; Damanpour et al, 2009). Organizations and the specific business environment in emerging market economies strongly differs from that of developed economies (Khanna & Palepu, 2007). Unlike business environments in developed countries, firms within emerging markets face conflicting pressures stemming from the constantly changing institutional environment (Khanna & Palepu, 2007). Based on such bottlenecks and the weak internal innovation capabilities, innovation practices conducted in emerging market economies also differ from that of developed countries (Luo & Tung, 2007; Govindarajan; Ramamurti, 2011). Consequently, studies focusing on the innovation-performance relationship of MNCs from industrialized countries cannot be utilized to understand the specific impact of innovation on performance levels of emerging market companies, as they lack in taking account the differences in both business environment and innovative approach.
When looking at existing research on the innovation-performance relationship, several factors were found to moderate the link between innovation and firm performance levels. According to Tsai (2001) and Kostopoulos et al. (2011) a firm’s absorptive capacity has a significant impact on the innovation and performance relationship of a company. Thus, the higher the level of absorptive capacity, the greater the absorption of knowledge, hence the success of innovative activities and subsequently a firms performance. In addition to that firm size and firm age were also identified as factors moderating a firm’s innovation and performance levels, while smaller and younger firms have been identified as having a more positive impact than older and bigger firms (Rosenbusch; Brinkmann & Bausch, 2011). Again other studies found evidence that the level of innovative success and improved firm performance also depends on the type of innovation which is conducted as well as the cultural and environmental context in which a firm operates (Damanpour, 2009; Rosenbusch et al., 2011 & Fernandes et al., 2013).
crucial role in multiple emerging markets by serving as substitutes for missing institutions enabling numerous companies to successfully engage in business operations within emerging market economies (Carney et al., 2011; Khanna et al., 2005).
However, none of the past scholars has done research in this particular field. Thus, the moderating role of business group affiliation on the relationship between innovation and firm performance is rather uncertain and unexplored. Although prior research has provided critical insights on the relationship of innovation and firm performance, there are still gaps in the field of how innovative activities within emerging market firms affects performance and how this relationship is moderated by business groups within emerging market economies.
2.1 Innovations within emerging markets
During the past decades emerging market economies have gained increased attention, while the number of businesses engaging in innovative activities is continually increasing (Petrick & Juntiwasarakij (2011:24). The locus of entrepreneurial activity recently started shifting from industrialized to emerging economies based on ongoing industrial and urban revolutions, leading the worlds’ economy towards markets that have never been witnessed before (McKinsey & Company, 2015a). Besides, technological changes and economic liberalization resulting from the “flattening” of the global economy, has brought multiple successful emerging market firms to the fore (Govindarajan & Ramamurti, 2011). What is more, emerging market firms are in a development to catch up with the firms of the industrialized countries by investing into innovation for low and medium-end markets, thus are no longer solely seen as a target for Western MNCs.
knowledge and result in products and services that have never been developed before, thus mainly aiming for a small target of early adopters within a niche market (Iyer et al., 2006).
The reason for this difference is based on several factors and their dissimilarity within both emerging and advanced markets. Market characteristics, institutional development and potential business and consumer culture within emerging markets are fundamentally different from that of developed economies. Based on low levels of average income per capita, the majority of the mass market within emerging economies requires products and services with improved price-performance features, hence demanding for “affordability innovations” (Zeschky et al., 2011; Govindarajan & Ramamurti, 2011). Therefore, innovations conducted in emerging markets also involve improving of already existing products and services as well as novel ideas in the ultimate distribution, sales and finance of these products and services, to meet locally harsh environmental and technological conditions (Khanna & Palepu, 2005). To conclude, innovative practices and types of innovation that are conducted in advanced markets may not be applicable when considering innovations in emerging market economies.
2.2 Business Groups
Multiple studies have identified business groups as a fundamental organizational form within a majority of emerging market economies, carrying the role of substitutes for external markets (Khanna & Palepu, 1997 in Yiu et al., 2007). Literature provides several reasons concerning the emergence of business groups. One stream of literature describes business groups as a response to imperfect factor market situations (Khanna & Palepu, 2005 in Gubbi et al., 2015). Other scholars characterize business groups as a governmental body created to achieve state-related objectives in terms of politics and economy (Sing & Gaur, 2009). Again others argue that business groups can also be seen as social institutions that govern opportunities and access to markets by influencing economic exchanges between companies (Granovetter, 1995 in Gubbi et al., 2015). Even though multiple explanations exist, the majority of the literature agrees on the notion that business groups can be characterized as an organizational form emerged as a response to institutional conditions within diverse economies (Khanna & Yafeh, 2007 in Gubbi et al., 2015).
industries (Economist, 2011). According to Barney (1991) and the theory of the resource-based view, organizations depend on valuable, rare, inimitable and non-substitutional firm-specific resources to create competitive edge. Given the supportive nature and internal functioning of business groups, this study argues that business group affiliation positively influences innovative activities and ultimately performance levels of firms.
I base this premise on three specific arguments. First, intragroup ties allow affiliates of a business group to coordinate strategic action and behavior as well as resources. Based on shared norms that are embedded within group structures, such ties are able to decrease transaction costs by simplifying the access and transfer of group internal resources between affiliates (Granovetter 1995 in Chang, Chung & Mahmood, 2006). By allowing access to group specific resources such as technology, capital and complementary information about products and services, business group affiliation may result in greater innovativeness for member firms than for independent firms (Mahmood & Mitchell, 2004). Additionally, by drawing on group-internal financial assets, business groups can also provide financial aid for affiliates with greater ease than standalone firms may access, thus increasing the level of engagement and mobility for affiliates to actually participate in innovative opportunities and ultimately increase firm performance levels (Khanna & Yafeh, 2005). Second, business groups also facilitate the transfer and adoption of knowledge concerning different industry types between affiliates, as a majority of firms within the group are operating in diverse industries (Khanna & Rivkin, 2001). Hence, affiliates are provided with critical insights about different sectors and are more likely be able to identify an industry’s needs in terms of innovative improvements. Third, the richness of information and knowledge shared within business group levels coupled with competitiveness among affiliates results in greater flexibility to changes within the environmental context and occurring innovative opportunities, thus may foster a firm’s performance levels (Lamin, 2013 in Gubbi et al., 2015).
Subsequently, due to group specific characteristics such as resource-sharing and access to superior and complementary information and knowledge skills, business groups provide a beneficial platform for affiliates to be at the technological frontier by fostering innovative success and ultimately, firm performance levels. Based on these findings the hypothesis is as follows:
As previously discussed, business groups play a fundamental role within most of the emerging market economies. Taking into account the availability of business group specific capabilities such as an unlimited resource base, complementary knowledge and information comprising multiple industries and sectors, business groups are in a favorable strategic position within local markets (Mahmood & Mitchell, 2004; Gubbi et al., 2015). Based on their distinctive features and strategic positioning certainly adapted to local markets and the prevailing market and institutional context, business group affiliates conducting innovation within domestic markets may be more likely to succeed and achieve higher levels of firm performance than in radically different foreign markets where they do not, or are less likely to possess extensive market specific know-how and experience.
Since domestic innovation is subject to domestic regulations and quality standards, business groups hold an edge over non-business group affiliated firms within local markets. On the contrary, when practicing foreign innovation in foreign markets, business groups may have no advantage to non-business group affiliated firms, as they might face critical challenges and cut-through competition from competitive foreign firms and are more likely to be crushed or outperformed by global MNCs with unlimited resources and knowledge about the specific foreign market. Based on this line of thought the second hypothesis looks as follows:
H2: The positive effect of BG-affiliation on the relationship between innovation and firm performance will be more pronounced for domestic innovation than for foreign innovation.
2.2.1 Business Group diversity
One of the most essential characteristics of business groups within emerging market economies is the diversity of operations in multiple industries. For instance, the Indian Tata Group established in 1874 is presently operating in multiple industries and business sectors such as materials, engineering, energy, chemicals, communication and information systems, services, and consumer goods (Economist, 2011).
Mitchell, 2004). Looking at this issue from the resource-based view (Barney, 1991), diverse business groups may play the role of an intermediary to allocate and generate information and complementary means by leveraging on each other’s resources (Gosh, 2010; George & Kabir, 2012; Zaheer & Bell, 2005). Thus, highly diversified business groups may be more effective when exploiting spillovers, acquiring integral information and resources more cheaply and easily by gaining access to group member’s resource bases, which in turn may facilitate and promote the development of innovative activities. Based on common goals and the diffusion of innovative practices within network ties such as business groups (Granovetter, 1985), affiliates may successfully utilize novel ideas and information to innovate. Contrary to highly diverse business groups, less diversified firms may not be able to profit when engaging in innovation and ultimately achieve lower levels of firm performance (Khanna & Rivkin, 2001; Gosh, 2010).
Hence, building on their heterogeneous knowledge and resource base, business groups are able to achieve a pioneering position within the market by creating novel responses to technological developments. Thus, I argue that the more diversified a business group, the greater the diversity of knowledge and experience from a variety of industries and economic activities, hence the greater the probability of achieving peerless success when engaged in innovation and ultimately firm performance. Consequently, this paper proposes that the relationship between innovation and firm performance, is positively moderated by the business group diversity. Based on this line of thought the third hypothesis looks as follows:
H3: Among business group affiliated firms the positive effect of innovation on firm performance will be more pronounced at higher levels of business group diversity.
2.3 Internationalization
2.3.1 Learning by exporting
Emerging market firms derive crucial advantages when engaged in export activities in larger, sophisticated markets such as access to foreign products and process technologies, as well as complementary expertise and knowledge, thereby leading to greater economies of scale and ultimately increasing the firm’s capability for learning-by-doing (Smith, 2014; Yiu et al., 2007). In fact, exports cannot only be seen as enlarging the level of sales for existing products, but go beyond such well known firm outcomes, thus can be characterized as a strategic tool to facilitate innovation within the firm and its domestic markets by enabling the adoption of novel and complementary sources of foreign knowledge and access to factor inputs and innovation intermediaries (Salomon & Shaver, 2005; Wu et al., 2015; Smith, 2014 in Grossman & Helpmann, 1991).
Looking at this issue from the context of business groups, access and membership to a specific business group may provide affiliates with the export experience of other affiliates within the group that have already established a presence in foreign countries, thus may also be more likely facilitate the adoption of sophisticated innovation practices and ideas based on shared information and experience within business group levels (DiMaggio & Powell, 1983 in Guillén, 2002). Consequently, affiliates may benefit from two advantages when engaged in increased export activity, namely the access to complementary sources of foreign knowledge and information about foreign markets as well as possible innovation opportunities that might provide potential to capitalize on.
2.3.2 Reducing costs of innovation
Exports can be characterized as a straightforward way to enter foreign markets, as it is based on lower levels of commitment and risk (Cassimann & Golovko, 2011; Alvarez & Robertson, 2004). Taking into account the resource constraints such as limited capital and absence of innovation supporting institutions within domestic markets, firms within emerging market economies may be more likely to invest in operations abroad to spread the costs arising from R&D over the larger volume of sales that is achieved through exports in foreign markets. Considering the increased level of profitability via exports and the accompanied benefits such as cost reduction of innovative activities and the increase in firm’s internal capabilities via knowledge spillovers, this study assumes that higher levels of exports within business groups will strengthen the positive relationship between innovation and firm performance. Thus, the fourth hypothesis looks as follows.
3 CONCEPTUAL MODELS
3.1 Model (1) Business group vs. non-business group-affiliated firms
3.2 Model (2) Business groups only
Moderator BG diversity IV* Innovation DV* Firm Performance Controls BG size BG age Leverage BG Sales R&D intensity Business group total
assets Lagged ROS
Moderator
BG exports
*IV= Independent Variable DV Dependent Variable IV* Innovation DV* Firm Performance Foreign innovation Domestic innovation Controls Firm size Firm age Leverage Sales R&D intensity Export intensity Lagged ROS
*IV= Independent Variable DV Dependent Variable
Moderator
4 METHODOLOGY
4.1 Empirical Context: India
As for the empirical analysis, the Indian drug and pharmaceutical industry was identified as the context base for two main reasons. First, India provides a perfect context to test the hypotheses that have been formulated as most of the industries are dominated by the existence of business groups that play a major role within the economy (Khanna & Palepu, 1997). Secondly, India can be characterized as one of the most emerging pharmaceutical markets worldwide with an estimated value of US $ 8.2 billion. In fact, the Indian drug and pharmaceutical industry can be seen as one of the largest providers of generic drugs and holds the position of being the third largest industry in terms of volume and fourteenth by value (as most of the drugs are marketed to very low prices) worldwide (Economist, 2012; BloombergView, 2014). According to the Economist (2012), Indian drugs account for almost 20% of the volume that is exported globally, resulting in a striking export turnover of $10 bn. spread over 200 countries worldwide. At the same time Indian pharmaceutical firms also dominate the local industries within the boundaries of their own country, where generic drugs make up for 70 to 80% of the market volume (Deloitte, 2014; McKinsey & Company, 2015a). With a population of over one billion, low labor and research costs to design and manufacture novel generic drugs, India represents a major market for innovative pharmaceutical companies and is expected to grow at an average growth rate of 10% between the period of 2014 to 2020 by ultimately achieving a forecasted market size of 55 bn. in 2020 (Deloitte, 2014; McKinsey & Company, 2015a). The reason behind this immense market power is based on the interplay of several factors that are characteristic to the unique macroeconomic environment in India. Namely, the expanding GDP level, the increase in wealthy consumers with raising spending power, the demand for innovative drugs based on the Indian populations’ adoption of a Western lifestyle resulting in typical Western style chronical diseases as well as the government structure of India that has encouraged the production of generics by keeping price levels relatively low and introducing product patent regimes to the country to support the industry in maintaining its high growth pace despite economic downturns (Bloomberg, 2011).
related to safety issues and the compliance of the drug and clinical trial quality with the world’s standards, as well as the law including the patenting of newly developed products and ultimately also serve the effect for the purpose. Accordingly, the discovery and the development of innovative drugs is characterized with high levels of uncertainty. The process of discovering and implementing a specific drug may extend more than 15 years, raising the costs to more than $500 million for a drug. Out of every 10,000 compounds screened, only 2.5% get into the next step of preclinical testing. Out of these pre-clinically tested compounds only 2% make it into clinical testing, while 80% pass phase I, 30 % phase II and again 80% phase III. Thus, out of every 10,000 compounds only one drug is ultimately approved, meaning that the likelihood of a discovered molecule to be marketed as a commercialized product accounts only 0.01% (Rothaermel & Deeds, 2004). To overcome these hurdles and to be a good player in the global market, firms within the Indian pharmaceutical industry are forced to steadily advance internal R&D skills and at the same time, invest in its Contract & Research Manufacturing Services (CRAMS) segment (Deloitte, 2014). As a consequence of the industries conditions, a collaborative approach and partnerships within firms such as business groups are more likely to result in value-added products for the global market.
To put it in a nutshell, India and its drug and pharmaceutical industry represent an ideal case to investigate the impact of business groups on the relationship between innovation and firm performance levels.
4.2 Data sources and sample
ties such as business groups may have played a major role in the survival and progress of many firms. Hence, the period of this study is of fundamental importance for the Indian drug and pharmaceutical industry itself and may result in crucial insights regarding the context of this study.
The initial sample size consisted of 201 firms with a total of 2742 firm-year observations. In order to create a homogenous sample, some observations were omitted from the dataset. First, as this study focuses on Indian firms, all foreign-owned firms were removed from the sample, which reduced the sample size by 12 firms and 75 firm-year observations, respectively. Second, firms with missing firm and industry related data were eliminated. This reduced the sample by another 35 firms and 161 firm-year observations. Third, this study also excluded firms with fewer than six observations over the period 2005-2011, and also eliminated smaller firms based on assets to create a more consistent sample and avoid the emergence of outliers on other variables. Again this process decreased the sample size by additional 43 firm-year observations. As for the fourth and last step, outliers with inadequate values on key variables were excluded, which in turn reduced the sample size by six observations. As a result of this process a firm-level panel dataset of specific companies was created comprising a total sample of 145 firms with 983 firm-year observations, while 102 being independent firms and 43 business group affiliated firms operating within the Indian pharmaceutical industry.
The second dataset was based on announcements made on the BSE by publicly listed firms within the Indian pharmaceutical sector. In order to identify new product launches within the industry, this study relied on a compilation of product launches obtained from news releases made by the top pharmaceutical firms on the BSE during the time frame between 2005 and 2011. Examples about typical announcements that were classified as a new product launch can be found in the Appendix (Table 8). Overall, - of the 1399 announcements made between the studied time frame, the dataset contained 273 observations on new product-country launches that could be clearly classified as drugs that were introduced to the market for the first time anywhere within the world (including repeat new product launches by the same firm). Of the 273 “new product launch” - announcements, 95 were made within domestic markets, thus within India, while 178 were introduced to the foreign markets such as United States, Canada, Japan and the UK.
4.3 Measurement
4.3.1 Dependent variable
Based on the theoretical model, the variable of firm performance is identified as the dependent variable. Prior scholars distinguish between accounting- based measures such as return on invested capital (ROIC), return on (total) assets (ROA/ROTA) and return on sales (ROS) as well as market- based measures of firm performance using for instance the market-to-book value (Kim et al., 2010). This study utilized ROTA, as it is used as a standard measure in majority of the studies and can be characterized as one of the most common indicators of firm performance. For instance, Kim et al. (2010) investigated the impact of market-oriented institutional changes on firm performance levels of business-group-affiliated multinationals and included ROTA as an indicator to measure firm performance. Other studies also use ROTA to investigate firm performance, as it is easily retrievable from the majority of the databases for many different firms. Taking these factors into account, firm performance was measured by the ratio of profits after taxes to total assets (ROTA).
4.3.2 Independent variable
As for this study the method of content analysis was performed in order to measure the independent variable. A detailed description about the employed content analysis and procedure will follow below. Innovation was measured using the launch of new products which is conceptualized by two specific variables, namely domestic (i.e. India) and foreign new product launch (i.e. introduction in other markets outside of India). This operationalization offers a crucial advantage when compared to other standard indicators of innovation. First, this measure includes innovations that are not patented but still in the production process (Liu & Buck, 2007 in Wang et al., 2015). Even though patents can be a good indicator for technological developments, they are not likely to mirror the value of these technologies and do not consider that many innovations are also not protected by patents. This trend is especially prevalent within the majority of emerging market economies, where constant innovations occur due to the urge for improved products and processes that are triggered by the rapid and disruptive changes within the institutional environment. Second, by distinguishing the variable of new product launch into domestic and foreign new product launches, this study also enables to analyze the impact of innovations within a specific target country on firm performance levels.
entity within a target country before they ultimately can be marketed. However, a regulatory approval does not necessarily ensure the commercial availability of a new chemical entity, thus formal approval may not always lead to the commercial launch of a new product.
Having these difficulties in mind, a content analysis of a dataset retrieved from the BSE with a total of 1399 company and newspaper announcements for the period between 2005 and 2011 was conducted. Referring to Elo & Kyngäs (2008) a content analysis is a systematic technique in order to compress many words of text into content categories based on definite categorization and coding rules. In order to minimize and prevent errors regarding the content analysis, a manual coding procedure was adopted instead of using existing automatic programs. The content analysis followed several steps. The first step, thus the coding procedure has been practiced as follows: Announcements that could be clearly classified as new products introduced to the market for the first time anywhere within the world were coded as “1” (see Appendix: Table 8), while the remaining news releases that were related to other firm-specific activities and were not identifiable as an introduction of a new chemical entity were coded as “0”. This approach has been performed manually for each company within the sample by reviewing each announcement made between the given period. As a next step following the coding procedure I also decided to use the cumulated values by adding up the new product launches within a specific company for each following year. To minimize undercounting of new product launches, two selection criteria’s have been utilized. First, the product had to be introduced to the worldwide market for the first time. Second, the product had to receive the final approval by regulatory bodies (e.g. USFDA) to be marketed immediately in specified target countries. This process yielded a total sample of 273 “new product launch” announcements for the period between 2005 and 2011.
4.3.3 Moderators and controls
This study includes three moderators, namely business group (BG)-affiliation, BG diversity and BG exports. BG affiliation was operationalized to a dummy variable, which equals “0” if the firm is not affiliated to a firm and “1” if the firm is independent. BG diversity was measured by the total number of different industries in which a business group operates referring to the two digit SIC industries. The last and third moderator of BG exports was measured by using the export intensity as an indicator. Export intensity was measured by the ratio of foreign sales to total sales.
included to control for the effects of economies of scale on firm performance (Kim et al., 2010). The variable is operationalized as the logarithm of total assets owned by the firm. Firm age is also considered to be an important determinant of firm performance. Scholars widely argue that older firms possess greater experience, thus receive benefits of learning and are more likely associated with first mover advantages and positive firm performance levels (Leiponen et al., 2010; Plen-Dujowich, 2009). However, it is also argued that older firms may be prone to inertia, thus less flexible to adapt to competition and engage in innovative activities which may influence the performance levels negatively (Capon, Farley & Hoenig, 1990). In order to control for this effect, this study utilized the variable firm age defined as the difference between the focal year and the firm’s incorporation. As used by many other studies investigating the effects of firm performance levels, leverage (ratio of total debts/ to total sales) was employed as another control variable. This study also included sales as a proxy for the firm size, since firm size has an impact on firm performance levels and a proxy for firm size is used in almost all studies explaining firm performance. Prior studies have also stressed the association between intangible assets on firm performance levels (Kim et al., 2010). Thus, this study utilized R&D intensity (ratio of R&D expenditures/ to total sales) as an indicator for technological assets to capture the effects on firm performance. Group total assets (the sum of group total assets of listed firms affiliated to a business group) can be seen as a proxy for business group diversity and business group size and were also controlled, as it may also impact the specific firm performance levels. Ultimately, BG size (number of listed firms in a group) and BG age (years between the incorporation of the oldest affiliate to the group) were controlled since firms affiliated to a business group may result in differences concerning firm performance levels.
4.3.4 Firm and year dummies
4.4 Model and Estimation
In order to measure for the effects of the independent and moderator variables on the dependent variable of firm performance, a linear regression analysis was performed. The following equations were utilized to test the formulated hypotheses for the (1) business group vs. non-business group model and (2) non-business group model only:
(1) BG vs. non-BG model:
1. ROTAInd + BG = α +β1 (BG) + β2 (BG x Innov) + ε
2. ROTAInd + BG = α + β1 (BG) + β2 (BG x domestic Innov) + β3 (BG x foreign Innov) +ε
(2) BG model only:
ROTABG = α +β1 (BG) + β2 (BGD x Innov) + β3 (BGE x Innov) + ε
ROTAInd + ROTAInd + BG stands for the dependent variable of firm performance for both (1)
business group-affiliated firms and non-business group-affiliated, while ROTABG stands for the
dependent variable of firm performance and only includes (2) business group level data. α is the intercept and β is the parameter. The variable “BG” stands for business group-affiliation. “BG x Innov” measures the interaction effect of the independent variable innovation and the moderator of business group-affiliation. “BG x domestic Innov” stands for the interaction effect between the independent variable of domestic innovation and the moderating variable of business group-affiliation. Whereas “BG x foreign Innov” measures the interaction between the independent variable of foreign innovation and the moderator of business group-affiliation. Additionally, “BGD x Innov” shows the interaction effect between the independent variable of innovation and the moderator of business group diversity, while “BGE x Innov” measures the interaction between the moderating variable of business group export intensity and the independent variable of innovation.
measured variables will be tested to control for the relationship of observed data values. In order to conduct an OLS regression analysis, several assumptions need to be made, namely normality, homoscedasticity, multicollinearity and endogeneity. Further information about the assumptions and the results for the tested data set will be explained in the following.
The first assumption is that the values of the model are normally distributed, as it is assumed that ROTA is a linear function and a normally distributed variable. Looking at the histograms (Appendix: Figure 1 & 2) one can see that the dependent variable of ROTA can be described as normally distributed but shows an upward shift in both the first (1) and second (2) model. However, some of the other variables are not normally distributed but they also do not appear to be random. Referring to the Gauss-Markov theorem, even though when normality is not given for all variables, the method of OLS can still be adopted when the values are uncorrelated and homoscedastic in nature (Schaffer, 1991). Thus, it can be concluded that the first assumption of normality can be confirmed.
As for the second assumption of multicollinearity, it is expected that the values are uncorrelated with each other, meaning there’s no correlation or autocorrelation between the given variables of the models. I detected for multicollinearity using the variance inflation factor (VIF= 1/ tolerance). Based on the results (Appendix: Table 6 & 7) it can be confirmed that there’s no correlation between the variables for both models.
For the third assumption of homoscedasticity the independent variables should not be random. Based on the analysis it can be confirmed that there is no heteroscedasticity, thus homoscedasticity can also be accepted for both models.
Table 1:Descriptive Statistiscs and Correlations - (1) Business group vs. non-business group-affiliated firms
Variables Mean S.D. Min Max 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.
1. ROTA 4.25 10.40 -51.91 55.67 1 2. Innovation 1.15 4.87 0.00 52.00 0.86 1 3. Domestic innovation 0.46 2.01 0.00 24.00 0.89 0.68 1 4. Foreign innovation 0.69 3.76 0.00 44.00 0.61 0.91 0.33 1 5. BG affiliation 0.28 0.45 0.00 1.00 0.78 0.26 0.19 0.24 1 6. Firm size a 3.14 2.08 -2.21 8.16 0.26 0.36 0.26 0.32 0.42 1 7. Firm age 24.81 15.91 0.00 110.00 0.10 0.66 0.04 0.63 0.29 0.25 1 8. Sales a 24.81 189.05 0.00 1457.48 0.19 0.57 0.41 0.52 0.37 0.66 0.26 1 9. Leverage 1.01 1.74 0.00 22.37 0.14 -0.03 -0.04 -0.01 0.61 0.12 -0.96 -0.01 1 10. Export intensity 0.24 0.26 0.00 0.97 0.14 0.21 0.14 0.19 0.08 0.48 0.03 0.36 -0.01 1 11. R&D intensity 0.05 0.45 0.00 6.64 0.11 0.93 0.06 0.9 0.13 0.15 0.02 0.11 0.14 0.07 1 12. Lagged ROS -39.28 626.09 -13200.00 97.91 0.12 0.19 0.17 0.15 0.44 0.88 0.05 0.04 -0.07 0.69 0.01 1 n = 987 * p < 0.05
a Figures in billion USD.
Table 2:Descriptive Statistiscs and Correlations - (2) Business groups only
Variables Mean S.D. Min Max 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.
1. ROTA 5.90 9.11 -25.14 55.68 1 2. Innovation 3.22 8.50 0 52 0.11 1 3. BG size a 7.64 9.02 0 61 -0.21 0.05 1 4. BG age 44.00 25.54 12 140 0.37 -0.20 0.50 1 5. Firm size a 4.57 1.77 0.97 8.16 0.35 0.46 0.01 0.01 1 6. Firm age 32.14 19.40 0 104.00 0.93 -0.06 -0.06 0.63 0.15 1 7. R&D intensity 0.15 0.84 0 6.64 0.01 0.06 0.02 -0.01 0.17 -0.03 1 8. BG export intensity 27.15 118.67 0 863.61 -0.17 -0.05 0.63 0.13 -0.15 -0.20 0.12 1 9. Leverage 1.19 2.24 0 22.37 -0.28 -0.06 0.09 -0.09 -0.09 -0.13 0.20 0.03 1 10. Lagged ROS 4.50 19.64 -211.32 44.03 0.50 0.17 -0.17 0.06 0.31 0.10 0.04 -0.13 -0.21 1 n = 277 * p < 0.05
5 RESULTS
Table 1 and 2 illustrate the descriptive statistics and the correlation coefficients of the observed variables for both models. The theoretical concept of this study is based on two distinct models examining the relationship between innovation and firm performance. The first one focuses on the differences of (1) business group affiliates vs. non-business group/independent firms, while the second subsample only considers the effects within (2) business groups. Hence, results and the details of the respective analyses will be explained for both models (1) and (2) separately.
Looking at the fit parameter of R2 in each model, it can be seen that the values are increasing by each step of the analysis, meaning that the stepwise addition of the variables leads to a better fit between data and statistical model. In addition to that, the R2 values remain between 42% and 60% for the specific models indicating a moderate to high explanation of the variance in the outcome variable of firm performance.
Based on the stepwise analysis, the first models only consists of the control variables. As a second step, the independent variables were added, followed by models showing the direct impact in the case for each interaction variable, before summarizing all variables in the fully specified models. The stepwise regression results and a detailed overview can be found in table 3 and 4.
5.1 Business group vs. non-business group model
In this model, the full sample was utilized comprising a total of 982 firm-year observations about 145 firms with 43 being business group firms and 102 independent firms.
5.1.1 Controls
Table 3 Model 1 presents the effects of the control variables on the dependent variable of ROTA. The utilized variables for the observed data sample are firm age, firm size, sales, leverage,
export intensity, R&D intensity and lagged ROS values. Controls such as sales, R&D intensity
and lagged ROS seem to have a positive and significant impact on the dependent variable of
ROTA. On the contrary, firm age has a negative and significant impact on ROTA.
5.1.2 Independent variable effects
In Model 2 (Table 3) the independent variable of innovation was added to the regression analysis. The construct of the independent variable of innovation has a positive but not significant impact on the dependent variable of ROTA (β = 0.093; p = 0.320). When looking at the control variables R&D intensity and lagged ROS remain having a positive and significant impact on ROTA with β = 2.845; p < 0.5 and β = 0.001; p < 0.5, respectively. The control of
sales loses its significance in this model. Besides, also firm age remains having a negative but
significant impact on the dependent variable of ROTA with β = 0.263; p < 0.5. No change is recognized for the remaining controls such as firm size, leverage and export intensity, also having no significant impact on the dependent variable of ROTA in this model. In Model 3 (Table 3) two other independent variables were added to the regression analysis, namely
domestic innovation (β = 0.086; p = 0.782) and foreign innovation (β = 0.127; p = 0.453). Both
of these constructs have a positive but not significant impact on ROTA. Except some minor variations in the β coefficients none of the control variables change in this model.
5.1.3 Interaction effects
For the purpose of this study, the interaction effects between the independent variables of
innovation, domestic innovation and foreign innovation and the moderating variable of business group affiliation is of fundamental importance. To ultimately examine the previously
formulated hypotheses the interactions have been constructed and tested. Model 4 – 6 (Table 3) present the direct effect on ROTA in the case of each interaction variable. Model 7 as the full model includes all the observed variables as well as the interactions between the independent variables and the moderating variable.
In hypothesis H1, I proposed that the positive effect of innovation on firm performance
will be stronger for business group affiliated firms than for non-business group affiliated firms. The results (Table 3, Model 4) were not as expected. According to the findings, the interaction between innovation and business group affiliation has a positive but nonsignificant impact on ROTA (β = 0.461; p = 0.478). Thus, hypothesis H1 can be rejected.
As for the hypothesis H2, I predicted that the positive effect of business group affiliation
Looking at the regression outcomes for the third interaction variable between foreign innovation and business group affiliation, the fully specified model (Model 7) shows a negative and significant impact of the interaction on the dependent variable of ROTA (β = -1.425; p = 0.077), also backing hypothesis H2. In terms of hypotheses H2, this means that the positive effect of
business group affiliation on the relationship between innovation and firm performance is indeed more pronounced for domestic innovation than for foreign innovation.
5.2 Business group model only
For this model a subsample of 43 business group affiliated firms was used, spanning a total of 277 firm-year observations.
5.2.1 Controls
As for the second model focusing on business group affiliates only, seven control variables have been used, namely business group size, business group age, firm age, firm size, leverage,
R&D intensity and lagged ROS values. According to the results (Table 4, Model 1) only firm size shows a significant impact on ROTA, while business group size, firm age, leverage and R&D intensity are negatively nonsignificant, whereas business group age and lagged ROS
positively nonsignificant.
5.2.2 Independent variable effects
As previously done in the business group vs. non-business group model, also for the second data set considering business group firms only, the independent variable has been added to the regression analysis. When looking at Model 2 (Table 4) it appears that the independent variable of innovation has a positive but nonsignificant impact on the dependent variable of ROTA (β = 0.018; p = 0.755). Comparing the results for the control variables one can see that the findings remain the same as in Model 1, with business group size (β = - 0.102; p = 0.147); firm age (β = - 0.007; p = 0.866); leverage (β = - 0.052; p = 0.780) and R&D intensity (β = - 0.073; p = 0.872) having a negative but not significant and business group age (β = 0.040; p = 0.293) as well as lagged ROS (β = 0.015; p = 0.563) a positive but not significant impact on ROTA. The control variable of firm size also remains positively significant but only when the p value is raised to p = 0.1 (β = 0.059; p < 0.1).
5.2.3 Interaction effects
In hypothesis H3, I proposed that among business group affiliated firms, the positive effect of
the dependent variable of ROTA, but shows no significance (β = 0.112; p = 0.665 & β = 0.368; p = 0.185). Thus, not supporting hypothesis H3, as such it can be rejected. For hypothesis H3
this implies that the positive effect of innovation on firm performance among business group affiliated firms will not be more pronounced at higher levels of business group diversity. In hypothesis H4, I predicted that among business group affiliated firms the positive effect of
innovation on firm performance will be more pronounced at higher levels of export intensity. The findings were unexpected. The coefficients for the interaction variable of innovation and export intensity show a negative but significant impact on the dependent variable of ROTA in both the direct (β = - 10.369; p < 0.01) and the fully specified model (β = -10.775; p < 0.01) (Table 4, Model 6 + 7). This shows no support for hypothesis H4. Despite the significance of
the last interaction variable, hypothesis H4 will be rejected as it is in contradiction to the findings
Table 3:(1) Business group vs. non-business group-affiliated firms Ordinary least squares regression on Firm Performance (ROTA)
Variables Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7
Firm age -0.26* (0.103) -0.263* (0.103) -0.263* (0.104) -0.249* (0.105) -0.227* (0.104) -0.264* (0.105) -0.25* (0.105) Firm size -0.174 (0.367) -0.157 (0.367) -0.157 (0.367) -0.125 (1.849) -0.073 (0.368) -0.16 (0.369) -0.103 (0.369) Sales 0.008* (0.004) 0.006 (0.004) 0.006 (0.004) 0.006 (0.004) 0.007* (0.004) 0.006 (0.004) 0.005 (0.004) Leverage -0.296 (0.191) -0.287 (0.191) -0.287 (0.191) -0.286 (0.191) -0.288 (0.191) -0.287 (0.191) -0.282 (0.191) Export intensity 2.202 (2.327) 2.073 (2.33) 2.072 (2.332) 2.034 (2.332) 1.9 (2.325) 2.07 (2.332) 1.747 (2.326) R&D intensity 2.839* (1,028) 2.845* (1.028) 2.846* (1.029) 2.859* (1.028) 2.844* (1.026) 2.855* (1.029) 2.821* (1.026) Lagged ROS 0.001* (0.000) 0.001* (0.000) 0.001* (0.000) 0.001* (0.000) 0.001* (0.000) 0.001* (0.000) 0.001* (0.000) Business group affiliation Dummy 2.288 (1.826) 2.008 (1.848) 2.012 (1.856) 1.97 (1.849) 2.232 (1.848) 2.086 (1.845) 2.47 (1.858)
Innovation 0.093 (0.094) -0.266 (1.113)
Domestic innovation 0.086 (0.312) -2.395* (1.323) -3.855* (1.537) Foreign innovation 0.127 (0.034) 0.488 (1.109) 2.425* (1.300) Innovation x Business group affiliation 0.461 (0.65) Domestic innovation x Business group affiliation 1.927* (0.819) 2.930* (1.007) Foreign innovation x Business group affiliation -0.048 (0.655) -1.425* (0.805)
Observations 982 982 982 982 982 982 982
Adjusted R² 0.427 0.427 0.426 0.427 0.43 0.426 0.431
Note: Standard errors are in parentheses
n = 987 * p < 0.10 * p < 0.05 * * p < 0.01
Table 4:(2) Business group only
Ordinary least squares regression on Firm Performance (ROTA)
Variables Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7
BG age 0.036 (0.035) 0.040 (0.038) 0.036 (0.035) 0.036 (0.036) 0.047 (0.041) 0.058 (0.037) 0.085* (0.042) BG size -0.095 (0.067) -0.102 (0.070) -0.095 (0.067) -0.103 (0.083) -0.116 (0.077) -0.223** (0.087) -0.305** (0.106) Firm age -0.004 (0.039) -0.007 (0.040) -0.004 (0.039) -0.003 (0.039) -0.014 (0.044) -0.005 (0.039) -0.027 (0.042) Firm size 0.640* (0.297) 0.597* (0.327) 0.640* (0.297) 0.649* (0.303) 0.609 (0.328) 0.839** (0.319) 0.911** (0.323) Leverage -0.057 (0.186) -0.052 (0.187) -0.057 (0.186) -0.057 (0.187) -0.052 (0.187) 0.208 (0.187) 0.220 (0.187) R&D intensity -0.077 (0.450) -0.073 (0.451) -0.077 (0.450) -0.091 (0.458) -0.077 (0.452) -0.016* (0.026) 0.945 (0.489) Lagged ROS 0.016 (0.026) 0.015 (0.026) 0.016 (0.026) 0.016 (0.026) 0.015 (0.026) 0.969 (0.489) -0.018 (0.026) Innovation 0.018 (0.057) 0.116 (0.489) -1.866** (0.622) -2.019** (0.632) BG diversity -0.095 (0.067) -0.116 (0.077) -0.305 (0.106) BG export intensity 0.001 (0.164) -2.162** (0.743) -1.881* (0.771) Innovation x BG diversity 0.112 (0.259) 0.368 (0.277)
Innovation x BG export intensity -10.369** -10.775** (2.126)
Observations 277 277 277 277 277 277 277
Adjusted R² 0.620 0.620 0.620 0.620 0.620 0.654 0.656
Note: Standard errors are in parentheses
5.3 Robustness Test
6 DISCUSSION
This study sought to address the research question whether and when business groups, that are of central importance within a majority of emerging market economies, lead to better firm performance levels when engaged in innovative activities. To address this specific research question the empirical context of the Indian drug and pharmaceutical industry was chosen, as it consists of numerous business groups as well as independent firms that have directed the spotlight onto a market that is accountable for a remarkable deal of innovative products and has gained both immense attention and reputation between the period of our study, 2005-2011. In order to control for the differences between independent firms and business group-affiliated firms, this study has conceptualized two theoretical models with the first one comparing (1) business group-affiliated firms to non-business group-affiliated/independent firms and the second one (2) only considering business group firm level data.
changes and conditions as well as potential competitors within local markets, resulting in greater performance levels (Khanna, Palepu & Sinha, 2005).
Second, as for the second model focusing on the business group level only, the outcomes for hypothesis H3 further unveiled that the diversity of business groups does not influence the relationship between innovation and firm performance. As this study has utilized the number of affiliates within a business group as a proxy for business group diversity, it can be concluded that the size of a business group has no impact on the relationship between innovation and firm performance. However, this finding contradicts the majority of the existing literature that argues for the negative impact of larger business groups on both innovativeness and firm performance outcomes by fostering conflicts in decision making processes and group intern strategies (Khanna et al., 2005; Mahmood & Mitchell). Consequently, it is recommended for further research to investigate this specific field in a more detailed manner.
Third, this study did not find supporting evidence for the moderating role of business group exports to positively influence the relationship between innovation and firm performance. In contrast, the results for hypothesis H4 revealed that among business group affiliates, exports have a negative impact on firm performance levels when engaged in innovation. This finding stands in contradiction to a number of literature that characterize exports as a way of organizational learning and a method to reduce or outsource the costs of innovation across own country boundaries (Cassiman & Golovko, 2011). A likely approach to explain this finding may be that the positive role of internationalization to foster innovation and firm performance outcomes may not be applicable in the context of business groups. As innovations practiced by firms in emerging markets are usually adapted to local institutional and environmental conditions that are radically different from that of foreign markets, exports may not be a good method for business group affiliates to acquire knowledge about specific innovation patterns that are based on a different approach than it is practiced in domestic markets. As a consequence, higher levels of business group exports might not lead to both improved innovative success and firm performance outcomes in the context of business groups.
domestic innovation or foreign innovation on respective firm performance outcomes. By doing so, this paper delivers insights on which type of innovations may result in greater performance outcomes within business group affiliated firms. Future research in the field of business group and innovation literature should consider the impact of business groups and innovation on other performance or profitability indicators to redress and resolve the lack of literature in this field. 6.1 Limitations
Even though this quantitative study has provided some critical evidence to the existing literature, the concept of this study underlies some specific limitations that need to be taken into account when discussing the results. For reasons of clarity and ease, this paper has only focused on one particular national context and industry. Although this may have facilitated finding reliable and complete data sets, the results may not be applicable and generalizable to the global business environment or specifically to other industries within a different emerging market economy other than India. Secondly, this study has only observed the time frame between 2005 and 2011. Other time frames may lead to different results. Thus, future studies should also investigate this research question by considering a longer period of time. Third, the independent variable of innovation was conceptualized and coded by considering a firm’s new product announcements. Since the given announcement information was often not clear enough, it was not always possible to explicitly distinguish whether a product has been introduced to the market or not and also whether the product has been launched within domestic country boundaries or is also simultaneously sold in foreign markets. As a fourth limitation, I would also advise future studies to use a bigger sample size for the analysis. Especially our second model focusing on (2) business groups only had a very limited sample size of 43 business groups in total. Extending the sample size may lead to more reliable and valuable results. Considering these aspects, it must be said that this study may not guarantee a full representation of the observed phenomenon and future research needs to be done to extend and re-examine this particular study.
6.2 Conclusion
