Master Thesis
Environmental CSR and Corporate Reputation
How does environmental CSR affect the corporate reputation and is
this relationship moderated by institutional quality?
University of Groningen
Faculty of Economics and Business
Groningen, June 17
th2019
Supervisor: Dr Olof Lindahl
Co-assessor: Dr Rieneke Slager
Submitted by
Neslihan Bekdemir
S3504069
ABSTRACT
In the last decades, consumers tend to care more about environmentally responsible behaviour. Therefore, corporate sustainability and environmentalism are getting increasingly important and popular for companies. The effect of Corporate Social Responsibility (CSR) on the firm’s corporate reputation has been investigated extensively in the literature. However, the relationship between the environmental pillar and corporate reputation has not been investigated broadly. Therefore, this study aims to contribute to prior research on CSR and corporate reputation with a novel approach. Compared to prior studies, the thesis analyses one component from the multidimensional construct CSR, which allows an in-depth understanding of the effects of environmental dimensions of a firm on corporate reputation. More precisely, the influences of environmental CSR (ECSR) dimensions comprising the three pillars of Asset4 (emission reduction, resource reduction, product innovation) on corporate reputation, operationalised by brand value. Furthermore, the moderating effects of institutional quality on the relationship are analysed. Contrary to previous studies, investigating these indicators in isolation, this study aims to analyse the total effects. An Ordinary Least Square (OLS) has been conducted with the environmental Asset4 pillars and a sample of 119 companies from 27 countries and 11 sectors. In contrast to the postulated hypotheses, the relationships, including the moderating effects, were found insignificant. However, some significant results are found in the control variables (Industry, Firm Size, Financial Performance), which show certain effects.
Keywords: Corporate Social Responsibility, environmental CSR, Asset4, corporate
TABLE OF CONTENT
Abstract ... I Abbreviations ... IV
1. Introduction ... 1
2. Literature review ... 4
Corporate Social Responsibility ... 4
2.1.1 Corporate Social Responsibility ... 4
2.1.2 Environmental Corporate Social Responsibility ... 5
Corporate reputation ... 8
Relationship ECSR and corporate reputation ... 10
Institutional quality ... 11
Hypothesis development ... 12
2.5.1 The influence of emission reduction on corporate reputation ... 12
2.5.2 The influence of product innovation on corporate reputation ... 13
2.5.3 The influence of resource reduction on corporate reputation... 14
The moderating role of institutional quality ... 15
Conceptual model ... 17
3. Methodology ... 18
Data collection and sample ... 18
3.1.1 Measures ... 18 3.1.2 Sample ... 22 4. Preliminary analysis ... 24 Model description ... 24 Preliminary analysis ... 25 5. Results ... 26 Descriptive Statistics ... 26 Regression results ... 29 Moderator analysis ... 32 6. Robustness Check ... 34 7. Discussion ... 35 8. Conclusions ... 38 Theoretical implications ... 38 Managerial implications ... 38
Limitations and future research ... 39
References ... 42
9. Appendices ... i
Appendix 9.1. Frequencies of Countries ... i
Appendix 9.2. Frequencies of Industries ... ii
Appendix 9.3. Independence of errors ... ii
Appendix 9.4. Normality Test ... iii
Appendix 9.5. Homoscedasticity ... iv
Appendix 9.6. Linearity ... iv
Appendix 9.8. Outliers ... v
ABBREVIATIONS
Abbreviation
Corporate Social Responsibility CSR
Corruption Control CC
Environmental Corporate Social Responsibility ECSR
Global Industry Classification System GICS
Government Effectiveness GE
Ordinary Least Squares OLS
Regulative Quality RQ
Research and Development R&D
Return on Assets ROA
Rule of Law RL
1. INTRODUCTION
Corporate Social Responsibility (CSR) is a relatively old concept, dating back to the 20th
century (Carroll, 1999). However, CSR is getting increasingly important for companies and their business today (Baughn, Bodie, & McIntosh, 2007). A common definition for CSR is proposed by McWilliams & Siegel (2001: 117), which underlines “actions that appear to further some social good, beyond the interests of the firm and that which is required by law”, assuming that CSR is a concept which is dealing with all actions beyond a company’s interests and not demanded by law. Those actions include social aspects such as human rights, sustainability or philanthropy. CSR is also defined as multiple factors, which influence the brand building of firms (Chomvilailuk & Butcher, 2010). Drivers of CSR are divided into social, economic and environmental aspects (Jenkins, 2009), also known as the triple bottom line, suggested by Elkington (1998). Therefore, CSR is of tremendous importance for companies, as they are engaged with different interests in economic, cultural, environmental or social environments. Also, because they directly affect those interests in society with their business and activities (Dobers, 2009).
Consequently, firms are responsible for protecting long-term profitability without damaging the environment or disturbing the society in the short-term (Elkington, 1998). Therefore, companies have strong motives to engage in socially responsible behaviour. This is not only for their self-interest but also to introduce their company as socially responsible to their stakeholders. Apart from businesses, consumers demand responsibility for the environment and tend to care more about CSR and sustainability before purchasing products (Öberseder, Schlegelmilch, & Gruber, 2011). Another motive is, especially regarding environmental responsibility, companies started focussing more on sustainability aspects in the last decades. This new focus can be traced back to current issues like global warming (Stanghellini, Marchello, & Michetti, 2008). The increasing relevance of environmental protection leads companies to focus more on environmental aspects in their CSR activities. Due to this increasing importance of the subject matter of CSR, this topic is highly interesting for many scholars.
CSR (ECSR) and corporate reputation was not enough in the focus of scholars. Due to globalisation, deregulation, innovation, higher competition, and uncertainty (Aaker, 2005), amongst other aspects, reputation is an important factor for companies but also for scholars. A promising approach is not only to focus on the concept of ECSR as a whole but to rather analyse the different sub-dimensions of ECSR. The support of CSR actions is not only influencing buying and loyalty motives but also rankings made by consumers (Sen & Bhattacharya, 2001) as well as their reactions to the firm’s products (Brown & Dacin, 1997). Social Media and public online rankings lead companies to focus more on their reputation (Dijkmans, Kerkhof, & Beukeboom, 2015) as it can influence their financial performance (Waddock & Graves, 1997). Based on these assumptions, CSR can lead to an advantage in financial performance, well-being and reputation of the firm in different ways (Marquina Feldman & Vasquez-Parraga, 2013).
In addition to the trend of being more environmentally active, the reputation of companies is also gaining more relevance for business activities. Corporate reputation, as an intangible asset, is helping companies to gain a competitive advantage as reputation is nearly impossible to imitate (Surroca, Tribó, & Waddock, 2010). This is also affected by the fact that consumers care more about the responsibility of brands rather than only about the quality of the products (Gugler & Shi, 2009).
However, the relationship between ECSR and corporate reputation do not play out a vacuum, and one interesting question is how this association is affected by the institutional context. There is a reason to believe that this effect is relevant as, for instance, Marano & Kostova (2016) state that institutional complexity in the organisational environment of a firm is influencing the CSR adoption. They also argue that CSR adoption is helping to gain a good corporate reputation and avoid negative publicity. For instance, a home country with a weak institutional environment and high degrees of corruption is affecting both the firm’s CSR activities and reputation (Hond, Rehbein, Bakker, & Lankveld, 2014). Additionally, in a highly corrupt home country, the firm could lose trust, which may lead to the loss of interest of consumers and negatively influence their buying behaviour (Sukhtankar, 2015). Previous studies have dealt with the topic of corruption or regulatory environment in isolation, but the moderating effect of four variables on country-level is still interesting to investigate.
Based on the arguments above, the following research question will guide this study:
How does environmental CSR affect the corporate reputation and is this relationship moderated by institutional quality?
This study aims to answer this research question with a dataset covering global companies. Four ECSR variables, institutional quality, and corporate reputation are operationalised. Therefore, a dataset consisting of 119 companies from 27 countries and 11 sectors is used, and an Ordinary Least Square (OLS) has been conducted. The relationships will be tested based on the conceptual model, which is presented later in this study.
2. LITERATURE REVIEW
In the following section, relevant literature for this research is reviewed. First of all, the concept of CSR is explained in general, followed by a specific definition of ECSR in the literature. In the next steps, corporate reputation and the relationship with ECSR is explored. Finally, literature regarding the moderating effects of institutional quality and the specific variables are defined.
Various studies in the past decades focused on the concept of CSR. However, only a few scholars investigated the dimensions (social, environmental, economic) of CSR in isolation. Moreover, this paragraph presents hypotheses and introduces the conceptual framework.
Corporate Social Responsibility
This section provides a brief overview of the concept of CSR and then move on to a specified explanation of the environmental pillar.
2.1.1 Corporate Social Responsibility
and a business strategy of the companies. However, the concept has a direct impact on society (Wan-Jan, 2006). Bowen (1953),on the contrary, as one of the first scholars, who investigated CSR, introduces CSR as “obligations of businessmen”, who should not just follow the financial aims of the company, but also focus on the “objectives and values of our society” (Bowen, 1953: 6). Comparing these different definitions, a possible common point that firms engage in aspects beyond financial resources and interests and take into account the needs of their stakeholders can be found. In this study, I will follow the definition of McWilliams and Siegel (2001: 117): “actions that appear to further some social good, beyond the interests of
the firm and that which is required by law”. Accordingly, CSR is a concept, going beyond the
personal aims of the company and which is not forced by legal conditions.
Apart from the assumption that CSR as an overall concept is leading to success, another approach is to narrow the concept down to its three pillars. Elkington, as the founder of the triple bottom line theory, suggests that firms should shift from the traditional point of view by measuring success with financial performance, to a broader concept, which includes all social, economic and environmental bottom lines (Elkington, 1998). Savitz & Weber (2014), who adopted the triple bottom line theory and went further, state that a company has an impact on the world. Specifically, a business is not only about investment but also about using environmental resources and materials as well as influencing society with its activities (Savitz & Weber, 2014). Carroll suggested in 1979 that the aim of the company should first be gaining profits. The goal should be to sell the product or service and have an economic success on the one hand and following the law by being legally responsible on the other hand. Carroll introduced his three dimensions named as economic, legal, ethical (Carroll, 1979) and later added one more dimension called philanthropic, which resulted in a pyramid model (Carroll, 1991). Philanthropic responsibilities for a company are on a voluntary level “to promote human welfare or goodwill” (Carroll, 1991: 42). He recommends a firm to “be profitable, obey the law, be ethical and be a good corporate citizen.” (Carroll, 1991: 42). While many scholars focused on different dimensions for the concept of CSR I follow the triple bottom line (Elkington, 1998) and specifically measure the effects of the environmental pillar, which is described in the following section.
2.1.2 Environmental Corporate Social Responsibility
Following this definition, the environmental sustainability contains “Ensuring that the overall
productivity of accumulated human and physical capital sustainability resulting from development actions more than compensates for the direct or indirect loss or degradation of the environment” (World Bank Group, 2008: v). These definitions contain the responsibility
of companies to take into account their resources, emissions and the impact of their production on the natural environment. The definition of the World Bank is underlying in this section.
Another definition is proposed by Mazurkiewicz (2004: 2) “a duty to cover the environmental implications of the company’s operations, products and facilities” and reduce emissions and waste, increase resource efficiency and decrease actions that destroy resources in the long-term (Mazurkiewicz, 2004). He focuses on the fact that companies are forced to be more transparent nowadays. Consumers are judging the practices and environmental behaviour of companies due to available media and increase of environmental knowledge.
energy-efficiency (Ramus & Steger, 2000). In line with this assumption, Bansal (2003) underlines the coherence of individual values and corporate values. Engaging in ECSR requires the cooperation of employees as well as stakeholders. Even if expectations of stakeholders with regards to (environmental) CSR implementation are often perceived as pressures (e.g. Sharma & Henriques, 2005).
According to Sen & Bhattacharya (2001), consumers are also stakeholders, and they can influence product or service evaluations with negative or positive CSR associations (Brown and Dacin, 1997) as well as affecting the brand choice and recommendation (Sen & Bhattacharya, 2001). On the one hand, Sharma & Henriques (2005) argue from a stakeholder perspective, that different types of stakeholder pressure affect the environmental engagement of the firms, based on variables similar to those Hart (1995) named as pollution control, eco-efficiency, recirculation, eco-design, ecosystem stewardship, and business redefinition. On the other hand, companies could engage their stakeholders to proactively work on ECSR strategies (Sharma & Henriques, 2005). Stakeholders can be divided into two groups named as primary stakeholders (customers, employees, suppliers) and secondary stakeholders (government, the local community, civil organisations) (Rhee, Park, & Petersen, 2018). In accordance with the stakeholder theory which suggests that a firm's financial accomplishment is based on the ability to implement a business strategy which is supporting the relationship with stakeholders effectively (Donaldson & Preston, 1995) companies have several reasons to engage their stakeholders in their CSR-strategies. Aguilera, Rupp, Williams, & Ganathati (2007) summarised those motives: first, the self-interest of the firm (instrumental), second, the relationship with other group members (relational), and finally the moral and ethical standards of the firm (moral).
organisational outcomes. Lower pollution, energy consumption, and the frugal use of resources are saving costs such as charges, fees, disposal, and fines for waste, licenses or other environmental arrangements. For instance, recycling and reusing materials or saving water and other resources can also be rewarded with grants by the government or other initiatives, which is also a cost saving aspect for companies (Sharma & Vredenburg, 1998). Companies working in the same sector, often face similar challenges perceived as pressures with regards to CSR activities, indicating that specific industries could benefit more from CSR than others. Companies are sometimes forced by their industry to implement specific policies (Jackson & Apostolakou, 2010). Especially, the fact that CSR and sustainability have different priorities across the industries (Babiak & Trendafilova, 2011), it can be interpreted either as a benefit or a pressure.
Corporate reputation
Similar to the concept of CSR, corporate reputation is gaining more attention from firms as well as from scholars (Brammer & Pavelin, 2004). Due to Globalisation, deregulation, de-intermediation, innovation and new technological inventions, an intense competition arises (Aaker & Mills, 2005). Therefore, companies hand back on corporate reputation, as an intangible asset, to gain a competitive advantage (Melo & Garrido-Morgado, 2012; Sabate & Puente, 2003). As a result, intangible forms of assets, specifically reputation, make it difficult to be replicated by competitors (Surroca et al., 2010).Corporate reputation is often seen as the degree of being good, admired, respected or being identified in high esteem by the community. This fact results in the hypothesis that the more positive a company is evaluated based on these factors, the more the company is trusted (Dowling & Moran, 2012). Trust, as a consequence, is a helpful mechanism for companies as consumers believe in the company’s integrity or potential for the future. A good corporate reputation is also based on the trustworthiness of both sides, the seller and the buyer resulting in a decrease of risk and transaction costs. Thus, reputation is important for the company as a reason for competitive advantage. For the stakeholders, reputation is also an assurance of honesty and commitment to long-term agreements (Dowling & Moran, 2012).
informative transparency with which the firm develops relations with them” (Sabate & Puente, 2003: 280). Wartick (1992: 34) defined corporate reputation as “the aggregation of a single stakeholder´s perceptions of how well organisational responses are meeting the demands and expectations of many organisational stakeholders”. Similarly, (Fombrun, 1996: 72) introduces the definition that corporate reputation is about “a perceptual representation of a company’s past actions and prospects” that describes how stakeholders see firm’s initiatives and evaluate the ability to deliver success. These definitions highlight the cumulative nature of corporate reputation, which reflects the expectations of stakeholders and therefore indicates a broader concept of corporate performance. Corporate reputation definitions differ based on disciplines such as economics or marketing. The definition of the organisation theory field states that “Reputations are representations of companies that develop as stakeholders make sense of corporate activities” (Fombrun, Gardberg, & Sever, 2000: 243). Even if the definition of organisation theory covers the importance of stakeholders, the marketing definition seems more suitable for the present study with regards to the focus on brand value: “Reputation
describes the corporate associations that individuals establish with the company name.”
(Fombrun et al., 2000: 243). The brand value scores in this study take into account consumer’s perceptions of brand equity. Therefore, it is important to follow the definition from a consumer’s point of view.
Relationship ECSR and corporate reputation
The general concept of CSR is often associated with an improved corporate reputation (Gardberg & Fombrun, 2006). According to Branco & Rodrigues (2006), CSR can boost the reputation with the different stakeholders. Consequently, CSR is seen as a potential factor that influences corporate reputation. In prior studies (e.g. Brammer & Pavelin, 2004; Maden et al., 2012; Melo & Garrido-Morgado, 2012; Pavelin & Brammer, 2006; Šontaitė-Petkevičienė, 2015) the influence of CSR on corporate reputation has been investigated. Fombrun & Shanley (1990), as one of the earliest works, contribute to the discussion with the result that greater improvement in social activities leads to a better corporate reputation. While, Brammer & Pavelin (2006) argue that the effect of social performance on corporate reputation varies across stakeholders as well as sectors and types of CSR, by using evaluations of managers and market analysts. Underlining the use of CSR as a corporate strategy, Rettab et al. (2009) found a positive relationship even in emerging markets, where no positive effects were expected due to lack of skills and communication to stakeholders.
In line with prior papers, a more recent and long-term study with large Spanish companies found that a positive relationship is depending on the activities of the firm. Therefore, the energy and oil sector with a known negative impact on the environment has a higher CSR intensity to compensate those issues and consequently gain more reputation. However, a negative moderating effect of the financial crisis is not supported (Singh, Garcia Sanchez De Los Salmones, & Rodriguez del Bosque, 2008). On the contrary, CSR is also seen as a tool for reputation risk management (Jacob, 2012) as companies pursuit to conduct their reputational risks by using their CSR reports (Bebbington, Larrinaga, & Moneva, 2008). Consequently, even in negative scenarios, there is evidence for a positive relationship between those two concepts.
strategies could result in a competitive advantage due to the growth of firm-specific competencies. Also, Toms (2002), for instance, found out that there is a positive correlation between information quality on environmental aspects and a better corporate reputation. Jackson & Apostolakou (2014) measured all three dimensions of CSR, where firms scored separately on each. The environmental pillar here includes policies, reporting, and performance on a broad scale of environmental indicators by using the following criteria: environmental performance / eco-efficiency, environmental policy, and management, environmental reporting. They demonstrate that country-effects are different in terms of environmental dimension, where Anglo countries scored higher than Nordic or Latin countries. Likewise, the industry had a significant influence on the environmental dimension of CSR (Jackson & Apostolakou, 2014).
The sub-dimensions of ECSR used in this study are namely emission reduction, product innovation and resource reduction. Pollution is a costly burden for companies on the one hand and a competitive disadvantage, on the other hand. Therefore, emission reduction mostly leads to a cost-saving advantage (Hart, Ahuja, & Arbor, 1996). In 1975 a new program, Pollution Prevention Pays (3P) was adapted to prevent pollution in advance rather than trying to solve it afterwards. The program was successful and led to a reduction of nearly 50% between the years 1975 and 1990 (Hart et al., 1996). The study of Hart et al. (1996) found a positive relationship between emission reduction and firm performance. It is a benefit and cost saving process for firms to act towards pollution. Firms also invest in CSR with the innovation of their products and services. The new attributes of these products indicate concerns about environmental issues so that the consumers believe. This innovation can also lead to trust within consumers and end in competitive advantage (Mcwilliams & Siegel, 2000). The resource reduction is more an issue about resource protection. Natural resources are endangered due to the increase of production over the years (Ayoo, 2007). With reducing the use of natural resources, long-term usage can be admired (Mcwilliams & Siegel, 2000).
Institutional quality
(1992): “the manner in which power is exercised in the management of a country’s economic and social resources for development.” This definition includes a broad description and focuses on the public sector. To decide on a useful description for this study, it makes sense to define institutional quality as governance, which is:
“the traditions and institutions by which authority in a country is exercised. This includes (a) the process by which governments are selected, monitored and replaced; (b) the capacity of the government to effectively formulate and implement sound policies; and (c) the respect of citizens and the state for the institutions that govern economic and social interactions among them.”(Kaufmann, Kraay, & Mastruzzi, 2011).
The definition allows an understanding based on the measures used in this study, named as Government effectiveness (GE), Regulatory quality (RQ), Rule of law (RL) and Control of corruption (CC).
A more detailed description of the moderating role of institutional quality can be found in
Chapter 3.6.
Hypothesis development
2.5.1 The influence of emission reduction on corporate reputation
Hart et al. (1996) suggest that working towards emissions through pollution prevention, for instance, may lead to a benefit for the firm. They suppose that within a few years after the implementation of this strategy, it will pay out by positively affecting the financial performance. Consequently, the relationship between emissions reduction and financial success is influencing the reputation of the firm (Hart et al., 1996). Accordingly, the environmental profile of the firm is influencing reputation and market value (Barth & Mcnicholst, 1995).
Even if environmental activities seem to be costly, and emissions reduction is attached to high expenditure, a positive relation in terms of corporate reputation can be expected. As environmental activities are gaining more attention from stakeholders as well as consumers, it can be argued that companies invest more in ECSR activities. From a consumer’s point, it is not only important that companies take measures to protect the environment, but which specific measures are taken. Therefore, firms that can show decreasing emission levels may get even more attention of consumers. As the awareness may increase with growing interest, specifically the case of pollution reduction may lead to a better corporate reputation. Based on these arguments, the following hypothesis is introduced:
Hypothesis 1: The higher a firm’s emission reduction, the higher is the corporate reputation. 2.5.2 The influence of product innovation on corporate reputation
According to Hart (1995), a competitive advantage can arise from innovative environmental strategies of the firm, which helps to develop specific capabilities. Achievements such as pollution prevention and innovation of environmental technologies or products are complicated to differentiate from the company’s additional (environmental) activities (Hart, 1995). The differentiation with the use of CSR resources (recycled products, organic pest control, etc.) also includes R&D, which results in both process in general and product innovation. Each of these is essential for consumers and their decision-making (Mcwilliams & Siegel, 2001). For example, a label called “organic” or “pesticide-free” includes both the organic method, which is a process innovation and also the invention of a new product, which is product innovation. Both indicators trigger the attention of consumers as they expect socially responsible attributes resulting in product innovation (Mcwilliams & Siegel, 2001). Therefore, a company is creating a reputation for quality and reliability (Fombrun & Shanley, 1990).
reputation. The visibility of sustainable production or ingredients is important for potential buyers. With upcoming trends like vegan life and the growing concern about dangerous ingredients and animal testing, the labelling of products is even more important. Consumers need proof of safe products to decide on that brand. The labelling may also lead to better recognition of the brand itself and also to better evaluation. Brands with higher ratings lead to a better corporate reputation, which is expected to be due to the innovation of products. Therefore, the second hypothesis is developed:
Hypothesis 2: The higher a firm’s product innovation, the higher corporate reputation. 2.5.3 The influence of resource reduction on corporate reputation
Land, mineral or water as natural resources are essential for the existence and improvement of human beings (Ziran, 1999). The productive utilisation and minimisation of natural resources got increasingly important due to economic, social and ecological advantages (Wang et al., 2011), which results in national industrialisation (Shen, Cheng, Gunson, & Wan, 2005). Policymakers have often discussed the optimal way to utilise natural resources. Conflicting stakeholder interests and the complex nature of resource management lead to difficulties (Ayoo, 2007). For example, China is facing big challenges regarding the utilisation of water due to the high population and shortage of resources (Wang et al., 2011; Ziran, 1999). Those problems influenced sustainable development (social and economic) of the country and resulted in new rules and regulations to prevent further damage (Ziran, 1999).
Nevertheless, the use of natural resources has a remarkable role in corporate performance and economic values of the communities (Ayoo, 2007). As companies based their CSR strategies firstly to meet their consumer’s wishes and expectations and secondly to care for the future (Khojastehpour & Johns, 2014), it can be argued that environmental activities are linked to the performance of the firm (Hart et al., 1996). Therefore, it is important how the consumer perceives specific environmental protection activities of the company. It is fundamental for the customer’s decision-making process that the CSR activities are fully transparent and visible to result in a successful CSR differentiation (Mcwilliams & Siegel, 2001). If the buying intention of the customer is created, it can lead to an indirect effect while it results in a direct effect when the CSR activity mirrors the consumer’s beliefs and expectations regarding CSR (Öberseder et al., 2013).
knowledge about natural raw materials and their limitation in nature is considerably high. Consumers care more about sustainable and long-term resources. Resource reduction is more a protective way to avoid the limitation in our natural environment. As this is moving more in the spotlight, increasing attention on the use of resources in companies is remarkable. Consumers get more in charge and want to know how the production behind the brand is working. Therefore, the visibility of strategies against the wastefulness of resources and for renewable materials may move more in the focus of consumers before buying products and services. Based on these arguments, the third hypothesis is created:
Hypothesis 3: The higher a firm’s resource reduction, the higher the corporate reputation.
The moderating role of institutional quality
In the literature, institutional conditions are often used to address the institutional level of analysis concerning CSR activities. These conditions are based on the three pillars of institutions by Scott (1995): normative, cognitive and regulative conditions. Regarding laws and rules, the regulative environment should be taken into account (Scott, 1995). Moreover, standards and certifications are also influencing CSR actions, which is the reason why regulative conditions are important for this research. When it comes to society and stakeholders or consumers themselves and the external conditions of the firm, cognitive and normative conditions are shaping the firm’s environment (Scott, 1995). Institutions can be seen as authoritative guidelines (formal and informal) that manage individual’s actions (Powell and DiMaggio, 1991) and support stability to social behaviour at the same time (Scott, 1995).
With the public policies they constitute, governments have a great impact on CSR. With the power they have, agendas for CSR initiatives are set up as well as assets are allocated by governments. Moreover, they establish environmental laws, change disclosure requirements of corporations and also affect labour laws (Tschopp, Wells, & Barney, 2013). According to Campbell (2006), corporations are more likely interested in socially responsible actions if there are strong regulations available to ensure such behaviour. Especially, the process of developing regulations, for instance, result from negotiations with government and stakeholders. Despite the voluntary nature of CSR initiatives, governments can act as an important factor in promoting those to the global market (Tschopp et al., 2013).
also act towards corruption in a socially responsible way based on different tools such as codes of conduct, employee awareness or compliance with rules. Therefore, it is also an expectation that companies integrate anti-corruption standards in their CSR activities (Carr & Outhwaite, 2011). A corrupt environment in the home country of a company can also affect the consumer behaviour named as “social effect” (Wang & Sun, 2010). Corruption levels describe whether laws are implemented and whether regulations are needed to assist corruption (Dam & Scholtens, 2008).
Regarding CSR activities, the political environment of the firm is crucial. The rule of law or government effects could take over the decision-making process of the firm when planning to act towards CSR. Specific rules can discourage CSR policies, which then affect the reputation. Since CSR is getting more important for consumers of “non-CSR” or “anti-CSR” lead consumers to discontinue that company and affect the reputation negatively. Especially towards companies with home-countries, which are known to be corrupt, the trust of consumers decreases. The corrupt environment can be the reason for the consumer to lose trust and stop the business. Accordingly, the company’s reputation is at risk. It is difficult for firms to gain trust again once it is lost.
All these variables and their effects might be investigated in isolation. I suggest analysing these four factors in one as they might be equally distributed and affect the corporate reputation in the same manner.
Taking these arguments into account, I expect a negative moderating effect of institutional quality on the relationship and put forward the following hypotheses:
Hypothesis 4a: Institutional quality negatively moderates the relationship between emissions reduction and corporate reputation.
Hypothesis 4b: Institutional quality negatively moderates the relationship between product innovation and corporate reputation.
Conceptual model
Figure 1 presents the conceptual model, including all hypotheses resulting from the literature
review.
3. METHODOLOGY
In this section, I present the different variables underlying this thesis. Moreover, the research methods, as well as the sample, are introduced.
Data collection and sample
This study aims to investigate the effects of ECSR on a company’s reputation and how this relationship is influenced by institutional conditions. To test the proposed hypotheses, I will draw upon quantitative research by using secondary data from various databases.
3.1.1 Measures
In the following section, the independent-, dependent- and control variables used in this study are explained.
Independent variables
To find a sufficient database measuring ECSR, various databases are reviewed as there is not a standard or ordinary measure for it (Gond & Crane, 2010). Other scholars (e.g. Ameer & Othman, 2012) highlighted the absence of a common standard to monitor issues like sustainability or CSR. Therefore, I reviewed existing and prominent databases such as Kinder, Lydenberg, and Domini (KLD) Database, the Global Reporting Initiative (GRI) regarding their availability, limitations, and efficiency for this research. The KLD database suffers essentially from coverage of global data as it is mainly providing US-based companies, which is not in line with the present model. On the other hand, the GRI database encounters more limitations as it is based on the disclosure of the companies. The databases work with information, the companies submit on their own, which could lead to bias. Moneva, Archel, & Correa (2006) critically argue that it is a risk to rely on the responsibility of the firms reporting the truth. Besides, there is evidence that results in practice differ from the GRI expectations (Moneva et al., 2006).
which are the three variables of this paper based on previous research (Montiel, 2008; Khojastehpour and Johns, 2014). The scores are ratings between 0 and 100 for each company, differentiated by normalised and adjusted raw scores (Refinitiv, 2013). The collection of the data is done by analysists who assemble ESG data yearly and transfer them in components that are suitable for quantitative analysis.
Dependent variable
Based on the research question, the dependent variable is corporate reputation. Many scholars used to measure reputation with conducting surveys and content analysis. In contrast, in this study, only secondary data is used. Finding a suitable measure for corporate reputation proves to be difficult. A general or unique database for corporate reputation is not existent. Therefore, I reviewed numerous private companies that measure different kinds of reputation. Most of these failed due to data availability. Data was only partly available, which makes it even more difficult. For instance, the Brand Strength Index (BSI) seemed to be a suitable measure. The database considers several factors to determine the strength of a brand: Marketing Investment, Stakeholder Equity, and the impact on Business Performance. The company/brand is assigned a BSI score between 0 and 100. With this score as a base, every brand is appointed to a rating with a maximum of AAA+ (Brand Finance, 2018). The scores for the present sample were not available and therefore, insufficient for this study.
Moderator
To test the moderating effects of institutional conditions, the measures for institutional quality from the World Bank are chosen. The moderator is created by the weighted average (Marano & Kostova, 2016) of all four components of regulative quality, rule of law, government effectiveness and control of corruption to reach an equal weighting of them. Therefore, the created moderator is covering four important factors from the institutional environment of the home country, influencing the strength of the main relationship between ECSR and corporate reputation. The Worldwide Governance Indicator (WGI), which is a dataset concluding the opinions on the quality of governance. The information is retrieved from numerous respondents on surveys in developing countries. The WGI variables are explained in the following (Kaufmann et al., 2011): Government effectiveness (GE), which is obtaining perceptions of the quality regarding the independence from political pressures and the quality of policy formulation amongst others. Regulative quality (RQ) with the direction on the implementation and formulation of policies and regulations by the government. Rule of law (RL) with the focus on the quality of contract enforcement amongst others. Control of corruption (CC), which is providing insights on the degree of misusing public power for private gain.
As four different variables on country-level were chosen, the first step is to transform them on the company-level of the other variables. Therefore, I adjusted the values of each variable to each company’s country of origin and got a complete list of institutional quality measure for the sample. To decide on an accurate way to use these variables, I conducted the Cronbach’s alpha method as also used by previous scholars (Fombrun et al., 2000; Wang & Sun, 2010). With the use of SPSS, the Cronbach’s Alpha being 0,9, which is the maximum of acceptable values (Field, 2009). Based on that reliable outcome, I created a new moderator with the average of all four variables, which is sufficient in this case.
Control variables
To control the conducted research, the following variables are chosen, which are expected to influence corporate reputation as well. The variables are namely: Firm size, industry, and financial performance.
Firm size
introducing their intentions and plans (Fombrun, 1996). Larger firms, therefore, tend to be more visible to customers as they have higher brand recognition as well as a higher scope. Larger firms also draw public attention and have a broader influence on society (Sotorrío & Sánchez, 2010). The firm size is measured by the number of full-time employees the firm has (Erramilli & Rao, 1993) in the present sample. The number of full-time employees for the latest available year is retrieved from the database Compustat and double-checked and completed with the annual report of each company if needed.
Industry
The industry is chosen as a control variable as the level of corporate reputation can differ between industries (Mcwilliams & Siegel, 2000). To investigate the role of the industry, I created a dummy variable based on the Global Industry Classification System (GICS) as the present sample contains global companies with different industries. The GICS system is structured into 11 sectors and narrowed down into industry groups, industries, and sub-industries. The classification of industry-groups is used for this research to clarify the influence. In total, 11 industries can be distinguished in the sample (See Appendix 9.3).
Financial performance
Additionally, to control corporate reputation, the variable financial performance is chosen. This variable is measured by the return on assets (ROA) of the company. This is a common approach to test financial performance (e.g. Kotha, 2001; Nelling & Webb, 2009). The ROA is calculated as the ratio of net income to total assets with the formula:
𝑁𝑒𝑡 𝑖𝑛𝑐𝑜𝑚𝑒
𝑇𝑜𝑡𝑎𝑙 𝑎𝑠𝑠𝑒𝑡𝑠
∗ 100
Formula 1: ROA
3.1.2 Sample
The sample is created by using public companies from the Forbes Global 2000 list. That list shows the largest global companies with regards to sales, assets, market value and profits
(Forbes, 2019). Due to human and financial resources, it is to expect that larger firms take more responsibility to implement CSR-activities. Consequently, higher differences should be found in larger firms compared to small firms, as they have more resources and opportunities (Udayasankar, 2008). On the other side, large firms are more visible for consumers, especially with their advertisement and online accessibility, which leads to the fact that this sample could give insights for both CSR and corporate reputation.
Moreover, the list contains firms from different countries. Therefore, the study is not focussing on only one region. Table 1 shows the list of all countries represented in the sample.
Country N % % cumulative Australia 3 2.5 2.52 Brazil 5 4.2 6.72 Canada 2 1.7 8.40 China 20 16.8 25.21 France 12 10.1 35.29 Germany 8 6.7 42.02 Hong Kong 2 1.7 43.70 India 2 1.7 45.38 Italy 5 4.2 49.58 Japan 4 3.4 52.94 Russia 3 2.5 55.46 South Korea 1 0.8 56.30 Spain 4 3.4 59.66 Switzerland 2 1.7 61.34 The Netherlands 3 2.5 63.87 UK 7 5.9 69.75 US 36 30 100.00 17 119 100
To investigate the effect on corporate reputation, the minimum sample size is calculated with the sample size formula (Khan Academy, 2019).
𝑝̂ ± 𝑧 𝑎 2⁄ √𝑝̂(1 − 𝑝̂) 𝑛
Formula 2 sample size formula
The formula includes 𝑝̂ for the sample proportion, 𝑧 𝑎 2⁄ , which is the critical value and, n stands for the sample size. The following dimensions are taken into consideration: a confidence level of 80%, which results in 𝑧 𝑎 2⁄ = 1,28, the margin of error not more than +/- 5%, response distribution of 0,5 (50%) and a maximised sample size proportion of 0,5 as the sample size is not known yet.
The calculation with the provided data leads to a result of n = 163,84 and consequently, a sample size of 164.
1.28√0.5(1 − 0.5) 𝑛 ≤ 5
Formula 3 sample size formula for this study
4. PRELIMINARY ANALYSIS
Model description
Various steps are taken to analyse the current dataset. In this study, the statistical approach Ordinary Least Squares (OLS) multi-regression is used. The OLS is the most appropriate tool to test multiple relationships with numerous independent variables in one model. Moreover, multiple regression is suitable as well when the dependent variable is continuous. To run this analysis, I used the software package IBM SPSS version 25 as the statistical instrument. Before running the OLS, several assumptions need to be validated. These assumptions consist of homoscedasticity, linearity, multicollinearity, normality and significant outliers which need to be validated with regards to the independence of observations (Tabachnick & Fidell, 2001). If the assumptions are not checked in advance, it can affect the results and therefore, the conclusions, and final interpretations as these are based on the OLS results.
Furthermore, a centring method has been used for the independent variables. Centring is a typical approach used in multiple regressions (Field, 2009). It helps to centre all variables around 0. Especially for the moderator, the mean centring was a helpful step.
Preliminary analysis
A preliminary analysis is conducted in advance to investigate if the assumptions above are in line with the data. In this section, an overview of all assumptions and the test results are provided and explained. Appendices 9.3 – 9.8 presents detailed information about the tests and their results.
With a first step, the independence of errors has been investigated. Appendix 9.3 illustrates the Durbin-Watson test result is 1,530. Therefore, the assumption of independence has been passed as values between 1 and 3 indicate a strong correlation.
Second, a histogram has been developed through the use of SPSS to check the normality. As to observe in Figure 2 in Appendix 9.4, the present sample is not perfectly distributed. However, the right shows more observations than the left side. The sample follows a normal distribution further as it is the most sufficient when comparing with other dispersions such as Leptokurtic, Uniform or J. Bimodal.
Third, a scatterplot has been created to test the homoscedasticity. In Figure 3 in Appendix 9.5, homoscedasticity of almost the same variances is presented with the help of a linear fit line, which leads to a confirmation of this assumption.
Fourth, the assumption of linearity has been tested through a probability-probability plot (P-P Plot). As presented in Figure 4 in Appendix 9.6, the point almost draws upon the line. Therefore, it can be predicted that the points follow the line and pass the test.
Fifth, a multicollinearity check has been examined, where the variables are tested on correlations between each other. The multicollinearity assumption affirms that two or more variables should not be related to each other very closely linear and exhibit a strong correlation as this situation would result in an unreasonable analysis of unique estimates (Field, 2009). To test the assumption, a variance inflation factor has been analysed. The highest VIF = 1.249 and the highest Tolerance of ,396 between all conducted models show that strong relationships between the variables can be eliminated. Consequently, no multicollinearity is to observe in the present sample.
5. RESULTS
After I controlled for the OLS assumption, the collected results and analysis are presented in the following section. First, descriptive statistics are introduced. Afterwards, the results of the OLS regression are presented, followed by the significance testing of the conducted hypotheses.
Descriptive Statistics
As all assumptions are passed, the descriptive statistics are introduced below. In total, there are 119 observations, as Table 2 shows. The corresponding home countries are presented in
Appendix 9.1. The dependent variable Brand Value shows values between 5,088 to 12,618
and a mean of 9,255.
Moreover, a big difference between minimum and maximum in the variable environmental innovation score is to observe. The values differ from 1,012 to 99,667, which is a difference of nearly 98. The lowest variance is given in the variable brand value with 7,53, which is at an acceptable level. The minimum values of brand value are 5,088 and the maximum 12,618. The ROA values differ between -0,302 and 7,73. Considering the means of the independent variables, it is to observe that these are quite high and more on the right side of the relationship rather than in the middle of the range. The mean of the environmental innovation score is 76,671; the emissions score 83,792 and the resource use score even 86,579. On the contrary, the mean of the dependent variable brand value is quite low with 9,255.
Descriptive Statistics
N Minimum Maximum Mean Std. Deviation
Brand Value 119 5.088 12.618 9.255 1.447799250
Environmental Innovation
Score 119 1.012 99.667 76.671 23.14878971
Emissions Score 119 12.624 99.918 83.792 15.02217150
Resource Use Score 119 7 99.797 86.579 16.34573472
ROA 119 -0.30 7.73 0.509 1.225738872 Fulltime Employees 119 320 2300000 178157 241459.310 Financials 119 0 1 0.42 0.496 Information Technology 119 0 1 0.08 0.266 Energy 119 0 1 0.12 0.324 Consumer Discretionary 119 0 1 0.08 0.279 Consumer Staples 119 0 1 0.06 0.236 Healthcare 119 0 1 0.01 0.092 Utilities 119 0 1 0.03 0.181 Real Estate 119 0 1 0.02 0.129 Materials 119 0 1 0.01 0.092 Industrials 119 0 1 0.03 0.181 Communication Services 119 0 1 0.0 0.279 Valid N (listwise) 119
Table 2 Descriptive statistics of the data
Correlations Brand Value Innovation Score Emission Score Resource Use Score
Brand Value Pearson Correlation 1 ,113 -,038 ,078
Sig. (2-tailed) ,220 ,685 ,399
N 119 119 119 119
Innovation Score Pearson Correlation ,113 1 ,455** ,474**
Sig. (2-tailed) ,220 ,000 ,000
N 119 119 119 119
Emissions Score Pearson Correlation -,038 ,455** 1 ,747**
Sig. (2-tailed) ,685 ,000 ,000
N 119 119 119 119
Resource Use Score Pearson Correlation ,078 ,474** ,747** 1
Sig. (2-tailed) ,399 ,000 ,000
N 119 119 119 119
**. Correlation is significant at the 0.01 level (2-tailed).
Table 3 Correlation matrix
Table 3 shows the correlation matrix. First, the correlations seem to be low and not
significant. The multicollinearity assumption affirms that two or more variables should not be related to each other very closely linear and exhibit a strong correlation as this situation would result in an unreasonable analysis of unique estimates (Field, 2009). As remarkable in Table
3, no relationships superior to R +- 0,8 are present. However, the relationship between
Regression results
The output of the statistical analysis is demonstrated in Table 5. Five models have been conducted. Following prior studies in this field (e.g. Brammer & Pavelin, 2006), an OLS has been accomplished. The independent variables Environmental Innovation, Emission and Resources have been analysed in isolation. Likewise, the control variables ROA, firm size and industries, as well as the dependent variable corporate reputation, have been analysed as well.
Model 1 Model 2 Model 3 Model 4 Model 5
b SEb b SEb b SEb b SEb b SEb Constant 9,157 ,384 8,746 ,561 9,776 ,873 8,804 ,805 9,427 ,887 IND_FIN -,225 ,412 -,332 ,426 -,227 ,413 -,231 ,414 -,382 ,426 IND_IT 1,007** ,577 ,910 ,585 1,016 ,578 ,975 ,583 ,798 ,585 IND_EN -,382 ,517 -,432 ,519 -,400 ,518 -,385 ,519 -,517 ,518 IND_CS -,620 ,634 -,571 ,636 -,639 ,635 -,627 ,636 -,651 ,634 IND_HC -,363 1,432 -,545 1,444 -,289 1,438 -,422 1,442 -,553 1,435 IND_UT -1,514** ,784 -1,480 ,785 -1,580** ,790 -1,520 ,787 -1,723** ,792 IND_RE -1,436 1,047 -1,287 1,058 -1,664 1,088 -1,314 1,079 -1,571 1,080 IND_MAT -,835 1,437 -,924 1,440 -,795 1,440 -,841 1,442 -,830 1,432 IND_IN ,037 ,818 ,029 ,818 ,003 ,821 ,036 ,821 -,095 ,816 IND_CSE ,639 ,565 ,682 ,566 ,588 ,569 ,636 ,567 ,504 ,572 Employees 1,314E-6** ,000 1,234E-6** ,000 1,300E-6** ,000 1,329E-6** ,000 1,225E-6** ,000 ROA -,670 ,000 ,780 ,000 ,888 ,000 ,897 ,000 ,759 ,000 InnovationSc ,006 ,006 ,007 ,007 EmissionSc -,007 ,009 -,025 ,013 ResourceSc ,004 ,008 ,016 ,013 Observations (N) 119 119 119 119 119 R ,426 ,435 ,431 ,428 ,464 R – Square ,181 ,189 ,186 ,183 ,215 Adjusted R - Square ,088 ,088 ,084 ,081 ,100 F - Value 1,937** 1,866** 1,830** 1,795 1,864** Prob > F (Sig.) ,038 ,043 ,048 ,053 ,036 ***. p<.01, **. p<.05, *. p<.1
In Table 4, the regression results are shown. A total of 5 different models have been conducted. In the first model, the effects of control variables on corporate reputation are measured, which means that all dummy industries, the ROA and the number of full-time employees (firm size) are a part of this model. The industry with sector number 25 (Consumer Discretionary) was removed from SPSS during the regression due to unpredictability. The study continues with ten industry specifications.
Model 2,3 and four have analysed the relationship between independent variables separately and the dependent variable. These models are based on hypotheses 1-3. In the fifth model, the independent variables were included all together to analyse the consolidated effects of them on corporate reputation.
Model 1 presents a coefficient of determination R-squared of ,181. R-squared measures the fitness of a variable with the variance percentage. This percentage determines the amount of changeability in one variable (independent), which is shared by the other variable (dependent) (Field, 2009). In this case, it means that the R² = 0,181 reflects the variation of control variables explain the variance of 18,1%. The F-value explains the significance of the model. In the present case, the model 1 is with the probability of ,038 insignificant. Looking at the control variables, only the number of full-time employees (p<,05) have a significant effect on corporate reputation. Moreover, the model shows the single effects of the Information Technology and Utilities industries. The ROA has no influences on corporate reputation, which is explained by model 1.
In model 2, the variation slightly increased to 18,9% after adding the Environmental Innovation Score to the model. However, the adjusted R-squared stay the same (,088) in both models, which leads to the assumption that adding one independent variable would not change the significance of the model. The variable innovation score with the B-value of 0,006 has no significant relationship with a corporate reputation in an isolated observation, which means that in the case of increasing one unit, the corporate reputation would rise of 0,006. Compared to model 5, the value raises to 0,007, which is still very low. Model 2 is as insignificant as model 1 with, even with a significant F-Value of 1,866 (p<,05). Consequently, hypothesis 1 stating that environmental innovation has a positive effect on Corporate Reputation is not supported as the regression does no find evidence for a positive relationship between those variables.
they can be observed as similar values. The adjusted R-square decreased to ,084, which was ,088 in models 1 and 2. A slight relevance of this model can be detected with the F-Value of 1,866 (p<,05). The utilities sector is the only significant sector in model 3, which is like model 1, where utilities and information technology showed a significant relationship. Even with a relevant F-Value, the third model is insignificant. Therefore, hypothesis 2, with a statement of emissions have a positive relationship with corporate reputation, has been rejected.
Model 4 presents an R-squared value of 18,3%, which is lower than model 2 and three but still higher than model 1. In this model, the variable Resource Use Score has been added. The adjusted R-square decreased again to a value of 0,081, which is lower than in the rest of the models. In comparison to the other models, no relevance by reference to the F-Value can be found in model 4. Moreover, the model is not significant. Therefore, resource use does not influence corporate reputation. In this model, none of the industries has significant results. Similarly, the control variable firm size (number of full-time employees) has a significant outcome as it was in models 1-3. Based on these results, hypothesis 3 stating that resource use influences corporate reputation, has been rejected as well.
Finally, model 5 shows the output by adding all independent variables to the model. The R-square in this model reaches the highest value, with 21,5% and the highest adjusted R-R-square of 0,1. The utilities sector remain significant again. Moreover, there is a repeated relevance of the model with an F-Value of 1,864 (p<,05). Since it was observed in isolation in models 2-4, the three independent variables have no significant relationship with a corporate reputation as proven with the regression results.
Moderator analysis
After investigating the first three hypotheses separately, the moderating effects of institutional quality composed of government effectiveness, rule of law, regulative quality and corruption control will be analysed in this section. In Table 5, the results of the moderator analysis are presented. Deciphering the F-Values of all four models, it is visible that only model 1 (1,937) has a significance. Adding the independent variables one by one, lead to insignificant results in all cases.
Model 1, considering the influence of control variables only, conclude a significant outcome. Observing the variables in isolation, no industry dummy or the ROA have significant influences. Again, the control variables number of full-time employees is significant in all models. In the present model, the B-Value of full-time employees is 1,314E-6 (p<,05).
The regression results for moderating effects show no significance for in all conducted models even though model 4 shows a slight difference. Whereas models 2 and 3 are quite similar in their outcome. First, adding the variable environmental innovation score lead to a B-Value of ,003 and a standard error SEb of ,007. Testing the model with the variable Emissions Score
causes a negative B-Value of -,015 and a SEb of ,010. The next case with the variable resource
use score shows a B-Value of ,000 and a SEb of ,009, which is slightly similar to model 2.
These implications show that the relationship between these variables separately and the moderator are not statically significant.
Another effect is that the adjusted R-square does not change when testing with the first variable in model 2 compared to model 1.
Model 1 Model 2 Model 3 Model 4 b SEb b SEb b SEb b SEb Constant 9,157 ,384 9,084 ,406 ,395 (0,943) 9,012 ,396 IND_FIN -,225 ,412 -,165 ,443 -,086 ,417 -,087 ,421 IND_IT 1,007 ,577 ,921 ,585 ,959 ,573 ,957 ,581 IND_EN -,382 ,517 -,176 ,553 -,069 ,538 -,095 ,546 IND_CS -,620 ,634 -,629 ,646 -,760 ,634 -,654 ,637 IND_HC -,363 1,432 -,524 1,444 -,293 1,422 -,439 1,437 IND_UT -1,514 ,784 -1,367 ,792 -1,503 ,792 -1,332 ,794 IND_RE -1,436 1,047 -,987 1,092 -1,255 1,101 -1,097 1,104 IND_MAT -,835 1,437 -,373 1,492 -,088 1,485 -,218 1,483 IND_IN ,037 ,818 -,048 ,821 -,179 ,817 -,044 ,821 IND_CSE ,639 ,565 ,703 ,569 ,574 ,563 ,659 ,565
Employees 1,314E-6** ,000 1,335E-6** ,000 1,367E-6** ,000 1,388E-6** ,000
ROA -,857 ,000 1,050 ,000 1,265 ,000 1,398 ,000
InnovationSC ,003 ,007
EmissionsSC -,015 ,010
ResourcesSC ,000 ,009
Inst.qual. ,286 ,202 ,417 ,203 ,341 ,207
Inst. qual. X InnoSc ,002 ,007
Inst. qual. X EmmisSC -,002 ,011
Inst. qual. X ResourSC ,006 ,010
Observations (N) 119 119 119 119 R ,426 ,453 ,469 ,453 R – Square ,181 ,204 ,220 ,205 Adjusted R - Square ,088 ,088 ,106 ,089 F - Value 1,937** ,055 ,046 ,335 Prob > F (Sig.) 0,038 ,815 ,831 ,564 ***. p<.01, **. p<.05, *. p<.1
6. ROBUSTNESS CHECK
After the regression, a robustness test has been conducted by changing the independent variables. The reason for that test was to check if the Asset4 scores only lead to the results of the main relationship. To change these variables and check the robustness seemed to be a suitable approach as ECSR consists of many other dimensions than the three used in the main model.
Therefore, the Asset4 variables have been replaced by other environmental variables from the Thomson Reuters database. These are environmental materials sourcing, CO2 equivalent emissions and the product responsibility Score. The results of the robustness check are presented in Table 9.
The analysis of those results leads to the assumption that the outcome is roughly the same when changing the independent variables. All models are insignificant. Moreover, the industries Information Technology and Utilities are significant in model 1. The number of full-time employees is a significant control variable in the robustness test, similar to the main regression. In the following models, the sector Utilities show a bare significance.
7. DISCUSSION
The ambition of this study was to analyse the relationship between different ECSR issues and corporate reputation as proposed in hypotheses 1-3. I wanted to explicitly investigate, if a company that principally engages in ECSR issues has a higher corporate reputation in terms of brand value, compared to those who do not engage in this pillar. Previous scholars generally suggested a positive relationship based on their research results. The second part of the conceptual framework aimed to look deeper into the moderating effects of institutional quality. The institutional quality is conducted by a country-level approach to analyse the influences of different issues in the institutional environment on this relationship. Prior literature mainly suggested a negative effect of institutional quality on the relationship of ECSR and corporate reputation as proposed in hypotheses 4a-4c. As mentioned in the analysis part, this study does not find statistical evidence to support the illustrated hypotheses. Regardless of the main relationships, small effects in different industries as well as the control variable firm size have been investigated and will be explained in detail later on.
In the previous section, the results of the robustness test were explained. The insignificance of the relationship can be replicated in the robustness check.
Contrary to the general opinion of theoretical research, the relationship between ECSR variables and corporate reputation is not positive in this research. Prior studies (e.g. Khojastehpour & Johns, 2014) found a positive relationship between climate responsibility and natural resource utilisation on corporate reputation based on primary research.
long-term and harm the company (Zou, Zeng, Zeng, & Shi, 2015). This lagged effect might have consequences on statistical data. To build a positive reputation, companies might need years of effort. Therefore, the results might be time-shifted in the case of this sample.
Even though previous research states that consumers are highly interested in a company’s CSR activities, it might still be that the majority of customers does not know about reduced emissions, resources etc. Proactive firms might not have enough publicity to promote their positive actions, whereas negative actions are immediately a part of mass media. Consumers have a more positive attitude toward environmentally active firms, in case they are informed about it (Distefano, Pisano, & Galvagno, 2013). Therefore, a positive influence on corporate reputation is difficult to expect in this case.
On the other hand, a minority of consumers who do not have knowledge or interest in environmentalism might not care about the ECSR activities. Therefore, these consumers do not help to build a positive reputation. Whereas, consumers with a personal interest in sustainability, for instance, tend to deal with the engagement of companies in CSR specifically.
As all three variables of ECSR seem to be insignificant in this data analysis, it might be based on the isolation of variables. Isolating the distinct variables might have made statistical distinction less significant. Therefore, it is possible that there is no relationship between these environmental dimensions and corporate reputation in isolation.
