University of Groningen Master Thesis
To what extent does Collaboration with Non-traditional partners
affect firms’ Innovative Performance at Bottom of the Pyramid
markets?
Date: June 17, 2019
Word Count: 9780
Student:
Bakain Christina S3623254
Parklaan 26, 9724 AP Groningen
0610741808
c.p.bakain@rug.student.nl
Supervisor:
Dr. S.
Gubbi
Co-assessor:
Dr. H.U. Haq
List of Figures and Tables ...2
Abbreviations ...3
Chapter 1: Introduction ...4
Chapter 2: Theory and Hypotheses ...7
2.1 The BoP Environment ...7
2.2 Strategic Alliances in the BoP ... 11
2.2.1 Types ... 11
2.2.2 Incentives and Partners ... 13
Chapter 3: Methodology ... 17 3.1 Data Breakdown ... 17 3.2 Operationalization... 18 3.2.1 Dependent Variable ... 18 3.2.2 Explanatory Variables ... 19 3.2.3 Control Variables ... 20 3.2.4 Model Construction ... 21 3.2.5 Variable Correlation... 22 Chapter 4: Results ... 24
Chapter 5: Conclusion and Discussion ... 26
5.1 Limitations and Future Research ... 28
1
Abstract
While the markets of developing economies are becoming highly saturated, MNCs ought to shift towards emerging economies (EE). Specifically firms should invest in the bottom of the pyramid (BoP) markets as these will assist in leveraging their innovation capabilities and improve their performance and competitiveness. Nevertheless this is not an easy task for individual multinational corporations (MNCs) as the BoP is an unknown environment that conceals a multitude of institutional challenges and huge entrepreneurial opportunity at the same time. One strategy that MNCs are advised to follow in order to survive in this complex context is to forge alliances with other parties. However the success of the collaboration is determined by the nature of these partners. Collaborations may be devised with traditional partners, which are given as larger domestic companies that commonly participate in the formal economy, perceive the Western capitalistic system as legitimate and value its products and services, or with non-traditional partners, which are NGOs, local communities and groups, governmental authorities and small local companies. This thesis investigates how
2
List of Figures and Tables
Figure 1: The Indian Market as a Pyramid
Figure 2: Theoretical Framework
Table 1: Sample Characteristics Country of Origin
Table 2: Sample Characteristics Industry of Origin
Table 3: General Information of the Variables
Table 4: Correlation between the Variables
Table 5: Descriptive Statistics of the Variables
Table 6: Goodness of Fit of Negative Binomial Model
Table 7: Omnibus Test of Negative Binomial Model
Table 8: Tests of Model Effects
Table 9: Parameter Estimates of Negative Binomial Regression
3
Abbreviations
BoP- Bottom of the Pyramid
EE- Emerging Economies
NGO- Non-governmental Organisation
MNC- Multinational Corporation
4
Chapter 1: Introduction
Nearly 4 billion people in the developing world live on per capita income of less than $1,500 per year which in some cases translates to daily income of less than one dollar (Prahalad, 2011). This
population constitutes markets that have been given the name Bottom of the Pyramid (BoP) and exist within emerging economies (EE) (Prahalad, 2011, Anderson and Billou, 2007). The BoP markets offer a lucrative environment for business development due to the immense untapped potential that lies there which has been largely unexploited all these years (Prahalad and Hammond, 2002).
Research by Prahalad (2011) and Eyring, Johnson and Nair (2014) suggests that the environment in these markets is generally favorable for creative experimentation and exploration and as a result, BoP markets have been characterized as innovation laboratories.
To be more specific, firms can have the opportunity to create novel ideas by confronting the unique conditions of the BoP and surviving under the pressure of these markets. If successful, firms that achieve to cope with the disheartening circumstances of these markets will harvest the first mover advantages which include reputation, access to new product markets and domination of major distribution and communication channels (Hoskisson et. al., 2000). A study by Isobe, Makino and Montgomery (2000) finds that first movers and technology leaders indeed attain superior
performance. Further, it is widely accepted in the literature that innovation is an essential driver of competitiveness of the company, as well as a factor that determines growth objectives, survival and prosperity of the firm (Gaynor, 2002, Cefis and Marsili, 2006, Krugman, 1979). Multinational
companies need to grasp that essentially by investing in the BoP not only do they enable consumers to have access and participate in the markets by either consuming, distributing, laboring or being entrepreneurs themselves, they also increase their own growth and profit potential by enabling themselves to innovate effectively in a diverse, alternative environment (Prahalad and Hart, 1999, Prahalad and Hammond, 2002).
5
Such differences both constrain and facilitate strategies of entry for corporations that attempt to venture the BoP markets (Hoskisson et. al., 2000, Khanna and Palepu, 2005). In these conditions, companies are better off by externally collaborating with multiple players to sell and to achieve growth objectives in these markets. This is because in large EE the business environment is unknown and the institutions are complex and hard to understand and decipher, hence the need for creating networks with other parties arises, which can be firms or non-firms (Rondinelli and London, 2003). In order to compose an appropriate strategy and achieve the desired competitive advantage it is crucial for these markets to establish an alliance network or alliance portfolio, which is essentially the cooperation of two or more firms, which substantially facilitates knowledge sharing and grants easier access to resources (Lavie, 2007). Companies’ resources and knowledge may be largely
complementary, therefore jointly pooling and coordinating them creates a network that enables firms to achieve economies of scale and scope and allows organizational learning to occur (Rondinelli and London, 2003). Therefore, network contracts, personal ties and networks may be used as a hybrid strategy and may significantly reduce the uncertainty in the context of the BoP (Peng and Heath, 1996).
However, firms need to go beyond selecting traditional partners as these may not prove appropriate to leverage the full potential of the BoP markets. This is because firms that wish to improve their innovative performance may need to explore new knowledge sources emerges and seek new partners outside existing and familiar networks (Duysters et. al. 2011). These new non-traditional partners are usually identified as governmental authorities, local companies, nongovernmental organizations (NGOs) as well as local communities and groups (Hahn and Gold, 2014, Prahalad and Hart, 2001). Based on the above premise, the aim of this study is to investigate to what extent collaborative partnerships with nontraditional partners in the BoP offer greater innovative
performance, which is defined as the rate of introduction of novel products, processes and business models in the market (Freeman and Soete, 1997). The question that this thesis will attempt to answer is “To what extent does collaboration with non-traditional partners affect firms’ innovative
performance at Bottom of the Pyramid markets?”.
6
horizontal, depending on the needs of the company, the network position of the partner, the resources available and the formal and informal institutions of the country.
The empirical context of this paper refers to western top MNC performers catering the needs of the BoP population. The sample comprises of 120 MNCs chosen based on a list published by Forbes in 2017. The firms had to abide to criteria that are presented in the London and Hart (2014) paper and the variability of the sample is ensured since no more than thirteen percent (13%) of the sample originates from the same country and no more than ten percent (10%) of the sample originates from the same industry. The proposed first hypothesis states “Firms that collaborate with non-traditional
partners have higher innovative performance than firms that collaborate with traditional partners, when catering to BoP markets in developing countries”. This hypothesis is tested by regressing
innovative performance on collaborations with traditional or nontraditional partners, while
controlling for size, age and industry type. The second hypothesis states “Firms that have high RnD
intensity will have higher innovative performance than firms that have low RnD intensity, when catering to BoP markets in developing countries”. This is tested by regressing innovative performance
on RnD intensity given as a percentage of the total RnD expenditures over the value of sales of an MNC.
Finally, this thesis finds support for the claim that collaborating with non-traditional partners will contribute more to a firm’s innovative performance than collaborating with traditional partners. However the results are insignificant and thus do not support that firms with higher RnD intensity will have higher innovative performance as opposed to firms with lower RnD intensity. These results confirm that the BoP can be a lucrative environment for investment and flourishing business if the potential entrees select nontraditional partners and achieve efficient collaboration with them. Notably, MNCs need to utilize all resources, novel ideas and business models that emerge from these collaborations and invest cautiously in the development of new products and solutions for the BoP, that could possibly be readjusted to the needs of the developed world as well.
7
Chapter 2: Theory and Hypotheses
2.1 The BoP Environment
The Bottom of the Pyramid (BoP) market is the setting of this research and it is an increasingly popular topic nowadays. BoP markets and are characterized by low income, continuous and rapid growth of economic development and governmental policies that favor liberalization and free-market system (Hoskisson et. al. 2000). The pioneers in the research of the BoP free-market were C.K. Prahalad and Stuart Hart who, through their article (1999) and subsequent book “The Fortune at the Bottom of the Pyramid” (2001), exhibit how these low income markets constitute huge untapped opportunity for the multinational companies. In the face of this new opportunity firms realize the challenges that these markets offer as strategies that may have been profitable for developed markets will be considered old and tired in this modern context of development. As a result,
products and services that are traditionally served at top of the pyramid markets are not suitable for BoP markets for a multitude of reasons. This is because products and services that are traditionally sold to the developed markets are not affordable, not accessible and correspond to different tastes and different needs than those of the developing world (Prahalad, 2011, Eyring, Johnson and Nair, 2014, Pitta et. al., 2008). Therefore, this constitutes a challenge for companies and managers to think beyond conventional models and processes and create inter-firm networks in order to jointly face the challenges of the BoP through knowledge sharing and bundling of capabilities (London and Hart 2004).
For example, India and China are countries that have a high likelihood of large existence of BoP and will be used to illustrate the bigger scene of the BoP markets. These are typical EE comprised of culturally unique regions and therefore each region is assumed to have distinct characteristics that can be used to generalize over the whole BoP (Dheer et. al., 2015). India and China are such immense countries that provinces may retain their own unique customs, traditions and dialects, so there is considerable cultural variation even within the same country which makes it complex to understand and be familiar with all different types of institutions (Parmigiani and Rivera-Santos, 2015).
In this manner, the literature has identified certain general traits of the BoP markets that can be used to define them and the use of specific country examples will facilitate this procedure. Firstly,
8
is of crucial importance in the BoP markets. This is the case because it is either too costly, since excessive interest rates are charged which makes it hard for small and medium enterprises to get access to credit, or too time consuming, since the bureaucratic system often has ineffective lars or is corrupted and too extensive, to participate formally in the economy (London and Hart, 2004). Another common issue that can be noted is that companies that serve these markets offer very limited quality and quantity of products. These firms assume that due to the low income of the population, consumers in these markets do not hold or use significant purchasing power and therefore leave them be a responsibility of NGOs or the government (Prahalad and Hart, 1999, Prahalad, 2002).
9
Hart, 2004, Khanna and Palepu, 1997). Therefore MNCs may find themselves in need of assistance to be able to not only survive but also to take advantage of the enormous promising potential of the BoP market. This assistance, as it will be discussed, may come in the form of alliances with external partners.
Moving on, to identify the nature of the entrepreneurial opportunity in BoP markets it is essential to understand the structure and properties of these markets viewed as a pyramid. To illustrate this, a diagram of the Indian market is depicted in Figure 1, where the classic modern model of Western multinational is usually serving the consumers in Tier 1. The top tier is comprised mainly by affluent population and can be classified as the elite of the population. As illustrated, this layer has the lowest number of consumers, only 5 to 10 million only in India, but is the main focus of business managers and entrepreneurs of large companies (Prahalad, 2002). The second and third tiers constitute the middle class and amount to roughly 200 million in India. These tiers have 3 to over 10 million dollars yearly income and concern some organizations and mainly local firms but, without necessarily being a priority to these firms as they consider that layers 2 and 3 do not entail a great opportunity for them. The last two levels, tier 4 and 5, contain the greatest amount of the population, almost 700 million of the Indian BoP market , with purchasing parity spanning from 3 up to less than 2 million dollars yearly.
Figure 1: The Indian Market as a Pyramid
10
As a result, it can be deduced that MNCs don’t respond adequately to the demands of the largest share of the developing world, which comprises the BoP markets at the last 2 tiers of the respective world pyramid. This situation per se, shows that the Western corporate sector is incapable of taping the potential in these layers. Individual MNCs and their regular strategies cannot cater the needs of the BoP market as these largely differ from the needs they usually satisfy. To summarize, EEs are significantly heterogeneous with a highly diverse institutional landscape even within the same country, both formally and informally (Hoskisson et. al., 2000). To overcome these challenges, London and Hart (2004) suggest that partnerships, networks and alliances constitute the ideal mean to develop mechanisms to adjust and cope to the conditions posed by this environment. Through alliances western MNCs will gain a better understanding of what the population of the BoP wants and needs from their products and services. Alliances can further facilitate the MNC to assimilate to a foreign network of firms and entrepreneurs which therefore gain access to specific resources,
11
2.2 Strategic Alliances in the BoP
2.2.1 Types
12
Another general alternative categorization is vertical against horizontal alliances, which is essentially partnerships between buyer and supplier and partnerships on the same value level, for example between suppliers, respectively (Kale et. al., 2002). These alliances are particularly important in the context of entering BoP markets, as through these alliances MNCs can gain access to valuable resources and are allowed to claim competitive advantage. Specifically, according to the knowledge-based view, the continuous flow of knowledge within the organization will assist the firm to evolve a capability of managing knowledge flows thus to attain competitive and sustained competitive advantage (Dyer and Noboeka, 2000). In the cases of vertical alliances, networks of firms are formed as a result of repeated transactions, such as the Toyota network, with the purpose of sharing resources, technology and knowledge. This procedure can be complex though as network settings are facing knowledge-sharing dilemmas such as to adequately motivate members to participate and share openly, while avoiding negative spillovers, to prevent free riders and to reduce the costs of identifying and accessing new knowledge (Dyer and Noboeka, 2000). For horizontal alliances, a study by Wilhem and Sydow (2018) reveals the paradox of “coopetive” tensions that can evolve in, for example, triadic relationships comprised of partners where there is one buyer and two suppliers that are collaborating and being competitors simultaneously. The Toyota network again offers an
excellent example of managing such a paradox, as by continuously suggesting cost reductions to their suppliers it creates joint value and by accepting the coexistence of this double role the best
13
2.2.2 Incentives and Partners
Strategic partnerships are increasingly gaining popularity in the literature and that is because a multitude of incentives can be enumerated to enter a collaboration in order to invest in the BoP. Some of the reasons to enter any type of collaboration with a nontraditional or a traditional partner are sharing costs and uncertainty, access and learning of complementary and tacit technologies, monitoring of environmental changes, entering to foreign markets and reducing innovation cycle (Hagedoorn and Schakenraad, 1990, Meyer, et. al., 2009). These last two points are the main focus of this research because initially many EMs have weak market-supporting institutions and firms are in need of local resources to be more competitive, which leads to alliances commonly used as a form of entry to such markets (Meyer, et. al., 2009, Webb et al., 2010, De Mattos, 2002).
The BoP markets offer a plethora of external and unexplored knowledge in the form of local partners and consumer needs and preferences, which has been increasingly recognized as a source of
innovation of a firm (Duysters and Lokshin, 2011, Prahalad and Hart, 2001, Prahalad 2011). The ideal partners are not large established multinational companies anymore, but rather smaller domestic companies, governmental authorities, local communities and groups, as well as NGOs which are partners that constitute a better partnering option for a firm that is in need of local resources and knowledge of relational capabilities and institutions (Collinson and Narula, 2014, Hahn and Gold, 2014). In a research by Gebauer and Reynoso (2013), these partners are characterized as non-traditional partners and they are highly associated with driving and creating innovation at the BoP market. Alliances with NGOs, government agencies and local community groups are called cross-sector alliances and these are substantially different from the collaborations discussed above (London and Rondinelli, 2003). This is due to the largely differing missions of the separate partners and the different governance structures, which may cause issues with alignment of interests and mutual understanding of the cultures of the partners (Gutiérrez, 2016, London and Rondinelli, 2003). Nevertheless, if the complexity of the portfolio of alliances of an MNC and the intricate conditions of the current institutions of the BoP environment are managed appropriately, the exploration of the BoP market may be transformed to a highly creative procedure which can to increase the innovative performance of the focal firm (Prahalad and Hammond, 2002, London and Rondinelli, 2003, Dyer and Noboeka, 2000). This procedure can be facilitated if MNCs decide to partner with the
14
One issue that is of major significance when a firm decides to achieve strategic objectives access in any market through collaborative ventures, especially when it is with nontraditional partners, is identifying the suitable partner and managing as well as maintaining the alliance (Dyer and Noboeka, 2000). Originally, as per the existing literature, the focal company will look at criteria such as
technology, managerial capability and international experience of the partner (Hitt et. al., 2004, Nielsen 2003, Roy and Oliver, 2009). However, in such a multiplex environment the firm ought to pay attention to alternative criteria such as the network position of the partner and how this firm’s network may offer complementarities to the focal company (Shi, et. al., 2012, Carayannis, 2000). Subsequently, in order to upgrade performance a firm must have re-combining or bundling
capabilities that allow the company to identify and select partners with complementary firm-specific assets and a certain degree of reciprocity (Collinson and Liu, 2017). Another non-traditional criterion that is critical for the entrance, adaptation and survival of the MNC is the knowledge of within-country institutional variations and relational governance, which will reduce liability of foreignness for the focal firm and avoid issues with misinterpretation or misconception of institutions (Shi, et. al., 2012, Collinson and Liu, 2017). Hence, MNCs ought to impose contemporary partner selection methods which will more likely lead them away from the traditional established companies and towards collaboration with nontraditional parties that have an established network and good knowledge of sub-country institutional variations.
In light of the above considerations, each collaboration has distinct characteristics regarding the type of alliance, the purpose of it and the partners themselves thus, depending on various factors,
15
To be more precise, as development at the BoP does not follow the conventional patterns of the developed world, non-traditional partners offer greater advantages than traditional partners, because the later players have a very limited outlook on what is effective and appropriate (London and Hart, 2004). Actors such as NGOs and other socially oriented organizations and institutions have the capacity to understand better the pace of business development in the developing countries. For example, Grameen Bank and its wholly owned subsidiary Grameen Telecom launched a pilot
program that allowed women members of the Grameen Bank to retail telecommunication services in rural areas of Bangladesh. The completion of this program offers employment to 40,000 people with the combined yearly net income of $24 million USD. Also, the benefits of Grameen Telecom’s Village Phone program include the enhancement of social and economic development in these areas, as well as provide valuable knowledge regarding how can private sector development significantly
contribute to the reduction of poverty (Richardson et. al., 2000). The aforementioned example illustrates a case of a company that recognized the growth and revenue potential in a market where other companies neglected. Hence the firm combines commercial and non-for-profit operations to successfully realize the potential for innovation and development by partnering with the community of women in the rural Bangladesh. Thus the following hypothesis is proposed:
Hypothesis 1: “Firms that collaborate with non-traditional partners have higher innovative
performance than firms that collaborate with traditional partners, when catering to BoP markets in developing countries.”
In general, due to the increasingly integrated and knowledge-centered nature of the global business scene it is of fundamental importance to ensure competitiveness by continuously generating novel and better products, services or practices (Hu, 2001). When it comes to new product development and product introduction to the markets such as the BoP, which is directly linked with a firm’s long term innovative performance, firms will face numerous uncertainties and challenges. These depend on the specialization and available legal infrastructure of each country, as well as the competition and network of each firm, which lead MNCs’ to vary in their decisions to invest in developing new products and services (Viscusi and Moore, 1993). Especially when it comes to operations in the BoP, which is an environment that requires frugal solutions to meet the demands, MNCs cannot gamble with their investments in RnD. One way to explain this is by thinking that in case of product liability the MNC is directly exposed to severe costs and impaired reputation (Spence, 1977). Principally, RnD is an essential component of most innovative firms, especially the ones that are operating in
16
and demands the MNC should be cautious when balancing expenditures on RnD against innovative performance to avoid negative reputation and ensure competitive advantage (Spence, 1977). Firms in the BoP cannot afford to invest too much in conventional RnD as this will increase the cost of products and solutions. Therefore, firms will have to be very careful and thrifty in terms of investment in RnD in an unexplored market such as the BoP.
One way to measure RnD is to take the total amount of expenditures in RnD and divide it by total sales which will generate an index called RnD intensity. When inspecting the literature it can be remarked that MNCs’ capacity to invest in RnD will increase the rate of introduction of new products and services to the market, which will improve their overall innovative performance (Simai, 2003, Spence, 1977). This will also be apparent in the developing world, when the focal company is thriving through a collaboration with a nontraditional partner. Hence, it is suggested that a firm’s spending on RnD will overall prove to be beneficial to the performance of the firm that is operating in the BoP and is likely to lead to increased capability of innovation. Thus, it can be assumed that the higher intensity of RnD will likely lead to higher innovative performance. Thus in this paper the following hypothesis is introduced:
Hypothesis 2: “Firms that have high RnD intensity will have higher innovative performance than firms that have low RnD intensity, when catering to BoP markets in developing countries. ”
After introducing the three key hypotheses that this paper will empirically explore, the general theoretical framework of this thesis is illustrated in the diagram of Figure 2 below.
17
Chapter 3: Methodology
3.1 Data Breakdown
To pursue the research objectives of this study the secondary data is drawn on the MNC level from the database of Orbis. In addition, primary data collection is executed using content analysis to transform qualitative data into quantifiable data to allow empirical investigation of the research question posed.
Initially, the companies to be observed had to abide by the standards of the London and Hart (2004) paper, which portrays the model of a large classic western MNC attempting to enter the BoP. Further, the firms were chosen based on a list published by Forbes ranking the 2017 top MNC performers, reassuring that all MNCs belong to EU, US or Canada. Out of this list the best candidates were selected based on the conditions described in the London and Hart (2004) paper, while
excluding the inapplicable candidates. To determine that the selected MNC has entered or will attempt to enter the BoP, the size and the industry type contribute crucially. For example, it is more likely that a big MNC in the food processing industry, like Nestle, will attempt to serve the BoP than a semiconductor MNC like ASML holding. Nevertheless, after selecting the companies, their actions were cautiously researched to determine which companies entered the BoP. The year of observation is 2017, which is not ideal as this assumes simultaneous formation of partnership and innovation performance, but was selected because it is recent enough and due to the availability of the annual reports.
The sample comprises of top multinational performers and the industry sectors that they represent vary from low innovative capability such as diversified metals and mining, specialized/diversified chemicals, construction materials semiconductors and others, to higher innovative capability such as pharmaceuticals, food processing, household, personal care, as well as auto and truck manufacturers (Frishammar and Horte, 2005). Likewise, the countries of origin of the MNCs are quite diverse, ranging from a multitude of European countries to firms from the US and Canada. No industry or country of origin represents more than ten and thirteen per cent, respectively, of the collected sample and the vast majority of the sample was large MNCs, meaning that they have above five hundred employees.
However, this cannot be assumed to be a perfectly random sample, as a perfectly random sample is very difficult to achieve practically and the sample is already restricted to large, western MNCs that do not originate from the EEs. Nevertheless, the above practices ensure a certain degree of
18
from which the sample was derived (Miller and Friesen, 1982). For a more detailed description of the characteristics of the sample see Table 1 and 2 Annex 1.
Finally, the annual reports of initially panel sample size of 120 MNCs were carefully inspected to extract, categorize and code all the necessary data to collect and produce a valid sample, in order to conduct an empirical analysis. However some of the values were not available when extracting the secondary data from the Orbis database which restricted the sample size. Henceforth, after
accounting for missing values the final sample constitutes of 102 companies out of the 120 originally collected sample of MNCs. This decision is further supported by preceding research by Miller and Friesen (1982) who conducted research utilizing a similar sample size. Finally, the procedure of operationalization of the variables is elaborated in the following sections.
3.2 Operationalization
3.2.1 Dependent Variable
19
spreadsheet. This set of coding rules is applied to all text reports. A similar data collection strategy was also used by Rondinelli and London (2003).
3.2.2 Explanatory Variables
Following that, this research paper poses two independent variables of interest, namely these are collaborations and RnD intensity. The first independent variable, which is collaborations is a discrete dichotomous variable. Collaborations is a dummy variable taking the value of 1 if the company under observation has partnered with a non-traditional partner, as these are defined in section 2.2 section above, and taking the value of 0 if the MNC hasn’t had any such collaborations. The procedure to identify whether the partnership is with a traditional or nontraditional partner is depicted in an example in Annex 2. The collecting of the data for this variable followed a similar pattern as for the process for the dependent variable and was executed using content analysis. The same text units were utilized to extract the data, which are the 2017 annual reports of 120 chosen western MNCs. The procedure encompasses setting two categories, particularly these are Traditional Partnership and Non-traditional Partnership, in which the collected elements will be classified. Then the reports are searched for the words “joint”, “alliance”, “cooperation”, “collaboration” and “partner” to identify partnerships of the company with other parties. To differentiate between traditional and nontraditional partners the mixed resulted partners were individually researched to determine the nature of the organization found, i.e. if traditional or nontraditional (see Annex 2). In order to make an appropriate choice, the coding rules are fixed which in this case imply that the types of
collaborations include but are not limited to any types of research oriented agreement, strategic agreement, equity purchase agreement and joint venture. All mergers and acquisitions that the companies materialized during the study’s examined period are excluded from this variable. Lastly, keeping in mind of the context of the text, the partnerships are assigned to the correct category and the text is then coded to 0 and 1 values. The same coding rules apply to all text units analyzed (Boyd et. al., 2005, Rosenberg et. al. 1990).
20
3.2.3 Control Variables
Following previous work, this model incorporates three control variables, of which two are continuous discrete variables and one is a categorical variable. Initially, this study controls for firm size, which is operationalized as the logarithm of the total number of full-time employees in the fiscal year 2017 (Un et. al., 2010). Thus, this variable cannot be negative or take the value of 0. This
variable is frequently used in many similar researches and it is one of major importance as it is generally assumed that the larger the firm the higher the capability it has to invest in RnD and generate more innovation (Shi, et. al., 2012, Zhou, 2012, Cohen, 1996). To obtain the data for firm size, Orbis database was used, from which the number of employees was readily derived as a numerical value from which then the logarithm was taken.
The second control variable of this research is age of the focal firm, which is another non-zero and non-negative variable that can only take the form of integers. This is frequently used as a control in the literature that has been utilized in multiple similar studies (Collinson and Liu, 2017, Zhou, 2012). This is because it is expected that firms that are younger in age will be more inclined to form partnerships as they are facing a liability of newness that makes them prone to look for collaborations (Freeman and Hannan, 1989).
The third control that this study accounts for is the industry type and this is measured by setting up a dummy variable that determines whether the nature of the industry of each focal company is
21
3.2.4 Model Construction
After the variables have been explained and adjusted, a regression analysis will be performed to test the predictions stated in the hypotheses of the previous section in order to deduce inference. The specific analytical technique is chosen due to the nature of the empirical data. To be more precise, when the outcome variable is observed it can be seen that the collected data may fit the
assumptions of the Poisson regression (Kennedy 1998, Greene 2003). Thus given the predictor variables, the dependent variable which is innovative performance is measured in finite count outcomes which requires the use of Poisson regression. In this case, the mean count of the outcome variable is 9,39 which is less than 10 and thus, it is a small enough value that allows for Poisson model. To continue, in this model it is assumed that one or more of the independent variables is measured in a nominal scale. At first glance this assumption is met as the main explanatory variable is collaboration which is a categorical variable measured in 0 and 1 for traditional and non-traditional partners respectively. Another assumption of major importance for the Poisson regression is that the mean and the variance of the model are identical, which results in equidispersion (Bianco et. al., 2013). Thus,
𝑉𝑎𝑟(𝐼𝑛𝑛𝑜𝑣𝑎𝑡𝑖𝑣𝑒_𝑝𝑒𝑟𝑓𝑜𝑟𝑚𝑎𝑛𝑐𝑒) = 𝐸 (𝐼𝑛𝑛𝑜𝑣𝑎𝑡𝑖𝑣𝑒_𝑝𝑒𝑟𝑓𝑜𝑟𝑚𝑎𝑛𝑐𝑒) = 𝜇
This essentially means that the allocation of counts follows a Poisson distribution, meaning that the observed and expected counts are almost identical. Theoretically, innovative performance can take the value of any integer number infinitely. However, after performing nonparametric tests on the econometric software and checking the descriptive statistics of the dependent variable it was observed that the sample appears to be overdispersed (see Table 10, Annex 5). This is because the mean and the variance vary greatly, they are 9,39 and 49,2 respectively. This can be noticed if a One-Sample KS test is run on the dependent variable to check whether the response variable is following a Poisson distribution. Henceforth, with such overdispersed data it is suggested to use the Negative Binomial Regression. An illustration of the model specification is provided below as it is constructed in accordance to the theory. The subscript i suggests individual MNCs , whereas εi is given as the standard error.
𝐼𝑛𝑛𝑜𝑣𝑎𝑡𝑖𝑣𝑒_𝑝𝑒𝑟𝑓𝑜𝑟𝑚𝑎𝑛𝑐𝑒𝑖
22
3.2.5 Variable Correlation
Initally, Table 3 in Annex 3 offers an efficient summary of all variables operationalized above as well as some of their key characteristics and where they are obtained from. The elements of this table will be used to compute the regression analysis in the upcoming section.
Next, after defining and describing all the variables that will be the components of the regression analysis, the correlation between these variables will be evaluated. This was conducted by running a Bivarieta Pearson’s correlation analysis. Initially it can be observed that there were no high
correlations between the variables, which makes all the variables suitable to be added in the regression. Moreover since no correlation coefficient has a value greater than 0.7, there is no evidence of multicollinearity between the variables.
More specifically, there is relatively high positive correlation between the dependent variable, innovative performance and collaboration (r=0.485, p<0.001) and it is the only significant correlation. Then, there is weak positive correlation between size and innovative performance (r=0.148,
p<0.001). Also there is very weak positive correlation between RnD intensity and innovative
23
Table 4: Correlation between the Variables Variable 1 2 3 4 6 7 1.Collaboration 1.00 2.Innovative Performance 0.485** 1.00 3.RnD Intensity -0.030 0.047 1.00 4.Size 0.170 0.148 0.013 1.00 5.Industry Type 0.056 -0.022 0.055 0.91 1.00 6.Age Experience 0.040 0.018 0.133 0.138 -0.31 1.00
**. Correlation is significant at the p<0.05.
Moving on, to better summarize the collected data set, Table 5 below will provide some essential descriptive statistics which include measures of central tendency and variability or spread of the sample. The first column provides the number of observations, while the second and third column give the minimum and maximum of each of the variables, followed by the mean and the standard deviation.
Table 5: Descriptive Statistics of the Variables
N Minimum Maximum Mean SD
24
Chapter 4: Results
This dissertation follows a deductive approach and builds a Negative Binomial regression to test whether collaborations with non-traditional partners positively affect western MNCs’ innovative performance at the BoP market, as opposed to collaborations with traditional partners. To perform the data analysis, the software IBM SPSS Statistics 24 will be used. After the data spreadsheet is entered to the appropriate software, the SPSS regression output is interpreted.
Initially the Goodness of Fit table is provided in Table 6 of Annex 4, from which it is observed that the ratio of Value to degrees of freedom is satisfying since this ratio should give the value of 1 so that the assumption of equidispersion is not violated. Looking at the “Value/df” column for the row “Pearson Chi-square” this value is very close to 1, which, as identified above, indicates overdispersion. This result indicates greater variability and statistical dispersion in the dataset which is not ideal for the given model.
Moving on, the next table provides the Omnibus Test (Table 7, Annex 4), which essentially calculates a ratio that explains whether the independent variables overall improve this model, as opposed to the intercept-only model which doesn’t contain any predictor variables. Here it can be noticed that the model indicates statistical significance, since the p-value is 0.00. Now it is acknowledged that the independent variables collectively contribute to the statistical significance of the model and
therefore improve it, it remains to be seen which of the independent variables are statistically significant.
25
Table 9: Parameter Estimates of Negative Binomial Regression
*Coefficient is significant at p<.10
Furthermore, looking at the incidence rate ratio from the Parameter Estimates table generated by the software, it can be deduced that the rate of innovative performance will increase by 0,2 times from each extra partner that a given firm collaborates. Alternatively, there is a 20% increase in the rate of innovative performance for each new partner that the company gets.
In this fashion, Hypothesis 1 suggested that firms that collaborate with non-traditional partners have higher innovative performance than firms that collaborate with traditional partners. In the Negative Binomial model there is statistically significant support for hypothesis 1.
Hypothesis 2 claims that firms that have high RnD intensity will have higher innovative performance than firms that have low RnD intensity. The result for variable in the model is statistically insignificant which provides no support for hypothesis 2.
Dependent Variable: Innovative Performance
Independent Variables Model 1 Model 2 Model 3 Model 4
26
Chapter 5: Conclusion
and Discussion
This thesis has sought to shed light on the impact of collaborative partnerships between MNCs and traditional as well as non-traditional partners on the consequent innovative performance within the multiplex context of the BoP. Within a fast paced era of development, innovation is a major driver of competitiveness and essential for the survival of the firms. Furthermore, from the inspection of the relevant literature it can be deduced that the BoP is a highly demanding environment that according to recent work can assist the MNC become more competitive through the use of non-traditional partners (London and Hart, 2004, Karnani, 2007). This paper empirically analyzes how MNCs form collaborations with various types of partners and how the nature of each of these partnerships affects performance related to innovation. For that reason a negative binomial regression was run with a diverse sample of 120 large western MNCs to predict how innovative performance changes with differences in collaborations and RnD.
The aim of the first hypothesis was to discover whether non-traditional actors will significantly increase a western MNC’s innovative performance as opposed to traditional partners. To test the first hypothesis the innovative performance variable, is regressed on the explanatory variable, which is the collaborations of the MNCs with either non-traditional partners or traditional, given as a dummy variable. The first group of partners is identified to be mainly NGOs, local communities and groups, governmental authorities, as well as small local companies, whereas the second group is given as a small subset of larger domestic companies that commonly participate in the formal economy, perceive the Western capitalistic system as legitimate and value its products and services (London and Hart, 2004, Rondinelli and London, 2003). The expected relationship between the two factors was positive and significant, which translates to higher number of collaborations with non-traditional partners will lead to higher innovative performance. This outcome was achieved since the results for the first hypothesis appeared to be positive and significant. Nevertheless this result may have been altered if there was access to more data concerning the new product development and upgrades of the selected MNCs. This is because each firm decides to include different amount and type of data in their yearly reports. Often, this results in unreliable content analysis of the reports which eventually translates into faulty or low degree of innovative performance of the firm. If there had been access to a more reliable database that contained a complete and detailed archive of the announcements of innovations this study’s inference may have varied.
27
verifies that the presence of strong emphasis on RnD indicates higher innovative performance as per Frishammar and Horte (2005). To understand this better, one should consider the importance of size of the MNC. Specifically, larger firms can exploit economies of scale, have the financial resources to undertake bigger and riskier RnD projects and the marketing and distribution channels that allows them market penetration abilities which in turn assists to achieve dominance in the BoP related to innovation (Corsino et. al., 2011). On the other side of the coin, smaller firms may enjoy other types of advantages when engaging in innovation related activities. In firms with smaller size, the
28
5.1 Limitations and Future Research
Subsequently, this section will acknowledge limiting factors and suggest future research paths based on the research question posed.
Concerning the data collection process manually coded content analysis was utilized to convert qualitative text into quantifiable data. Nevertheless, it is suggested that human coding systems may not be as reliable as computerized systems (Short and Palmer, 2008). This is due to the fact that computers possess the ability to always apply the same rules in the exact same manner, while at the same time it eliminates the possibility of mistakes. On the other side of the spectrum, the
introduction of the human factor in the analysis adds the element judging according to the context and drawing inference according to cognition. This creates a trade-off between the quality of content analysis. In addition, content analysis, as opposed to similar methods such as surveys, is an
inexpensive and unobtrusive technique that utilized material that are already publicly available without violating privacy (Morris, 1994).
One major weakness of this paper is that the second independent variable does not necessarily fully match the context of the research. The variable RnD intensity includes all RnD data and it is therefore not specific to BoP. This was not a negligence but rather a compromise and this data was not
available in the scale of this study. A future researcher should proceed with gathering data on RnD expenditures and total sales that is explicitly for the BoP and the collaboration of MNCs in the context of the EMs, in order to achieve intuitively more accurate results.
29
Furthermore, another critical constraint of this research was time pressure, given the tight and limited schedule that all the activities of this research had to be managed.
30
References
Anderson, J. and Billou, N. (2007). Serving the world's poor: innovation at the base of the economic pyramid. Journal of Business Strategy, 28(2), pp.14-21.
Baum, J. A., Calabrese, T., & Silverman, B. S. (2000). Don't go it alone: Alliance network composition and startups' performance in Canadian biotechnology. Strategic management journal, 21(3), 267-294.
Bianco, A. M., Boente, G., & Rodrigues, I. M. (2013). Resistant estimators in Poisson and Gamma models with missing responses and an application to outlier detection. Journal of Multivariate
Analysis, 114, 209-226.
Boyd, B. K., Gove, S., & Hitt, M. A. (2005). Construct measurement in strategic management research: illusion or reality?. Strategic Management Journal, 26(3), 239-257.
Brettel, M., & Cleven, N. J. (2011). Innovation culture, collaboration with external partners and NPD performance. Creativity and Innovation Management, 20(4), 253-272.
Carayannis, E. G., Alexander, J., & Ioannidis, A. (2000). Leveraging knowledge, learning, and innovation in forming strategic government–university–industry (GUI) R&D partnerships in the US, Germany, and France. Technovation, 20(9), 477-488.
Cefis, E., & Marsili, O. (2006). Survivor: The role of innovation in firms’ survival. Research
policy, 35(5), 626-641.
Cohen, W. M. (1995). Empirical studies of innovative activity. Handbook of the economics of
innovation and technological change.
Cohen, W. M., & Klepper, S. (1996). Firm size and the nature of innovation within industries: the case of process and product R&D. Review of Economics and Statistics, 78(2), 232-243.
Collinson, S. and Liu, Y. (2017). Recombination for innovation: performance outcomes from international partnerships in China. R&D Management, 49(1), pp.46-63.
Collinson, S. C., & Narula, R. (2014). Asset recombination in international partnerships as a source of improved innovation capabilities in China. The Multinational Business Review, 22(4), 394-417. Corsino, M., Espa, G., & Micciolo, R. (2011). R&D, firm size and incremental product
innovation. Economics of Innovation and New Technology, 20(5), 423-443.
31
Dhanaraj, C., Suram, B. and Vemuri, P. (2014). Godrej Chotukool: A Cooling Solution for Mass Markets. Harvard Business Publishing, pp.1-12.
Dheer, R., Lenartowicz, T. and Peterson, M. (2015). Mapping India’s regional subcultures:
Implications for international management. Journal of International Business Studies, 46(4), pp.443-467.
Dussauge, P., & Garrette, B. (1999). Cooperative strategy-Competing successfully through strategic
alliances (No. hal-00706251).
Duysters, G., & Lokshin, B. (2011). Determinants of alliance portfolio complexity and its effect on innovative performance of companies. Journal of Product Innovation Management, 28(4), 570-585. Dyer, J. H., & Nobeoka, K. (2000). Creating and managing a high‐performance knowledge‐sharing network: the Toyota case. Strategic management journal, 21(3), 345-367.
Eyring, M., Johnson, M. and Nair, H. (2014). New business models in emerging markets. IEEE
Engineering Management Review, 42(2), pp.19-26.
Faems, D., Janssens, M., Madhok, A. and Looy, B. (2008). Toward An Integrative Perspective on Alliance Governance: Connecting Contract Design, Trust Dynamics, and Contract
Application. Academy of Management Journal, 51(6), pp.1053-1078.
Freeman, J., & Hannan, M. T. (1989). Setting the record straight on organizational ecology: Rebuttal
to Young. 425-439
Freeman, C., Soete, L. (1997). The economics of industrial innovation. Pinter, London.
Frishammar, J., & Åke Hörte, S. (2005). Managing external information in manufacturing firms: The impact on innovation performance. Journal of Product innovation management, 22(3), 251-266. Gaynor, G. (2002). Innovation by design: what it takes to keep your company on the cutting edge. Amacom.
Gebauer, H. and Reynoso, J. (2013). An agenda for service research at the base of the pyramid.
Journal of Service Management, 24(5), pp.482-502.
Greene, W. H. (2003). Econometric analysis. Pearson Education India.
32
Gutiérrez, R., Márquez, P., & Reficco, E. (2016). Configuration and development of alliance portfolios: a comparison of same-sector and cross-sector partnerships. Journal of business ethics, 135(1), 55-69. Hagedoorn, J. (1996). Trends and patterns in strategic technology partnering since the early
seventies. Review of industrial Organization, 11(5), 601-616.
Hahn, R. and Gold, S. (2014). Resources and governance in “base of the pyramid”-partnerships: Assessing collaborations between businesses and non-business actors. Journal of Business Research, 67(7), pp.1321-1333.
Hart, S. L., & Christensen, C. M. (2002). The great leap: Driving innovation from the base of the pyramid. MIT Sloan management review, 44(1), 51.
Hitt, M. A., Ahlstrom, D., Dacin, M. T., Levitas, E. and Svobodina, L. (2004). ‘The institutional effects on strategic alliance partner selection in transition economies: China vs. Russia’. Organization
Science, 15, 173–85.
Hoq, M., & Norbani, H. (2009). Innovativeness: Its antecedents and impact on SME business performance. International Journal of Business and Management, 4(11), 50-53.
Hoskisson, R., Eden, L., Lau, C. and Wright, M. (2000). Strategy in Emerging Economies. Academy of
Management Journal, 43(3), pp.249-267.
Hu, A. G. (2003). R&D organization, monitoring intensity, and innovation performance in Chinese industry. Economics of Innovation and New Technology, 12(2), 117-144.
Isobe, T., Makino, S. and Montgomery, D. (2000). Resource Commitment, Entry Timing, and Market Performance of Foreign Direct Investments in Emerging Economies: The Case of Japanese
International Joint Ventures in China. Academy of Management Journal, 43(3), pp.468-484. Kale, P., Dyer, J. H., & Singh, H. (2002). Alliance capability, stock market response, and long‐term alliance success: the role of the alliance function. Strategic management journal, 23(8), 747-767. Karnani, A. (2007). The mirage of marketing to the bottom of the pyramid: How the private sector can help alleviate poverty. California management review, 49(4), 90-111.
Kennedy, P. E. (1998). Teaching undergraduate econometrics: A suggestion for fundamental change. The American Economic Review, 88(2), 487-492.
33
Khanna, T., & Palepu, K. G. (2005). Spotting institutional voids in emerging markets.
Khanna, T., Palepu, K. G., & Sinha, J. (2005). Strategies that fit emerging markets. Harvard business
review, 83(6), 4-19.
Krugman, P. (1979). A Model of Innovation, Technology Transfer, and the World Distribution of Income. Journal of Political Economy, 87 (2): 253-266.
Lavie, D. (2006). The competitive advantage of interconnected firms: An extension of the resource-based view. Academy of management review, 31(3), 638-658.
Lavie, D. (2007). Alliance portfolios and firm performance: A study of value creation and appropriation in the US software industry. Strategic management journal, 28(12), 1187-1212. Lavie, D., & Miller, S. R. (2008). Alliance portfolio internationalization and firm
performance. Organization science, 19(4), 623-646.
London, T. and Hart, S. (2004). Reinventing strategies for emerging markets: beyond the transnational model. Journal of International Business Studies, 35(5), pp.350-370.
London, T., & Rondinelli, D. (2003). Partnerships for learning: Managing tensions in nonprofit organizations’ alliances with corporations. Stanford Social Innovation Review, 1(3), 28-35. Meyer, K. E., Estrin, S., Bhaumik, S. K., & Peng, M. W. (2009). Institutions, resources, and entry strategies in emerging economies. Strategic management journal, 30(1), 61-80.
Miller, D., & Friesen, P. H. (1982). Innovation in conservative and entrepreneurial firms: Two models of strategic momentum. Strategic management journal, 3(1), 1-25.
Morris, R. (1994). Computerized content analysis in management research: A demonstration of advantages & limitations. Journal of management, 20(4), 903-931.
Nielsen, B. B. (2003). ‘An empirical investigation of the drivers of international strategic alliance formation’. European Management Journal, 21, 301–22.
Park, S. H., & Russo, M. V. (1996). When competition eclipses cooperation: An event history analysis of joint venture failure. Management science, 42(6), 875-890.
34
Peng, M. W., & Pleggenkuhle‐Miles, E. G. (2009). Current debates in global strategy. International
journal of management reviews, 11(1), 51-68.
Pitta, D., Guesalaga, R. and Marshall, P. (2008). The quest for the fortune at the bottom of the pyramid: potential and challenges. Journal of Consumer Marketing, 25(7), pp.393-401.
Prahalad, C. (2002). Strategies for the Bottom of the Economic Pyramid: India as a Source of Innovation. Reflections: The SoL Journal, 3(4), pp.6-17.
Prahalad, C. and Hammond, A. (2002). Serving the world's poor, profitably. Harvad Business Review, pp.1-11.
Prahalad, C. and Hart, S. (2001). The Fortune at the Bottom of the Pyramid. Strategy & Business, (26), pp.1-14
Prahalad, C. and Hart, S. (1999). Strategies for the Bottom of the Pyramid: Creating Sustainable Development. pp.1-26.
Prahalad, C. (2011). Bottom of the Pyramid as a Source of Breakthrough Innovations. Journal of
Product Innovation Management, 29(1), pp.6-12.
Richardson, D., Ramirez, R., & Haq, M. (2000). Grameen Telecom’s village phone programme in rural Bangladesh: a multi-media case study final report. CIDA, 1-104.
Rosenberg, S. D., Schnurr, P. P., & Oxman, T. E. (1990). Content analysis: A comparison of manual and computerized systems. Journal of personality assessment, 54(1-2), 298-310.
Roy, J. P., & Oliver, C. (2009). International joint venture partner selection: The role of the host-country legal environment. Journal of International Business Studies, 40(5), 779-801.
Sharma, S., Vredenburg, H., & Westley, F. (1994). Strategic bridging: A role for the multinational corporation in Third World development. The Journal of Applied Behavioral Science, 30(4), 458-476. Shi, W., Sun, S. L., & Peng, M. W. (2012). Sub-National Institutional Contingencies, Network Positions, and IJV Partner Selectionjoms_1058.
Short, J. C., & Palmer, T. B. (2008). The application of DICTION to content analysis research in strategic management. Organizational Research Methods, 11(4), 727-752.
35
Simanis, E., Hart, S. and Duke, D. (2008). The Base of the Pyramid Protocol: Beyond “Basic Needs” Business Strategies. Innovations: Technology, Governance, Globalization, 3(1), pp.57-84.
Schuster, T. and Holtbrügge, D. (2012). Market entry of multinational companies in markets at the bottom of the pyramid: A learning perspective. International Business Review, 21(5), pp.817-830. Spence, M. (1977). Consumer misperceptions, product failure and producer liability. The Review of
Economic Studies, 44(3), 561-572.
Un, C. A., Cuervo‐Cazurra, A., & Asakawa, K. (2010). R&D collaborations and product innovation. Journal of Product Innovation Management, 27(5), 673-689.
Van Beers, C., & Zand, F. (2014). R&D cooperation, partner diversity, and innovation performance: an empirical analysis. Journal of Product Innovation Management, 31(2), 292-312.
Varadarajan, R. (2009). Fortune at the bottom of the innovation pyramid: The strategic logic of incremental innovations. Business Horizons, 52(1), pp.21-29.
Viscusi, W. K., & Moore, M. J. (1993). Product liability, research and development, and innovation. Journal of Political Economy, 101(1), 161-184.
Webb, J., Kistruck, G., Ireland, R. and Ketchen, Jr., D. (2010). The Entrepreneurship Process in Base of the Pyramid Markets: The Case of Multinational Enterprise/Nongovernment Organization
Alliances. Entrepreneurship Theory and Practice, 34(3), pp.555-581.
Wilhelm, M. and Sydow, J. (2018). Managing Coopetition in Supplier Networks - A Paradox Perspective. Journal of Supply Chain Management, 54(3), pp.22-41.
Zhou, W. (2012). Determinants and effects of research partnerships in China's emerging market. Contemporary Economic Policy, 30(1), 129-147.
Online Sources:
36
Annexes
Annex 1
Table 1: Sample Characteristics Country of Origin
Country Company Switzerland 15 UK 15 USA 14 Sweden 13 Denmark 11 Belgium 9 Canada 8 Norway 7 Netherlands 5 Finland 5 Ireland 4 Spain 3 Italy 3 Australia 3 Luxembourg 3 France 2 Total 120
Table 2: Sample Characteristics Industry of Origin
Industry Company
Tobacco 6
Semiconductors 8
Pharmaceuticals 6
Household/Personal Care/ Appliances 10
Specialized/Diversified Chemicals 7
Diversified Metals & Mining 5
Business & Personal Services 5
Paper & Paper Products 4
Communications Equipment 5
Biotechs 10
Auto & Truck Manufacturers 4
Construction Materials 8
Food Processing 4
Medical Equipment & Supplies 2
Other Industrial Equipment 6
Heavy Equipment 5
Electrical Equipment 2
Specialty Stores 3
Oil & Gas Operations 4
Containers & Packaging 2
Apparel/Footwear 8
Investment Services 2
Banks 4
37
Annex 2
To identify whether a partnership is forged with a traditional or nontraditional partner the partner itself needs to be researched.
For example in the annual report of Danone for the fiscal year 2017 on page 8 out of 74 the following phrase is located:
‘LAUNCH OF FANMAXX
Fan Milk launches FanMaxx in Ghana, an innovative long-shelf-life, creamy drinkable yogurt and the first of its kind in West Africa. Danone and its partner Abraaj invest $25 million USD, adding three new production lines to its factory in Accra to sustain the growing demand for its products on the Ghanaian market.’
From the above extract it is confirmed that Danone launched FANMAXX and partnered with the company Abraaj.
38
Annex 3
Table 3: General Information of the Variables
Type Variable Obtained
from Type of Data Description Dependent Innovative performance Content Analysis of 2017 annual reports
Interval Discrete continuous variable that measures the rate of new product developments or upgrade or products.
Independent Collaboration Content Analysis of 2017 annual reports
Ordinal Dummy variable that takes the value of 1 if the company has partnered with a non-traditional partner and the value of 0 if the company doesn’t have such a partner. Independent R&D
Intensity
Orbis Database
Interval R&D expenses over Operating revenue given as a percentage.
Control Size Proxy Orbis Database
Interval Logarithm of number of full-time employees for the year 2017.
Control Industry Type Orbis Database
Ordinal Dummy variable that takes the value of 1 if the industry the company belongs to is technology intensive and the value of 0 if the industry is not technology intensive.
Control Age Orbis
Database
39
Annex 4
Table 6: Goodness of Fit of Negative Binomial Modela
Value df Value/df
Deviance 118,989 94 1,266
Scaled Deviance 118,989 94
Pearson Chi-Square 95,156 94 1,012
Scaled Pearson Chi-Square 95,156 94
Log Likelihoodb -317,608
Akaike's Information Criterion (AIC) 651,217
Finite Sample Corrected AIC (AICC) 652,765
Bayesian Information Criterion (BIC) 672,217
Consistent AIC (CAIC) 680,217
Dependent Variable: Innovative Perf.
Model: (Intercept), Collaboration (Dummy), RnD Inten. (exp/or) %, Size Proxy, Industry Type, Age Experience a. Information criteria are in smaller-is-better form.
b. The full log likelihood function is displayed and used in computing information criteria.
Table 7: Omnibus Test of Negative Binomial Modela
Likelihood Ratio Chi-Square df Sig.
32,131 6 ,000
Dependent Variable: Innovative Perf.
Model: (Intercept), Collaboration (Dummy), RnD Inten. (exp/or) %, Size Proxy, Industry Type, Age Experience
40
Annex 5
Table 8: Tests of Model Effects
Tests of Model Effects
Source Type III Wald Chi-Square df Sig. (Intercept) 6,752 1 ,009 Collaboration (Dummy) 34,168 1 ,000 RnD Inten. (exp/or) % 1,296 1 ,255 Size Proxy ,316 1 ,574 Industry Type ,279 1 ,597 Age Experience ,052 1 ,819
Dependent Variable: Innovative Perf.
Model: (Intercept), Collaboration (Dummy), RnD Inten. (exp/or) %, Size Proxy, Industry Type, Age Experience
Table 10: Dependent Variable Test for Equidispersion One-Sample Kolmogorov-Smirnov Test
Innovative Perf.
N 120
Poisson Parametera,b Mean 9,39
Most Extreme Differences Absolute ,285
Positive ,285
Negative -,211
Kolmogorov-Smirnov Z 3,121
Asymp. Sig. (2-tailed) ,000
