1
MASTER THESIS BUSINESS ADMINISTRATION
INTERNATIONAL MANAGEMENT
Multinational enterprise home country development level
and human rights abuses: industrialized- versus emerging
market responses to corporate abuse allegations
Supervisor: Michelle Westermann-Behaylo Second reader: Carsten Gelhard
Amsterdam Business School
Anouk Pruijn 11111089 29-07-2016
2 Statement of Originality
This document is written by Student Anouk Pruijn who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Abstract
This research aims to elaborate on variance in responding to corporate abuse allegations by multinational enterprises coming from either industrialized- or emerging market home countries. Various arguments with respect to the likeliness of incurring allegations for either type of corporation can be found in the existing scientific literature; however, past studies tend to focus mainly on corporate commitment statements regarding abuses rather than actual handling of abuse allegations. This research attempts to look into the addressing of corporate abuse allegations by industrialized multinationals and emerging multinationals respectively by analyzing coded narratives based upon real life allegations taken from the Corporate Human Rights Database. Analysis of 516 allegations results in no major significant findings for a relationship between a firm’s home country’s level of development and responding to allegations, suggesting that other contextual factors might account for variance.
3 Index
Introduction ... 4
Literature review - International business and human rights ... 7
Theoretical framework – MNE home country development level and CAAs responses ... 18
CAA responses and remedies ... 19
MNE home country development level and CAAs ... 21
MNEs’ ownership over foreign subsidiaries ... 23
Explaining variance in CAAs responses by IM MNEs versus EM MNEs ... 25
Method ... 29
Independent variable: MNE home country development level ... 30
Dependent variable: CAA response ... 32
Moderating variable: Ownership ... 34
Control variables ... 34
Sample selection ... 35
Results ... 37
Discussion and limitations ... 43
References ... 48
4 Introduction
As stated by Ruggie (2013), identifying a common ground between international business and human rights has been problematic within the scientific realm for years now. His efforts have been grounded in the 2011 United Nations Guiding Principles on Business and Human Rights, which distinguish between the state’s duty to protect human rights, the corporate responsibility to respect human rights and access to remedy. The general principles state that ‘These Guiding Principles apply to all States and to all business enterprises, both transnational and others, regardless of their size, sector, location, ownership and structure.’ Ever since Ruggie proposed the United Nations Guiding Principles on Business and Human Rights (2011), various scholars have addressed the role of corporations in human rights allegations. Up to this point in time, there is still much to learn regarding wider patterns behind corporate human rights abuses and moderating or contributing factors. Examples include risk factors that underlie actual or potential human rights violation which need to be assessed further and the role of due diligence and grievance mechanisms. Next to these issues, there are questions that relate to the perpetration of violations, with respect to industry sectors, country-based and regional differences. Moreover, corporate-, state- or global policies can be distinguished and how they respectively incite, prevent or reduce human rights allegations. Human rights are defined here by following Gewirth (1982), defining them as ‘rights of every human being to the necessary condition of human action, i.e. those conditions that must be fulfilled if human action is to be possible either at all or with general chances of success in achieving the purposes for which humans act’ (p.3).
With respect to the wider patterns behind corporate human rights abuses that are to be investigated as mentioned above, this thesis focus specifically on the multinational aspect of business. Various scholars discuss responsibilities and effects of multinational enterprises
5 (MNEs) with respect to international human rights (Muchlinski, 2001; Meyer, 2006), with Weissbrodt and Kruger (2003) touching upon the notion that MNEs are of specific relevance with regard to global trends in international human rights since ‘… they are active in some of the most dynamic sectors of national economies… They bring new jobs, capital and
technology.’ (p. 901). The authors highlight that whereas some of these MNEs truly try to adhere to local development by meeting international standards, others do not respect these minimum international human rights standards. In the section below, several factors are discussed that have been put forward in literature when it comes to corporate respect for international human rights practices. From that evaluation, it follows that whereas some research has been done regarding the MNE’s country of origin, labeled as MNE home country development level here, no universal outcomes exist with respect to MNE home countries and the link with human rights abuses. This thesis aims to contribute to the question why variance exist in MNE respect for and promotion of international human rights by attempting to fill the scientific gap when it comes to the role of MNE home country development level as a wider pattern, based upon coded narratives from real-life corporate abuse allegations.
This study will evaluate the role of MNE nationality in relation to corporate human rights abuses by ranking this nationality based on a MNE’s home country’s level of
development. Two broad categories of MNEs in the existing literature can be found here, elaborating on MNEs originating from industrialized markets or MNEs originating from emerging markets and their relation to respect for and promotion of international human rights. From previous works, arguments that emphasize that either type of MNE is likely to incur so called corporate abuse allegations (CAAs) are presented and discussed; these theories are further investigated by testing how MNE home country development level is linked with corporate responses towards CAAs. Testing is done in a quantitative manner, using data that
6 has been accumulated by collecting coded narratives of reports on CAAs worldwide.
Through this analysis, the proposed study aims to contribute to the existing literature on wider patterns behind global corporate human rights abuses, presenting MNE home country development level as a factor. As for practical implications, such insights can
contribute to both global and more local policy design, taking into account that variance might exist when it comes to either industrialized MNEs or emerging MNEs dealing with CAAs.
This thesis is structured in the following way. In the Literature Review, findings and theories from the existing literature are presented which discuss patterns and factors related to (international) business and human rights violations. Implications regarding these patterns and factors with respect to corporate responses to CAAs are further elaborated upon in the
Theoretical Framework section. The Method section explains the design of the study, its sample and measures; next, the quantitative analysis and its following results are given and interpreted. The final sections of this work discusses the findings and their theoretical and practical implications; moreover, limitations that were found in this work are addressed.
7 Literature review - International business and human rights
Several academic works investigate firm-level concepts and their relationship with human rights. Cuesta et al. (2012) look into the observation of human rights by corporations and to what extent they are being violated by business. By evaluating company commitment statements and policies and procedures that are in place in order to ensure compliance, they find that companies have a formal commitment to specific human rights at most, those being labor rights and consumer- and environment protection. When it comes to differences in commitment and compliance with human rights between companies, financial markets and reputational implications are addressed as key external drivers (Cuesta et al., 2012).
Regarding the conceptualization of human rights at the company level, Preuss and Brown (2011) evaluate large corporations and their adoption of human rights policies, the policy content and possible sectoral activity differences regarding the construction of such policies. Similarly to Cuesta et al. (2012), the authors find that corporate commitments to human rights are mainly superficial and that a distinction between positive rights and negative rights is being made, with companies focusing in a narrow manner on the latter. Whereas positive rights involve initiatives to protect or fulfill human rights, negative ones merely implicate corporate respecting of human rights. Again, experiencing exposure to public pressure rather than the gravity of the human rights allegation for the corporation or industry is put forward as an external driver. With respect to observed human rights due diligence, Hamann et al. (2009) point out that firm size and sector appear to be of little influence. Interestingly, these authors also find that corporate participation in the UN Global Compact and JSE Socially Responsible Investment Index is statistically insignificant when it comes to human rights due diligence for the top 100 South-African firms that are listed on the Johannesburg Stock Exchange. From their results, explicit leadership commitment, government regulations and stock exchange listing rules come forward as the most important mechanisms for human
8 rights due diligence by corporations. These authors explain due diligence as considerations taken by corporations to verify whether human rights challenges exist and whether their operations would (in)directly impact human rights (Hamann et al., 2009).
Whereas the findings that are briefly presented above are interesting, it can be concluded that many more insights are to be gained within the context of underlying relationships regarding the firm-level and corporate human rights allegations. The role of companies with respect to human rights allegations appears to be of particular importance in the international business context, which specific attention is given to in the following paragraphs.
Buckley and Casson (2009) define a MNE as ‘a firm that owns and controls activities in two or more different countries’ (p. 1564), engaging in foreign direct investment (FDI). Wettstein (2010) emphasizes that corporations increasingly gain political influence, which is an alarming development when related to the ‘structural interconnectedness of the global economy’ (p.35): MNEs are argued to have a hard time in critically evaluating whether their core business is indirectly related to human rights violations. In such indirect involvement, the corporation itself is not accused of violating human rights but being complicit in those
violations by other companies or entities. Such complicity is being argued to be difficult to point out according to Wettstein (2010): ‘It takes quite a bit of critical capacity for a company to see and acknowledge that despite not doing anything wrong directly, its very core business processes might be connected to the violation of people’s most basic rights.’ (p. 35)
Complicity in human rights violations is said to occur when businesses fail to reasonably assess that their activities are not linked to human rights violations. Whether or not there was a malicious intent to do any harm is not relevant in complicity matters; intent to participate, for example through regular business conduct, is what counts here (see Wettstein, 2010).
9 Blanton and Blanton (2006; 2007a; 2007b and 2009) investigate links between
international business and human rights in particular, stating that human rights conditions have a significant influence on dyadic trade flows (Blanton & Blanton, 2007a). More specifically, in their 2007b study focusing on non-OECD countries, they find that corporate respect for human rights has a positive impact, thus encouraging FDI. These findings build upon Blanton and Blanton’s research from 2006 which states that ‘The protection of human rights reduces risk and contributes toward economic efficiency and effectiveness’ (p. 464), indicating that ‘Human rights are a significant determinant of the amount of FDI inflows’ (p. 464). With regard to sectoral differences for the relationship between human rights and FDI, the authors put forward the significance that is given to the skill level of the workforce and corporate integration with the host society (Blanton & Blanton, 2009): sectors that value higher skills and integration are significantly linked with human rights with respect to FDI determinants.
Findings from the field of business and human rights are specifically interesting when considering the context of MNE operations in host markets that can be regarded as developing countries. Giuliani and Macchi (2013) explain how governments compete to attract FDI from MNEs due to the belief that those investments will bring economic advantages in the form of employment and technological transfers. The authors highlight that for some developing countries, the government disregards ‘human rights abuses perpetrated by foreign investors, provided the latter are satisfying certain economic targets’ (p. 480), thus prioritizing economic development over human rights. The authors evaluate mediating factors of MNE impact in developing countries, being both external and internal. External factors that are discussed are host country characteristics and industry-level characteristics. Host country characteristics are social capability, state capacity and the MNE’s engagement with civil society, which all positively impact human rights. Social capability refers to general education- and technical
10 competence levels; state capacity is defined as the state’s ability to ‘enforce contracts and regulate markets and to guarantee a strong and impartial legal system’ (Giuliani & Macchi, 2013). Industry-level characteristics on the other hand have a negative impact, consisting of high competition and low technology. MNE nationality, strategic investment motivations, subsidiary innovation and entrepreneurship and subsidiary autonomy are presented as internal factors. Of these factors, only subsidiary innovation and MNE nationality do not impact human rights in a negative way. MNE nationality appears to have a mixed impact: interestingly, Giuliani and Macchi (2013) touch upon the idea that MNEs holding an
industrialized country nationality ‘… will be more likely to generate spillovers and respect the host country’s human rights.’ (p. 495), which is explained by industrialized country MNEs being more able ‘… to transfer higher-quality knowledge and generate productivity and export spillovers’ (p. 495), and those MNEs being aligned with their advanced legal systems pushing for transparent management practices and respect for the law. Contrasting to this idea, the authors point out that several studies have produced differing results when it comes to positive spillover effects from industrialized market MNEs’ (IM MNEs) activities in developing countries, which might be accounted for by home and host country differences with respect to the cultural, social and legal context.
With the premises described above in mind, one might wonder about the variance with respect to the impact of human rights abuse accusations directed to MNEs holding different nationalities. More specifically, this thesis aims to classify variance with respect to MNEs coming from either markets that can be regarded as industrialized or emerging.
Luo and Tung (2007) define emerging market MNEs (EM MNEs) as ‘international companies that originated from emerging markets and are engaged in outward FDI, where they exercise effective control and undertake value-adding activities in one or more foreign
11 countries’ (p. 482), with emerging markets being nations with frail, inconsistent legal systems that demonstrate fast economic and industrial development. When comparing country-specific advantages and firm-specific advantages held by EM MNEs and IM MNEs, Amighini et al. (2015) argue that EM MNEs often benefit from production capabilities (such as low cost production factors), home country social networks and availability of capital. Advantages held by IM MNEs on the other hand include ownership of technologies, brand names and
intellectual property (see Amighini et al., 2015). Luo and Tung (2007) state that cooperation with global players and transfers of technological and organizational skills are advantageous for EM MNEs, which are argued to use international expansion as a springboard. Amighini et al. (2015) confirm this notion that EM MNEs are likely to use global linkages in order to leverage existing resources and lower costs, whilst simultaneously learning about ‘… new sources of competitive advantages and how to operate internationally’ (p. 349). Similar motives for international expansion by EM MNEs presented by Fey et al. (2016) include desires to acquire technology, brand names and raw materials, next to growing in size and to diversify risk. Luo and Tung’s (2007) international expansion as a springboard perspective posits that in order to overcome their latecomer disadvantage, EM MNEs take aggressive and risk-taking approaches in order to gain strategic resources and to decrease limitations related to their native institutional frameworks and market structures. Such a springboard approach to compensate for competitive weakness is moderated by national government, the extent to which industrialized key players offer strategic resources and global economic- and production integration (Luo and Tung, 2007). Motivations for EM MNEs behind their springboarding behavior are described as either asset- or opportunity seeking. Madhok and Keyhani (2012) elaborate on this springboard perspective by posing that EM MNEs acquire IM MNEs for their competitive catching-up, upgrading resources and capabilities and learning through such moves.
12 How would such a springboard approach relate to human rights allegations? One could argue that aggressive and risk-taking measures could (in)directly have implications for human rights violations. Luo and Tung (2007) describe several issues and challenges that come with international springboarding by EM MNEs, some of those being related to a ‘lack of global experience, managerial competence and professional expertise’ (p. 482) and ‘poor corporate governance’ (p. 482): such challenges might link to human rights allegations, or, as before mentioned by Wettstein (2010), issues related to judging indirect corporate involvement in human rights violations.
On the other hand, research indicates that legitimacy and reputation are key in operating in international markets, such as Fiaschi et al. (2012) who find that EM MNEs, compared to MNEs from developed countries, can be regarded as more eager to manage human rights and commit to corporate social responsibility (CSR). Whereas aggressive and risk-taking measures taken in a springboard approach do not appear to align with human rights promotion at first glance, EM MNEs might need to use CSR policies including the promotion of human rights in order to strengthen their competitive position in international business, following Gugler and Shi (2009). Marano et al. (2016) confirm the use of CSR reporting strategies by EM MNEs to overcome legitimacy challenges. CSR is defined here following McWilliams et al. (2006), as ‘situations where the firm goes beyond compliance and engages in actions that appear to further some social good, beyond the interests of the firm and that which is required by law.’
Ardichvili et al. (2012) however indicate that significant differences exist among the four largest emerging economies (Brazil, Russia, India and China or BRIC countries) with respect to perceptions of ethical business cultures in large business organizations, which raises expectations regarding human rights allegations for these countries. To illustrate, Brazilian and Indian firms were found to be more concerned about ethical cultures than Chinese and
13 Russian respondents. Christie et al. (2003) point out that such ethical attitudes held by
business managers are strongly influenced by national culture, employing Hofstede’s cultural dimensions in India, Korea and the United States.
Evaluating the ideas put forward by Giuliani and Macchi (2013) in an earlier section of this work, IM MNEs are generally thought of as being more respectful towards human rights than for example EM MNEs or MNEs originating from developing markets (DM MNEs). These authors claim that IM MNEs have respectable legal systems and more
transparent management practices, thus relying on an explanation that builds upon differences in institutional backgrounds when it comes to corporate human rights violations by MNEs coming from various countries of origin. However, IM MNEs which are active in developing countries with significantly lower legal standards might have a difficult time avoiding
(in)direct involvement in corporate human rights violations due to exactly those large differences in institutional frameworks and possible presence of corruption. Fasterling and Demuijnck (2013) explain that MNE activity in non-democratic countries to that extent can even be justified by posing that there will always be competing MNEs that are willing to take a corporation’s place if it decides to leave, due to for example human rights violations. MNE activities that (in)directly support an oppressive regime thus consequently do ‘not increase the number of victims there would be anyway’ (p. 805). Moreover, another perspective that can be applied here holds that IM MNEs are facing public legitimacy and cost efficiency issues in their home market, leading them to transfer misconduct and ill-practices in order to gain cost advantages to specific foreign subsidiaries with lower stakeholder expectations (Surroca et al., 2013). Thus, following this it can be argued that IM MNEs are likely to incur human rights allegations in host countries through this transferring behavior. Adding this idea to
Wettstein’s (2010) notion that IM MNEs might have a difficult time avoiding (in)direct involvement in human rights allegations due to large differences in institutional frameworks
14 and corruption, a strong claim can be made that IM MNEs are not always dealing with human rights that well.
Even though some prior research on the relationship between MNE nationality and human rights violations exists, the issue appears to be that there are no universal outcomes as elaborated upon in above paragraphs. MNE nationality is classified based on a country’s level of development in this work, being either MNEs coming from industrialized-, emerging- or developing markets. Summarizing, on one side there are Giuliani and Macchi (2013)
highlighting an argument that considers IM MNEs in general being good at avoiding human rights abuses. This contrasts with IM MNEs’ transferring behavior as described by Surroca et al. (2013) and the involvement issue raised by Wettstein (2010) and Fasterling and Demuijnck (2013). On the EM MNE side, Fiaschi et al. (2012) stress the need to manage human rights; however, similar to the IM MNE case, EM MNEs’ springboarding behavior as elaborated upon by Luo and Tung (2007) generates contradicting expectations when it comes to dealing with human rights here.
Cumulating all above arguments and findings, the proposed research aims to clarify which arguments and theories could be valid when it comes to MNE home country
development level and corporate human rights allegations. In order to differentiate the study from the findings by Giuliani and Macchi (2013) and Fiaschi et al. (2012), actual post-human rights allegations behavior by MNEs from both industrialized- and emerging markets would be considered in this case. Giuliani and Macchi (2013) do not actually test their proposed arguments, but come to their conclusions through a literature review; moreover, MNE nationality is only a fraction of the overall focus in their study, next to other factors that are being considered. Fiaschi et al. (2012) on the other hand do focus on various countries
15 evaluating the presented scientific contributions and theorized characteristics on IM MNE and EM MNE behavior respectively, it is apparent that a research gap exists with respect to MNE home country level of development and corporate human rights allegations specifically, which thus leads to the following research question:
How does MNE home country development level relate to corporate human rights
allegations?
Previous research by Salcito et al. (2015) partially looks into this research gap, through an analysis on human rights policies by the biggest firms in the finance-, mining-, oil and gas-, food and beverage-, apparel- and agribusiness sectors. Rather than evaluating links between these corporations and actual dealing with human rights allegations, the authors score the extent to which the firms state their commitment regarding their responsibility to respect human rights, to conduct human rights due diligence, and to give access to remedies for human rights allegations related to their business (Salcito et al., 2015). Firms from Asia, the Middle East and North Africa score the lowest when analyzing their commitment to the responsibilities as described above; companies coming from industrialized areas such as Europe, Canada and Australia ‘demonstrate the highest adoption rate of human rights duties’ (p. 673). Interestingly though, companies that originate from the United States score rather poor on human rights duties adoption in Salcito et al.’s (2015) study, falling in a middle group between high- and low performers.
In another related study, Kamminga (2016) evaluates actual responses by corporations after they have been alleged of a human rights violation. Besides calculating the response rate to allegations, the study looks into differences in industrial sectors and commitment to
international instruments when it comes to providing a response or not. A response rate of 70% was found, indicating that ‘… 30 per cent of companies essentially failed to respond to meet perhaps the most basic part of their due diligence responsibility…’ (Kamminga, 2015 p.
16 99). Most interestingly for the purpose of the proposed study are the calculated response rates per corporate home country: whereas some home countries scored above the average 70% response rate, others remained far below. To illustrate, the Netherlands scored 90%; Italy 62%; Russia 39% and South Africa 93%. Furthermore, quite some home countries score close to the 70% average, overall indicating variance with respect to corporate home country and responding to alleged human rights violations.
Whereas both Salcito et al. (2015) and Kamminga (2016) look into corporate home country and association with human rights abuses, they do not classify the home countries based on level of development; however, if this would be taken into account, neither study is able to highlight a pattern for either IM MNEs or EM MNEs. To illustrate, Kamminga (2016) mentions that ‘There is no similarity in response rate between [Brazil, Russa, India, China and South Africa]’ (p. 103), which could all be regarded as emerging markets. Contrastingly, building upon arguments and behavioral explanations presented in previous sections, this thesis aims to further clarify whether links between MNE home country development level and human rights violations specifically exist in order to evaluate whether human rights allegations might be explained or related to either springboarding behavior through aggressive and risk-taking measures by EM MNEs, or transferring behavior and differences in
institutional backgrounds experienced by IM MNEs. Rather than merely evaluating corporate statements regarding commitment to respect and promote human rights as performed by Salcito et al. (2015), the suggested research will be based upon analysis of actual reports on human rights allegations worldwide in order to draw conclusions based upon real life events in a similar fashion to Kamminga’s (2016) research. Through investigation of actual corporate responses after the allegation has taken place, the aim is to make a valuable contribution to the existing works presented in previous paragraphs of this section, one which reflects actual corporate statements post-allegation rather than corporate stated commitment pre-allegation.
17 Next to evaluating response rates in general though, more attention will be given to the
content of the responses themselves, thus going beyond Kamminga’s (2016) efforts.
Moreover, in contrast to the studies by both Salcito et al. (2015) and Kamminga (2016), rather than presenting findings per country, this study aims to highlight a wider global pattern with respect to the addressing of human rights abuses based upon a MNE’s home country’s level of development. Another difference is that the proposed research focuses on multinational
corporations exclusively, not taking into account behavior or violations that took place within the same host- and home country.
The incorporation of a springboard perspective versus a transferring perspective in this matter is a unique one and could explain for possible variance in human rights allegations by EM MNEs and IM MNEs. A focus on EM MNEs specifically would contribute to scientific works on human rights, since emerging economies have experienced ‘rapid growth and remarkable transformation’ (Luo & Tung, p.481), and are accounting for a significant rise in FDI outflows. If the wider patterns behind corporate human rights abuses and its moderating or contributing factors and the role that companies have in this is to be further investigated, the impact of EM MNEs compared to IM MNEs is not to be left out.
18 Theoretical framework – MNE home country development level and CAAs responses
This research aims to clarify the research question How does MNE home country development level relate to corporate human rights allegations? by elaborating on MNE
responses to corporate abuse allegations (CAAs) specifically, investigating whether significant differences exist for MNEs coming from either industrialized- or emerging markets. The level of ownership over foreign subsidiaries is taken as a moderating factor in establishing whether variety is present. The main argument for investigating this matter is based upon the literature review above: both for IM MNEs and EM MNEs, studies have put forward theories and findings for a higher or lower likelihood of incurring human rights violations. This notion then is taken further here by implying differences in approaching CAAs by either IM MNEs or EM MNEs. Based upon cases reported from real life, it is evaluated whether significant variance exists in providing a response in the first place and what type of response to CAAs is given based upon MNE home country development level.
The relation that will be analyzed is thus between MNE home country development level, CAAs responses by the afflicted MNE and the level of ownership the MNE has over the foreign subsidiary. The independent variable is MNE home country development level, with CAAs response being the dependent variable and the relation argued to be moderated by ownership levels. MNE home country development level is categorized as either MNE’s coming from industrialized markets, emerging markets or developing markets.
In the paragraphs below, elaboration will follow on CAAs responses as a dependent variable and arguments will be provided to make predictions for both EM MNEs and IM MNEs likeliness to provide responses to CAAs. The role of ownership as a moderating variable is also being explained in further detail. The section will be concluded with a conceptual overview and the propositions that are formulated for further analysis.
19 CAA responses and remedies
Ruggie’s Protect, Respect and Remedy framework (2008) poses that whilst it is the state’s duty to protect human rights, it is the responsibility of corporations to respect those rights and to provide a remedy when human rights are being violated. This is to be done through a continued due diligence approach where companies assess their impact on human rights (Ruggie, 2008).
When it comes to corporate responses to CAAs, involved companies are argued to provide a communicative response in order to secure their corporate image according to Bradford and Garrett (1995). These authors discuss five potential responses, being no response, denial, excuse, justification and concession. As for businesses not providing any response after incurring a CAA, it is posited that their perceived image held by third parties will be directly and adversely impacted. Moreover, Kamminga (2016) links Ruggie’s (2008) proposed due diligence approach for corporations with responding to CAAs: it is argued that failure to provide merely a response to human rights allegations does not align with due diligence. Whereas responses to CAAs do not necessarily have to be beneficial for human rights promotion, not providing a response at all ‘... indicates unwillingness to engage with civil society on important issues relating to the company’s conduct and implies a breach of the due diligence requirement contained in a company’s responsibility to respect human rights.’ (Kamminga, 2016, p.99)
Bradford and Garrett (1995) argue that perceptions of the image of the accused company will be more positive when the response communication matches the situation than when no response is given. Denials should be communicated when the corporation can prove that it did not commit the allegation, whereas it is advised to communicate an excuse when proof exists that there was no control over the incident or its impacts. Justifications are fit when the corporation has proof that the CAA was evaluated through inappropriate standards,
20 and the authors recommend concession when the raised CAA is valid (see Bradford &
Garrett, 1995).
This idea of corporate responses to CAAs and corporate images overlaps with a key external driver for commitment and compliance to human rights by businesses as discussed in the earlier section of this thesis, with Cuesta et al. (2012) pointing out reputational
implications and Preuss and Brown (2012) highlighting exposure to public pressure when it comes to businesses being involved with human rights abuses. It is then argued here that the likeliness of MNEs providing a response to a CAA increases as they attach more value to their corporate image. For this research then, an analysis pointing out significant differences in the extent to which IM MNEs and EM MNEs provide responses to CAAs could hint to the notion that IM MNEs and EM MNEs value their corporate images differently.
Whereas responding to a CAA can thus be regarded as a good start when it comes to addressing and promoting human rights as part of corporations’ due diligence responsibility, MNE statements following CAAs do not necessarily have to align with their further follow-up behavior regarding those CAAs. The proposed study does not evaluate actual corporate
respect for human rights when it comes to remedies that are provided for abuses. The UN Guiding Principles on Business and Human Rights recognize various forms of remedy for human rights violations, ranging from apologies to financial- and non-financial compensation. Grievance is defined as ‘a perceived injustice evoking an individual’s or a group’s sense of entitlement, which may be based on law, contract, explicit or implicit promises, customary practice, or general notions of fairness of aggrieved communities’, with a grievance
mechanism indicating ‘any routinized, state-based or non-state-based, judicial or non-judicial process through which grievances concerning business-related human rights abuse can be raised and remedy can be sought.’ (UNHR, 2011, p. 27) When it comes to the role of
21 businesses in providing remedy, they can apply their own grievance mechanism. Such
mechanisms are then mainly non-judicial, contrasting to judicial remedies which are provided for by the state. Criteria that are set for effective remedy are established as ‘legitimate,
accessible, predictable, equitable, transparent, rights-compatible, a source of continuous learning, and based on engagement and dialogue’ (UNHRC, 2011, p. 33- 34). For the scope of this research, remedies provided by MNEs to CAAs are not included. MNEs’ intend to
address and promote human rights based upon their responses towards CAAs can thus not be interpreted as their actual further or future corporate behavior.
MNE home country development level and CAAs
Two perspectives can be applied when arguing that IM MNEs or EM MNEs are likely to incur CAAs. However, for both perspectives, counter-arguments can also be found which argue why either IM MNEs or EM MNEs are less likely to incur CAAs. Due to these
opposing ideas, evaluating how responses to CAAs differ between IM MNEs and EM MNEs could provide valuable insights in the relation between international business and human rights violations and the associated underlying patterns.
The first perspective holds that IM MNEs are facing public legitimacy and cost efficiency issues in their home market, leading them to transfer misconduct and ill-practices in order to gain cost advantages to specific foreign subsidiaries with lower stakeholder expectations (Surroca et al., 2013). Thus, following this it can be argued that IM MNEs are likely to incur CAAs in host countries through this transferring behavior. Moreover, it is argued that IM MNEs might have a difficult time avoiding (in)direct involvement in CAAs due to large differences in institutional frameworks and corruption (Fasterling and
Demuijnck, 2013). Following this perspective, it is put forward here that stakeholders generally focus more on wholly owned subsidiaries for which information is easily available
22 (Surroca et al., 2013). Therefore, transferring misconduct and ill-practices to foreign
subsidiaries without risking lower overall MNE legitimacy should be more likely to occur for subsidiaries with a non-majority ownership when it comes to IM MNEs. This notion is
supported by Frey (1997) stating that the standard used to be that MNEs felt a lesser degree of responsibility to act upon human rights violations when they were not closely involved with that abuse. Only when the company moves closer ‘to the point of being the actual abuser, the responsibility to intervene becomes the greatest’ (p. 154). This would align with CAAs
reports from the Apparel and Textile industry (see Emmelhainz & Adams, 1999), in which the accused MNEs are usually indirectly involved in the abuse through sourcing from local
suppliers and factories.
On the other hand, Giuliani and Macchi (2013) argue that since IM MNEs originate from advanced legal systems, it is likely that they will push for transparent management practices and respect the laws. Whereas Fasterling and Demuijnck (2013) thus argue that institutional differences will lead to a great likeliness of incurring CAAs by IM MNEs, Giuliani and Macchi (2013) stress that such differences in institutional frameworks will actually motivate IM MNEs to remain true to their own, respectable institutional framework when operating abroad. This reasoning is argued in this thesis to mainly link back to key external drivers for commitment to human rights as found by Hamann et al. (2009) that were previously described, linking an advanced institutional environment from which IM MNEs originate to a desire to explicitly commit to leadership, to government regulations such as a properly functioning law system, and stock exchange listings.
The second perspective posits that EM MNEs use international expansion as a springboard in order to overcome their latecomer disadvantages with respect to IM MNEs, taking aggressive and risk-taking approaches to gain strategic resources and to decrease
23 limitations related to their native institutional frameworks and market structure (Luo & Tung, 2007; Amighini et al., 2015). The argument here would be that EM MNEs are likely to incur CAAs in host and home country through their aggressive and risk-taking way of doing
business as a sort of negative externality. A negative externality occurs when another party or entity than the company itself has to deal with costs brought by that company’s business activities (Cohen & Winn, 2007), being human rights abuses in this case such as
environmental harm or health issues. High levels of ownership and equity-based modes of entry are regarded as more risky and aggressive compared to lower ownership levels and non-equity modes (Pan & Tse, 2000; Meyer et al., 2009). Thus, for EM MNEs, incurring CAAs should be more likely to occur in subsidiaries which are majority owned.
Opposing this idea are some findings which were put forward earlier in the literature review: EM MNEs are said to be more eager to manage human rights and commit to human rights (Fiaschi et al., 2012; Marano et al., 2016), which is necessary to do in order to
strengthen EM MNEs’ competitive position in international business (Gugler & Shi, 2009). Key external drivers here would then be those which were brought up by Preuss and Brown (2012), indicating exposure to public pressure, and Cuesta et al. (2012) stressing reputational implications.
MNEs’ ownership over foreign subsidiaries
With respect to the two perspectives of transferring behavior by IM MNEs and springboarding behavior by EM MNEs and the implications of these behaviors for human rights, MNEs’ ownership over foreign subsidiaries is taken as a moderating factor for this research. As discussed earlier, EM MNEs demonstrating springboarding behavior engage in risk-taking measures and aggressive strategies. A high level of ownership is regarded as risky and more aggressive in comparison to lower ownership levels here, with equity-based entry
24 modes requiring more resource commitment, risk, return and control than non-equity modes (Pan & Tse, 2000). Gubbi et al. (2010) highlight the link between overseas acquisitions by EM MNEs as a springboard in order to gain access to strategic assets, underlining the
possibility to overcome disadvantages and to increase competitiveness. The authors note that overseas acquisitions bring complexity and risks, but allow for a leapfrog trajectory by bringing (inorganic) rapid growth and internalization of resources that are not easily obtained through factor markets or that would otherwise require maturation to develop. Moreover, EM MNEs can use this type of entry mode to overcome liabilities of newness and foreignness. Freeman et al. (1983) describe the liability of newness as costs that new corporations experience compared to older organizations, related to dependency on cooperation of strangers, low legitimacy levels and competitive disadvantages. The liability of foreignness can be described as the costs of doing business abroad that result in a competitive
disadvantage for a MNE compared to a local business (Sethi & Judge, 2009). In addition to these two concepts, Madhok and Keyhani (2012) mention that a so called liability of
emergingness is also to be overcome by EM MNEs, being an ‘… additional disadvantage that [emerging market multinationals] tend to suffer (over other foreign [industrialized market multinationals] by virtue of being from emerging economies’ (p. 28). Summarizing, Gubbi et al. (2010) propose that ‘…cross- border or international acquisitions enable such firms to fulfill some of their imperative needs, achieve accelerated internationalization, and in the process facilitate their strategic leap to "emerging multinational" status’ (p. 400). Meyer et al. (2009) elaborate by stating that entry modes in emerging markets are determined by the strength of institutions in host markets and the need for different types of local resources. Entry modes are then seen as risky or aggressive when acquisition or greenfield is chosen when institutional development is low, and when a greenfield approach is taken rather than an acquisition or joint venture when resources are needed (see Meyer et al., 2009).
25 Thus, for EM MNEs, a high level of ownership is regarded here as more risky and more aggressive than low levels of ownership or sourcing. On the other hand, as explained in earlier sections, IM MNEs that are likely to incur CAAs through transferring behavior are argued to opt for non-majority ownership in these subsidiaries or choose sourcing. Following Pan and Tse (2000), such non-majority ownership would be regarded as less risky and less aggressive. The moderating effect in this research would then be that EM MNEs which have a high level of ownership over their foreign subsidiaries are differently involved in CAAs responses than IM MNEs which have a non-majority ownership over their foreign subsidiaries or choose to engage in sourcing.
Explaining variance in CAAs responses by IM MNEs versus EM MNEs
Summarizing the arguments that were brought forward in the paragraphs above, the following conceptual framework is presented when it comes to IM MNEs and EM MNEs and CAAs in Table 1:
Table 1: Conceptual framework background theory
Perspective: IM MNEs Perspective: EM MNEs
Situation IM MNEs are likely to incur
CAAs due to their transferring behavior
EM MNEs are likely to incur CAAs due to their springboarding behavior
Level of ownership Minority ownership Majority ownership
Situation IM MNEs are unlikely to
incur CAAs due to their institutional background pushing for transparency and respect
EM MNEs are unlikely to incur CAAs due to their need to strengthen their international competitive position
Key drivers Commitment to leadership,
government regulations, stock exchange listings
Exposure to public pressure, reputational implications
Based upon the various perspectives of MNE home country development level inflicting with CAAs, it is suggested that variance exists in IM MNEs versus EM MNEs responding to
26 CAAs. Besides evaluating whether either IM MNEs or EM MNEs are more likely to provide a response in the first place, insights are also to be gained when it comes to the type of response that is given: this could range from a denial of the accusation to an
acknowledgement of or an addressing approach towards the claim.
When explaining for variance in CAAs responses, one might consider CSR adoption by EM MNEs in comparison to adoption by IM MNEs. Maon et al. (2009) highlight examples when it comes to CSR adoption, with MNEs communicating their commitment to CSR in reports based upon actual achievements and what they explicitly hope to achieve in the future. As touched upon in an earlier section, Gugler and Shi (2009) discuss that whilst it is usually the industrialized markets that set the standard and design CSR approaches and policies, EM MNEs overall do tend to adopt CSR. However, they are argued to do so in order to be able to leverage a competitive advantage: a precondition for doing business nowadays is to comply with CSR standards as seen from a more Western perspective (Gugler & Shi, 2009). Baskin (2006) points out that reported CSR in emerging markets can be regarded as more developed than generally thought of and even exceeding industrialized market standards at times when evaluating CSR behavior in South Africa, Brazil, India and some parts of Eastern Europe. Factors that were analyzed in that study are Dow Jones Sustainability Index membership, registration with the Global Reporting Initiative, ISO 14001 certification and corporate website statements. A 2010 study by Cheung et al. underlines that CSR is gaining importance in Asian emerging markets, looking into market valuation specifically. Li et al. (2010) on the other hand claim that developing countries adopt CSR behavior to a lesser extent than
developed countries due to their overall lower economic development. Evaluating CSR behavior of MNEs from Brazil, Russia, India and China specifically, they find that government policy is a key driver in stimulating for responsible practices. Their
27 governance environments are transformed in the direction of more transparency,
accountability, and public ordering at the macro level, we can also expect their firms to improve their CSR.’ (p. 649). Such reasoning could provide reinforcement that a country’s level of development affects how MNEs respond to CAAs and that variance could exist for IM MNEs versus EM MNEs. Other factors affecting CSR communications by EM MNEs that are found by Li et al. (2010) are industry-level and firm-level explanations, although those are regarded as less significant than the before mentioned country-level factor. Moreover, the authors address the notion that IM MNEs might have difficulties to put their CSR standards into practice in host countries due to differences in the macro institutional environment, which was touched upon in an earlier section of this work by putting forward a similar argument by Fasterling and Demuijnck (2013). Whereas Baskin (2006) thus found that ‘overall, there is not a vast difference in the approach to reported corporate responsibility between leading
companies in high-income OECD countries and their emerging-market peers’ (p.30), the proposed study that is described below seeks to find out whether variance does exist when it comes to actually handling CAAs rather than reported corporate commitment pre-allegation.
Overall, the arguments that have been presented lead to the following propositions, which will be analyzed in later sections of this study.
P1: IM MNEs are likely to respond differently to CAAs than EM MNEs, with this relation
being moderated by ownership level.
Analysis will clarify whether differences exist in providing a response to a CAA or not and whether this relationship is moderated by minority ownership for IM MNEs and majority ownership for EM MNEs respectively. As mentioned before, Kamminga (2016) evaluates whether or not a response to a CAA is provided by corporations in a similar manner.
28 countries based on their development level per se. Moreover, Kamminga (2016) focuses on variance in response rates with respect to industry sectors, companies themselves and
commitment to corporate instruments next to corporate home countries; the proposed research instead includes ownership levels as a factor. Finally, another contribution would be to
address the content of responses next to providing a response or not, something which Kamminga (2016) does not address, captured in the second proposition:
P2: MNE home country development level affects the corporate CAAs response type, with this
relation being moderated by ownership level.
Besides investigation of providing a response to CAAs in the first place, analysis will look into possible variance with respect to the types of responses that are provided by IM MNEs compared to EM MNEs, which is not addressed by Kamminga (2016). This is being
addressed since a MNE can provide a response to a CAA in the first place, but providing a denial response reflects differently on that corporation and its apparent attitude regarding human rights protection than a response that offers a strategy to better the situation and / or compensate possible victims.
29 Method
The proposed research aims to classify variance with respect to MNEs coming from markets that can be regarded as either industrialized or emerging and how they respond to CAAs. The conduction of the investigation for the proposed research question is done with the help of the Corporations and Human Rights Database (CHRD), which is constructed by coding cases of reported corporate human rights allegations worldwide. This project seeks to systematically register who makes claims regarding corporate human rights abuses, the responses by the corporations and whether claims are actually addressed. Coding is partially done through the efforts of various teams worldwide; however, some industries and continents still need to be addressed. Students from the Amsterdam Business School have received training and participated in the further systematic coding of the mining industry in Africa specifically. The database will contain claims from post-2000 addressed to over 5000 corporations, and takes data from the Business and Human Rights Resource Centre (the so called Company Abuse Claims). The outcome of the CAA coding acts as the dependent variable in the research, which is complemented with other existing secondary data to quantitatively test the model. MNE home country development level is taken as the
independent variable, with CAA responses as the dependent variable and level of ownership as the moderating variable; variables that will be controlled for are firm size and sector.
Based on the outlook of the CHRD, the sample is selected. For each MNE in the sample, there must be 1) a link with CAAs, which will be covered for since MNEs are taken from the CHRD, thus having a link per definition; 2) a clear establishment on the MNE nationality and corresponding development level; and 3) a percentage of ownership over the foreign subsidiary in question. Elaboration on 2) and 3) will follow in the paragraphs below.
30 Independent variable: MNE home country development level
For each MNE in the sample linked to a human right abuse claim, nationality and level of development must be established. Hoskisson et al. (2000) define emerging markets as ‘… low-income, rapid-growth countries using economic liberalization as their primary engine of growth’, with the nation’s government stimulating economic liberalization and a free-market system. Distinguishing countries based on their level of development appears to be a rather ambiguous undertaking, with no uniform agreement on classifications worldwide. For example, the United Nations system notes that ‘There is no established convention for the designation of “developed” and “developing” countries or areas,’ indicating Japan, Canada and the United States, Australia and New Zealand, and Europe being generally considered as “developed” regions (United Nations Statistics Division, 2013). Furthermore, the United Nations has developed the Inequality-adjusted Human Development Index (IHDI), which classifies nations worldwide with regard to human development in the dimensions of health, education and standard of living (income) whilst discounting for inequality levels (Human Development Reports, 2015). The World Economic Outlook (WEO) employed by the International Monetary Fund (IMF) distinguishes between advanced economies, emerging markets and developing economies based on per capita income level, export diversification and the degree of integration into the global financial system. The WEO though notes that ‘This classification is not based on strict criteria, economic or otherwise, and it has evolved over time. The objective is to facilitate analysis by providing a reasonably meaningful method of organizing data.’ (World Economic Outlook, 2015). Ultimately, classifying MNE origin as either industrialized, emerging or developing is a rather troublesome undertaking and comes down to combining from and using overlaps in these systems, reports and indices. The
31 Table 2 – MNE nationalities and level of development
MNE home country Classification level of development Entries in sample
Argentina Emerging 10 Australia Industrialized 11 Belgium Industrialized 1 Brazil Emerging 15 Canada Industrialized 66 Chile Emerging 1 China Emerging 19 Colombia Emerging 2 Cyprus Industrialized 2 Finland Industrialized 1 France Industrialized 17 Germany Industrialized 22
Hong Kong Industrialized 2
India Emerging 3 Israel Emerging 1 Italy Industrialized 8 Japan Industrialized 6 Luxembourg Industrialized 1 Malaysia Emerging 5 Mexico Emerging 2 Netherlands Industrialized 2 Panama Emerging 3 Peru Emerging 1 Russia Emerging 2
South Africa Emerging 25
South Korea Industrialized 8
32
Switzerland Industrialized 19
Taiwan Industrialized 8
Thailand Emerging 1
United Arab Emirates Industrialized 3
United Kingdom Industrialized 48
United States Industrialized 146
Venezuela Emerging 1
Multiple nationalities Industrialized 36
TOTAL
Home countries: 34 IM: 425 EM: 91 cases 516 cases
Dependent variable: CAA response
Corporate responses to CAAs are collected by making use of the CHRD. If a response to the allegation was provided, this database lists from what division it came from (local branch, national headquarters, international headquarters (this includes the parent company), or unable to determine (UTD)). Moreover, information can be found on who provided the response (CEO/President, Board of Directors, company PR/media relations, human resources/corporate services, company legal representation or UTD/other), the date of the response and a brief description. Various types of responses are being recognized, which are categorized into four groups for the purpose of this research in Table 3:
33 Table 3: CAAs Response Types
CHRD response type Categorized as
No response 1) No response
Denial: the claim was exaggerated
Denial: the claim does not take into account all of the facts Denial: the claim is blatantly false
Denial: other 2) Denial Acknowledgement of allegation Justification of action 3) Establishing Apology
Plan for Change
4) Addressing
Other Depending on context
categorized under either 2), 3) or 4)
To facilitate the analysis, all types of denial responses are put as one ‘Denial’ category. Acknowledgements and justifications are regarded as companies establishing that the allegation in fact took place, which are seen as a separate category from responses that involve an apology or plan for change since such responses are more addressing and promising towards the CAA. ‘Other’ types of responses are evaluated and placed under category 2), 3) or 4), depending of the context of the response. To illustrate, statements such as ‘The company spokesman refused to respond to the allegation’ are categorized under Denial response here, since it is apparent in this scenario that the company was aware about the opportunity to address the issue. To some extent, the classification here thus follows Bradford and Garrett (1995), although it places ‘excuse’ and ‘concession’ together in one category as an Addressing response type and adds ‘acknowledgement of the allegation’ to ‘justification’ as an Establishing response type category. In case a MNE provided a response to a CAA that includes multiple aspects, such as denying direct involvement but promising a plan for change, the response is categorized under the response type that reflects best on the MNE (for this example, category 4) would be selected).
34 Moderating variable: Ownership
Ownership of the foreign subsidiary is taken as the moderating variable in order to analyze whether a springboarding perspective as described by Luo and Tung (2007) could explain for possible variance in CAAs by EM MNEs, with a transferring perspective explaining variance for IM MNEs as described by Surroca et al. (2013). In order for a springboard perspective to uphold, EM MNEs must display aggressive or risk-taking
behavior: this is accounted for in this research when ownership over foreign subsidiaries gets closer to 100% (see Meyer et al., 2009). On the other hand, minority-owned subsidiaries were explained to be relevant in transferring behavior by IM MNEs linking to CAAs as referred to by Surroca et al. (2013). In this scenario, IM MNEs must have non-marjority ownership of foreign subsidiaries and thus employ non-aggressive and less risky behavior. Percentage of ownership for foreign subsidiaries data is accumulated by checking MNEs in the sample on Orbis, a database providing company information worldwide.
Especially in the Apparel and Textile industry, a lot of CAAs in the CHRD are based upon violations that took place in a supplying factory which is not a part of the alleged MNE itself. For such cases, entry mode was marked as ‘Sourcing’, which is regarded as non-aggressive and non-risk taking behavior since it concerns a non-equity mode of entry and involves 0% ownership (Pan & Tse, 2000).
Control variables
Industry type and firm size are taken as control variables in this research, which have also been applied (amongst other control variables) by Salcito et al. (2015); moreover,
Kamminga (2016) investigates CAAs response rates for various industries. In the latter study, most allegation reports were addressed to companies in the extractive industry, with response rates varying for different industries. Furthermore, Li et al. (2010) mention industry-level and
35 firm-level factors when it comes to CSR communications to be of minor significance next to major country-level factors. As mentioned before, a specific type of ownership might occur specifically in a certain industry (i.e. sourcing in Apparel and Textile, see Emmelhainz & Adams, 1999). Firms in the sample belong to either the Natural Resources industry, Apparel and Textile industry, Technology, Telecom and Electronics industry, or the Utilities industry, with the majority of the MNEs falling under either Natural Resources or Apparel and Textile. Industry type is provided for each CAA in the CHRD.
Moreover, firm size is argued to affect the likeliness and type of response that is provided to CAAs. A suggestion here could be that large companies are being exposed to more pressure to provide a suitable response: this notion is also put forward by Kamminga (2016), highlighting that larger companies might be more visible to their critics.
Furthermore, one can look into studies that review corporate inclination towards CSR behavior and firm size. For example, Udayasankar (2008) mentions that distinctions in CSR motivations exist for various firm sizes, ranging from very small to very large. Lepoutre and Heene (2006) on the other hand find that small businesses experience more problems in engaging in CSR behaviors than large companies. Size for entries in the sample is determined for the foreign subsidiary, whereas in the case of sourcing the size of the MNE itself is used. Information on firm size can be found on Orbis; sizes that can be found in the sample are Very Large, Large, Medium and Small.
Sample selection
When selecting the sample, first the nationality of the MNE owner of the foreign subsidiary is established. In case the MNE has a parent company, the nationality of the parent company is used. CAAs for which no MNE nationality could be established are removed from the sample. MNE nationality is labeled as ‘home country’, whereas the foreign subsidiary
36 location is labeled as ‘host country’ here; for each home country, the corresponding
development level is assigned. In doing so, the independent variable is covered for in the dataset. Companies having home countries which are regarded as developing markets are not considered in the sample: DM MNEs are being left out in this analysis due to both a lack of theoretical background on their relation to CAAs and a lack of DM MNEs representation in the used dataset. Moreover, for the purpose of this research, countries for which home country and host country of the CAA were equal are not selected either in order to guarantee
multinationalism.
Next, CAAs for which no data could be found from Orbis on corporate ownership over the foreign subsidiary or corporate firm size are removed from the sample. For the remaining firms, it is first established whether a corporate response regarding the CAA was provided in the first place or whether this was UTD. Entries which classified as ‘UTD’ are removed from the sample. The type of response that was provided regarding the CAA based upon the categorization explained in the paragraph above is taken from the CHRD.
The described selection leads to a total of 516 CAA entries in the sample, addressed to 425 IM MNEs cases and 91 IM MNEs cases coming from 34 countries. It is important to note here that 516 different CAA entries does not equal 516 different MNEs, but 516 different cases of allegations; to illustrate, the 146 CAAs addressed to companies originating from the United States included 13 different CAAs concerning energy corporation Chevron and 9 towards sports apparel business Nike, Inc.
37 Results
Descriptive statistics for skewness and kurtosis are performed first, with a normal distribution having skewness and kurtosis close to 0. Normal distribution displays an ideal case of random data; for the used dataset, skewness and kurtosis pose an issue and there are no normally distributed items since skewness and kurtosis are not close to 0. Unfortunately, this issue is hard to eliminate since only 91 suitable EM MNEs entries could be taken from the CHRD compared to 425 IM MNEs entries. From the means, it can be seen that there is an almost equal probability that MNEs from either industrialized or emerging markets provide a response to a CAA in the first place; moreover, the mean for the response type is close to 2 which indicates a denial response being provided in general. On average, the MNEs’ foreign subsidiaries are majority owned (mean 57,90%). When checking the correlations between the variables (Table 4), it appears that whether or not a response is provided to CAAs is
correlated with what was taken as the moderating variable Ownership and with what was taken as the control variable Industry sector at the 0.01 level. However, these correlations do not indicate a tendency to a positive- or negative relationship which requires a score between either 0.2 and 0.5 or -0.2 and -0.5 respectively, since the Pearson r for Ownership is 0.154 and -0.121 for Industry respectively. The type of response is not significantly correlated with any variable that was considered in the research except for the other dependent variable providing a response or not in the first place, which is a logical relation here.
38 Table 4: Variable correlations
**. Correlation is significant at the 0.01 level (2-tailed).
The first analysis that is performed further checks the relation between MNE home country development level and providing a response to CAAs or not providing a response to CAAs, covering proposition 1. Since both the independent variable MNE home country development level and the dependent variable CAA response are categorical, cross tabulations are run first (Table 5). With the Pearson’s Chi Square being p>0.05, it has to be concluded that there is no significant relation between MNE home country development level and whether or not a response regarding a CAA was provided by that MNE.
Table 5: Cross tabulation
MNE development level Response χ2 φ Yes No IM MNEs 242 183 0.052 -0.01 (-0.1) (0.1) EM MNEs 53 38 (0.1) (-0.2)
A binary logistic regression confirms this finding that MNE home country development level is of no statistical significance in the relationship (Table 6). From the Nagelkerke R2 it can be seen that the complete model including all variables only explains for 9.9% variance. In the first block, control variables Firm Size and Industry Sector are inserted. The second block adds the independent variable MNE home country development level, with
Mean St. deviation 1. MNE nat. 2. Resp. y/n 3. Resp. type 4. Ownership 5. Size 6. Sector 1. MNE development level 0.82 0.381 X -0.10 0.030 -0.243** 0.062 0.202**
2. Response y/n 0.57 0.495 -0.10 X 0.796** 0.154** 0.038 -0.121**
3. Response type 1.92 1.004 0.030 0.796** X 0.027 0.073 0.007
4. Ownership 57.90 42.688 -0.243** 0.154** 0.027 X -0.378** -0.655**
5. Firm size 2.721 1.3737 0.062 0.038 0.073 -0.378** X 0.346**
39 the third block accounting for the moderating variable Ownership. Predicted probability is implied for providing a response to CAAs rather than not providing a response; for Industry Sector, Natural resources is coded as (1), Apparel and Textile as (2), Technology, Telecom and Electronics as (3); for Firm Size, Small is coded as (1), Medium as (2), Large as (3). Industry Sector Utilities and Firm Size Very Large are taken as baseline categories. For MNE home country development level, EM MNEs are coded as (1) and IM MNEs as 0; moreover, for Responses, not providing a response at all is coded as (0) and providing a response as (1). Note that control variable Firm Size did generate significant results, as did moderator variable Ownership in the third block. This indicates that Firm Size and Ownership influence the likeliness of providing a response to a CAA rather than not providing a response. However, with their Exp(B) being close to 1, these effects are rather minor.
Next, a multinomial logistic regression gives insight how MNE home country
development level affects corporate CAA response type, i.e. if significant variance is present when it comes to IM MNEs versus EM MNEs giving Denial-, Establishing- or Addressing response types (Table 7), thus analyzing proposition 2. This model explains for only 6,5% of total variance in Response Types, by evaluating the Nagelkerke R2. Independent variable MNE home country development level and control variables Firm Size and Industry Sector are included as factors, with the moderating variable Ownership added as a covariate. The reference category that is being used in the analysis is category (1), providing no response at all towards CAAs. EM MNEs are coded as (0), with IM MNEs coded as (1). EM MNEs are found to have a significant lower likelihood (0.350) of providing an Establishing response to a CAA rather than No response; however, no significant results are found when it comes to MNE home country development level and the other Response Type categories of Denial- and Addressing responses compared to not providing a response to a CAA. A higher Ownership
40 percentage is found to have a significant greater likelihood of providing a Denial- or
Establishing response to a CAA rather than No response, with no significant result for providing an Addressing response rather than No response. The significant results that have been found are rather small, being very close to an Exp(B) of 1; to illustrate, an Exp(B) of 5 indicates a 5 times greater likelihood of a variable falling into a specific category.
This analysis also puts forward observed versus predicted frequencies related to the variables explaining for Response Types. Some of these have a Pearson Residual which is greater than 2, indicating that such a difference is remarkable. To illustrate, a Pearson Residual of 4.498 is found for providing an Establishing response to CAAs by Small companies in the Natural Resources Industry which are categorized as IM MNEs with 41% Ownership over their foreign subsidiary. This indicates that more Establishing responses were given to CAAs for this combination of variables than would be expected. For the complete oversight of observations versus predictions, see Appendix 1.
