Motives and sustainability performance
The effect of different motives for engaging in Corporate Social Responsibility on the
sustainability performance of business units in Dutch profit organisations
By
Kevin Delgado (S2610205) k.s.delgado@student.rug.nl
University of Groningen Faculty of Economics and Business MSc Organisational and Management Control
24 February 2017
Supervisor: Dr. H.J. van Elten Second assessor: Dr. Y. Karaibrahimoglu
2
ABSTRACT
Corporate Social Responsibility was introduced due to a growing concern among stakeholders about the sustainability level of businesses. As such, it has been seen as an extrinsically driven phenomenon for years. Over the last decade, researchers have also increasingly started to identify intrinsic motives. Various researchers have studied the differences between intrinsic and extrinsic motives, but no one has explored the differences in effect on performance between the two motives even though existing literature identifies various reasons for assuming that these differences do exist. This study thus addresses the following research question: What are the differences in effect on sustainability
performance between middle managers with intrinsic motives to engage in CSR and those with extrinsic motives, and how are these effects influenced by the use of formal controls?
To answer this question, survey study is conducted among 30 middle managers of Dutch profit organisations. The data is analysed using a multiple regression analysis. This current study distinguishes itself by examining differences in effect, rather than differences in characteristics or importance. Of the five hypotheses formulated about the relationship between motives for engaging in CSR and sustainability performance, support is found merely for hypothesis 1. Hypothesis 1 indicates that having intrinsic motives is positively related to sustainability performance. This study could contribute to existing literature by linking literature concerning extrinsic and intrinsic motives to literature about drivers of sustainability performance. Additionally, this study can create a certain awareness among middle managers concerning the manner their motives for CSR can unconsciously influence sustainability performance of their business. Nevertheless, further research is needed to answer the research questions, since the other four hypothesis are rejected. It is recommended that this research be performed again with adjustments that eliminate the limitations in this study.
3
1. INTRODUCTION
Over the last three decades, companies have been subjected to an enormous shift in the way they do business. For years they focussed merely on profit, but nowadays they try to anticipate pressures that derive from their entire operational environment (Carrol & Shabana, 2010). This new focus has to do with the fact that aside from shareholders, more and more of a company‟s
stakeholders have become concerned about its business (Lopez-Valeiras, Gomez-Conde & Naranjo-Gil, 2015). Such growing concern among other stakeholders stems from the fact that the side effects of companies
‟
business actions, such as economic inequalities and environmental pollution, become increasingly visible after industrialisation (Martinez et al., 2015). It is in response to such concerns that corporate social responsibility (CSR) was introduced.Since, CSR arose from a growing concern among stakeholders, many academic researchers have tended to consider it as an extrinsically driven phenomenon (Muller & Kolk, 2010). When examining drivers or motives for CSR, most researchers have merely highlighted institutional pressures or economic performance (Aguilera et al., 2007; Swanson, 1999; Miles & Covin, 2000). Over the last decade, however, more and more researchers have identified motives other than strategic ones for engagement in CSR. For example, Van de Ven and Graafland (2006) have examined the relationship between managers
‟
motives for engaging in CSR and their actual CSR-centred efforts. In their study, they find that managers can have various intrinsic and extrinsic motives for engaging in CSR. Graafland and Mazereeuw-Van der Duijn Schouten (2012) have analysed the motives of executives to take responsibility for labour-related, environmental and social aspects of their business; they distinguish various motives that can be characterised as extrinsic or intrinsic as well.Furthermore, Bron and Vidaver-Cohen (2009) have examined corporate motives for engaging in social activities and also identify both strategic and moral motives. Moral or intrinsic motives are seen as motives that derive from managers
‟
personal belief systems or their perceived moral duties towards society. Strategic or extrinsic motives are seen as motives that are considered to eventually contribute to a company‟s financial success.4
sustainability performance. The results of their study indicate that managers
‟
commitment to CSR is one of the most important drivers of CSP. This finding can be explained the fact that commitment to CSR measures the extent to which managers act independently, are consistent and accurate, treat others fairly and do their work properly for its own sake. These characteristics seem to be decisive for the success of CSR implementation and execution within companies (Rajapakse & Fernando, 2014). On the other hand, Dare (2016) has explored the relationship between motives for engaging in CSR and commitment. Dare (2016) supports the conclusion that individuals with moral motives forengagement in CSR are more committed to their role than individuals with instrumental motives. This conclusion is confirmed by van de Ven and Graafland (2009), who have found evidence that managers with intrinsic motives were more involved and willing to put effort into their CSR activities than managers with extrinsic motives. Furthermore, Muller and Kolk (2010) claim that firms with
managers having intrinsic motives for engaging in CSR, have more extensive and deeply rooted ethics programs than firms with managers who engage in CSR due to external pressure.
The findings of these papers reflect that a manager‟s level of commitment to CSR is an important driver of sustainability performance. On the other hand, the literature reflects that the level of commitment between managers with intrinsic and extrinsic motives differs considerably. These two findings indicate that having intrinsic motives is likely related differently to sustainability performance than having extrinsic motives. Nonetheless, the literature does not currently support this hypothesis. This survey study is thus conducted to investigate this omission, by examining the effect of having especially intrinsic and extrinsic motives on sustainability performance and the differences between these effects. In pursuit of this goal, this study tests how these effects are moderated by the use of formal controls. To examine the differences, the following research question is posed:
What are the differences in effect on sustainability performance between middle managers with intrinsic motives to engage in CSR and those with extrinsic motives, and how are these effects influenced by the use of formal controls?
The purpose of this research is to provide empirical evidence concerning the differences in effect on sustainability performance between managers having especially intrinsic motives and
managers having especially extrinsic motives for engagement in CSR. By examining these differences, this study attempts to link the conclusions of existing literature concerning motives for CSR to
5
necessarily mean that the differences can be explained by varying levels of commitment between the two types of middle managers. Future research should confirm if a different effect can be explained by different levels in managers‟ commitment. Additionally, this study could provide certain awareness among middle managers concerning the way their motives for engaging in CSR can unconsciously influence the sustainability performance of their business unit.
The research was conducted by means of a survey conducted among 30 middle managers of Dutch profit organisations listed on the transparency benchmark list. The decision to survey middle managers was made due to their autonomous role within a business unit and their influence on the operationalisation of sustainability strategies. To help ensure that the surveys were filled in sincerely and by the right individuals, the surveys were distributed and completed on site. The hypotheses were tested using a multiple regression analysis in which extrinsic and intrinsic motives were the
independent variables, sustainability performance was the dependent variable, and the use of formal management control was applied as a moderator.
In this study, five hypotheses were tested to identify an effectual difference between having especially intrinsic and especially extrinsic motives for engaging in CSR. Nevertheless, the only statement that can be made based on the outcome of this study, is that having intrinsic motives for engaging in CSR positively related to sustainability performance.
6
2. LITERATURE REVIEW AND THEORETICAL MODEL
This chapter offers an elaboration of the topics relevant for the current research. This chapter also acts as a basis for the theoretical model and hypotheses.
2.1 Corporate Social Responsibility
Companies all over the world were focussed merely on profit for years. In their business activities, they often did not account for related side effects. This attitude towards side effects derived mostly from either the fact that companies were simply unconcerned with these side effects or the fact that considering these side effects would not contribute to their profits. Moreover, considering these effects in their business activities could reduce their profits, since anticipating these side effects is often associated with costs (Carrol & Shabana, 2010). Over the last 25 years, this attitude has changed considerably. Lueg and Radlach (2016) claim that this change was mainly initiated by
industrialisation, since the side effects of a business`s activities became more visible after the industrialisation. The climate was apparently changing, environmental resources started to exhaust, and the economic inequalities were increasing all around the world. The change in attitude was enforced by the publication of the Brundtland Report (Krajnc & Glavic, 2005) composed by the World Commission on Environment and Development (WCED). The Brundtland Report addressed the imperative for companies to change their attitude towards their business activities on the basis that their activities can harm the global environment. The change of companies‟ attitudes towards the environment introduced social development (SD) which is concerned with optimising the level of sustainability of the services and products that companies provide. The WCED (1987) defines a development as sustainable when it „meets the needs of the present without compromising the ability of future generations to meet their own needs
‟
. As such, the CSR phenomenon was introduced. In some ways, CSR is synonymous with sustainability. Two other definitions are as follows:„Management‟s obligation to set policies, make decisions and follow courses of action beyond the requirements of the law that are desirable in terms of the values and objectives of society‟ (Mosley et al., 1996); and „Corporate social responsibility means that a corporation should be held accountable for any of its actions that affect people, their communities, and their environment‟ (Post et al., 1996).
7
operations and in interactions with stakeholders
‟
. Interaction with stakeholders is required because nowadays the demand for visible sustainability management is also derived from externalstakeholders, such as the customers, the media, the local governments and the society. Facing these external stakeholders, companies who desist from engaging in CSR can lose their sales, reputation and competitive advantage, or they can be sanctioned by local authorities (Lee & Saen, 2012).
Corporate social responsibility can be measured across by three dimensions: environmental performance, societal responsibility and economic contribution. These three levels are named as the triple bottom line (Krajnc & Glavic, 2005). The environmental dimensions refers to the minimisation of the environmental damage that business activities can cause, such as reducing the use of resources that cannot be renewed. The societal dimension refers to the optimisation of the well-being of
individuals and communities, such as social equity or a good working environment for employees. The economic dimension refers to optimisation of a company‟s financial efficiency (Krajnc & Glavic, 2005; Bansal, 2005). Hahn et al. (2010) argue that companies constantly have to strike a balance between the three dimensions. For example, in order to realize long-term social goals that can be associated with long-term financial goals, companies have to compromise on short term financial goals (Graafland, 2002).
2.2 Motives for Corporate Social Responsibility
As previously mentioned, the CSR implementation and practices vary by company. The variations derive from CSR‟s conceptual infancy, a lack of legislation and the fact that the implementation and practices of CSR is done mostly on a voluntary basis. Despite the voluntary character of CSR, companies can be sanctioned by the government when they are in conflict with, for example, the allowable C02 emissions or generally accepted labour conditions (Peloza, 2006; Barnet, 2006). Meeting the expectations of customers and society is hence important for a company to remain viable. This requirement can be explained by the „iron law of responsibility‟ phenomenon (Davis & Blomstrom, 1971). Davis and Blomstrom (1971) claim that „those who do not use power in a way society considers responsible will tend to lose it‟. This so-called law has become increasingly important since it was coined, because society and customers are today more involved in companies‟ business activities than they were before. Considering these findings, there can be a wide variety of motives for executives to engage in CSR.
8
The current study centres on examining how intrinsic and extrinsic motives are related to sustainability performance and how the effects of these motives differ. The independent variables are derived from a study conducted by Bron and Vidaver-Cohen (2009) who examined corporate motives for engaging in social initiatives. They identified three categories of motives: sustainability,
profitability and legitimacy. Sustainability is labelled as a moral (intrinsic) motive for engaging in social activities. Consistent with Van de Ven and Graafland (2009), moral motives are defined as motives that derive from the ethical duty to give back to society (Bron & Vidaver-Cohen, 2009). Profitability and legitimacy are labelled as strategic (extrinsic) motives. Bron and Vidaver-Cohen discuss that strategic motives are those pursued in order to warrant or optimise an organisation‟s reputation and financial performance. Strategic motives for engagement in CSR are often explained by the stakeholder, legitimacy or institutional theories.
Intrinsic motives can derive from two sources: a moral duty or altruism (Graafland &
Mazereeuw-Van der Duijn Schouten, 2012). Managers who act upon a moral duty are managers who engage in CSR because they feel internally obliged. This obligation can derive from ethical principles or religion. Managers with intrinsic motives due to altruism are managers who experience personal enjoyment from doing good or optimising the well-being of other individuals (Graafland &
Mazereeuw-Van der Duijn Schouten, 2012; Van de Ven & Graafland, 2009). Investing in developing countries, cooperating with non-profit organisations and protecting the environment proactively without considering stakeholder pressure or financial performance are typical initiatives of managers with moral motives (Carrol & Shabana, 2010).
Similar to intrinsic motives, extrinsic motives can also have two different characters: instrumental or institutional. Instrumental motives can be labelled as profitability motives. Bron and Vidaver-Cohen (2009) discuss that instrumental motives derive from managers‟ beliefs that social initiatives are positively associated with profit by warranting its consistency or increasing revenues. Various research has demonstrated that maintaining profits or increasing revenues could be
accomplished by social initiatives, because social initiatives can increase shareholder value, improve a company‟s market position, maintain a company‟s reputation or reduce transaction costs. For instance, Matten and Moon (2008) claim that the way in which companies finance their assets has changed considerably in recent years. Western companies now prefer equity financing over debt financing. Since engagement in CSR became an important investment criterion, being associated with social initiatives is crucial for gaining access to capital markets and increasing shareholder value.
Furthermore, Miles and Covin (2000) claim that CSR activities can positively influence the reputation of companies, as social responsibility has increasingly become an issue for companies‟ stakeholders, such as customers. In turn, a good reputation stimulates customers to buy goods or services, which increases a company‟s revenue and eventually its profits (Van de Ven & Graafland, 2009; Graafland, 2002). Although most studies show that CSR contributes to profit, some indicate that CSR is
9
companies that have overinvested in CSR (Cañón-de-Francia & Garcés-Ayerbe, 2009) or in whom customers perceived discrepancy between their green advertising and actual performance (Filbeck & Gorman, 2004). In addition to being characterised as instrumental, extrinsic motives can also be characterised as institutional. Bron and Vidaver-Cohen (2008), label institutional motives as
legitimacy motives. They argue that managers pursue institutional motives when they engage in CSR due to of institutional pressures. Institutional pressures are external forces that set boundaries concerning what is right or wrong. External pressures can take three forms: normative, mimetic and coercive. Normative pressures constitute pressures to meet specific norms; mimetic pressures are those that urge the copying of competitors to reduce uncertainty; and coercive pressures encompass pressure from formal and informal institutions to comply with rules or law (Liu, Ke, Wei & Chen, 2010). These pressures can be exerted by external parties, such as customers, suppliers, competitors, the
government, and the community. On the other hand, institutional pressures can derive from parties within the company including employees, shareholders, and the board of directors (Beddewela & Fairbrass, 2014). The fact that companies consider institutional pressures to be a motive for engaging in CSR can be explained by changes in the manner in which stakeholders assess a company‟s
legitimacy. Beddewela and Fairbrass (2014) argue that, engagement in CSR has today become an important criterion in stakeholder perceptions of a company‟s legitimacy, where legitimacy is defined as: „A generalized perception or assumption that the actions of an entity are desirable, proper, or appropriate within some socially constructed system of norms, values, beliefs, and definitions‟ (Suchman,1995). A change in a company‟s legitimacy has to do with changing social attitudes, stakeholder demands, and global concerns about sustainability. To illustrate, Bron and Vidaver-Cohen (2008) discuss that in the 21st century, customers expect companies to have a proper social agenda. Additionally, more and more stakeholders, such as employees, suppliers and investors prefer to be associated with companies that proactively care about the environment, support human rights, consider community demand, and monitor their social activities (Galaskiewicz & Colman, 2006; Aguilera et al., 2007). Not visibly engaging in CSR can result in consumer boycotts, penalisation by governments and sanctions from suppliers or shareholders (Peloza, 2006; Barnet, 2006). Furthermore, Barnett (2007) found that a good reputation benefits legitimacy. In turn, having a social agenda seems to be one of the most important criteria for maintaining a good reputation these days.
10
optimising their organisational performance. Furthermore, there is a significant difference in the level of commitment towards social activities between managers with intrinsic motives and managers with extrinsic motives. For instance, Van de Ven and Graafland (2009) find support reflecting that managers with intrinsic motives are much more willing to put effort into their CSR activities than managers with extrinsic motives. This finding is also confirmed by Dare (2016), who has explored the relationship between motives for engaging in CSR and managers` commitment to CSR. The outcomes of Dare‟s study indicate that individuals with moral motives are much more committed to their responsibilities and tasks than those with instrumental motives.
2.3 Sustainability Performance
Sustainability performance, which is synonymous with CSP, is defined by Barnett (2007) as „A snapshot of a firms` overall social performance at a particular point in time‟. Measuring social performance seems to be difficult due to of its qualitative nature. For example, in contrast to financial performance, which merely contains hard indicators (such as return on assets, return on equity), CSP contains both soft indicators (e.g. management‟s commitment to ethics, compliance, and stakeholders satisfaction) and hard indicators (such as CO2 emissions, water and energy waste) (Barnett, 2007). Corporate social performance is multidimensional resulting in a variety of CSP constructs, because the relevance of the dimensions varies (for example, by industry, company, country, size and society). To illustrate, Ioannou and Serafeim (2012) define overall social performance as a combination of social, environmental and governmental aspects, while Muller and Kolk (2010) explain CSP in terms of the extent to which the sustainability demands of external stakeholders are met. Furthermore, Lyon and Maxwell (2004) explain CSP vis-à-vis the level of safety and pollution, stating that the higher the safety and the lower the pollution caused by chemical products, the better the CSP. In the current study, CSP refers to the extent to which CSR goals are achieved, the quality of cooperation on the social agenda with other units, internal and external customer satisfaction concerning green products and services, employee satisfaction concerning green products and services, and compliance with the code of conduct deriving from a company‟s sustainability strategy.
11
(Bron & Vidaver-Cohen, 2008; Peloza, 2006; Barnet, 2006). Since the literature suggests that managers with intrinsic motives are associated with a high level of commitment to CSR, managers with extrinsic motives are associated with a high level of strategic trade-offs, and that CSP is driven by intrinsic and extrinsic motives, both variables are expected to be positively related to sustainability performance. As such, the following hypotheses were formulated:
Hypothesis 1: Having intrinsic motives to engage in CSR is positively related to sustainability performance
Hypothesis 2: Having extrinsic motives to engage in CSR is positively related to sustainability performance
Muller and Kolk (2010) studied drivers of CSP and found that managers` commitment to their CSR seems to be one of the most important drivers. Commitment measures the extent to which managers act independently, are consistent and accurate, treat others fairly and do their work properly for its own sake. These characteristics seem to be crucial in terms of CSP (Muller & Kolk, 2010). Rajapakse and Fernando (2014) confirm this by stating that executives‟ commitment to CSR is recognised as the key success factor for implementation and execution of social and environmental initiatives. They found that a lack of commitment to CSR of middle managers can act as a crucial barrier for proper implementation and execution of CSR. Given that the literature suggests that managers with intrinsic motives are more committed to CSR than managers with extrinsic motives, the positive effect of intrinsic motives on sustainability performance is expected to be stronger than the effect of extrinsic motives. Accordingly, the following hypothesis was formulated:
Hypothesis 3: The effect on sustainability performance of having intrinsic motives to engage in CSR is stronger than that of having extrinsic motives to engage in CSR
2.4 Formal Controls
Due to growing concern about sustainability, CSR is becoming an increasingly integrated part of organisations‟ corporate strategies. For this reason, it is essential for organisations to adapt
management controls to achieve goals related to CSR, as the use of controls is positively related to goal achievement (Miller et al., 2013; Epstein & Roy, 2001). Controls implemented for CSR management are labelled as sustainability management controls and can be divided into formal and informal controls. Anthony and Govindarajan (2007) define formal controls as: „regularized approaches to ensure the execution of individual activities by which managers influence other members of the organisation to implement organisation‟s strategies‟. They consist of both financial elements (such as budgets) and non-financial elements (including planning, performance
12
which can result in cooperation and coordination problems, since individuals within business units can serve deviating interests. Formal controls can nullify these problems by formulating boundaries with which every individual in a business unit should comply. These boundaries can vary from acceptable quality levels in products or operating procedures to production deadlines or financial targets. Formal controls are verifiable measures established to guide organisations towards desired outcomes. Anthony and Govindarajan (2007) define informal controls: „a system that fosters an organisational climate that is conductive to behaviour based on the organisation's values and beliefs of managers and employees. It is a system that is established by common values, beliefs, and traditions. This current study focuses solely on formal controls.
The existing literature contains reasons to assume that the effects of formal controls on performance can vary depending on middle managers‟ motives (intrinsic or extrinsic) for engaging in CSR. To illustrate, Van der Ven and Graafland (2006) state that managers with intrinsic motives, approach CSR as an end rather than as a means. This approach suggests that the goal of managers with intrinsic motives is to be as sustainable as possible, irrespective of institutional pressure. As such, these managers may strive to minimise their waste to zero units, while their shareholders expect them to minimise it to five units; they may also managers expect that their employees will continuously seek for sustainable solutions, even if this standard is not required by the organisation`s board. This
proactive and progressive attitude requires commitment (Rajapakse & Fernando, 2014). An
individual‟s commitment is strongly related to psychological empowerment. The more individuals are psychologically empowered, the more they will be committed to their tasks and responsibilities (Joo & Shim, 2010). Moulang (2015) confirms this commitment, stating that psychological empowerment is positively related to an individual‟s performance. Formal controls can inhibit psychological
empowerment. This can be explained by the fact that when individuals with intrinsic motives perceive external intervention as controlling, like formal controls are, their intrinsic motivation decreases (Van de Ven & Graafland, 2006). These findings indicate that the use of formal controls can nullify the decisive characteristics of both managers with intrinsic motives and their subordinates. For this reason, it is expected that the more managers use formal controls, the more they compromise on crucial characteristics, and the poorer they perform. Therefore, following hypothesis was formulated:
Hypothesis 4: The use of formal sustainability management controls negatively moderates the relation between having intrinsic motives and sustainability performance
13
otherwise be sanctioned by stakeholders (Peloza, 2006; Barnet, 2006). Such strategic trade-offs are mostly done on the basis of strategic goals (Van de Ven & Graafland, 2006). Since the use of formal controls contributes to achieving strategic goals, it is likely that the use of formal controls will help managers with strategic motives to realize their goals by providing structure and coordination. It demonstrates that the more these managers use formal controls, the better they are guided during their CSR activities and the better they perform. These managers often lack a proper level of intrinsic drive to perform as well as managers with intrinsic motives who do not make intensive use of formal controls (Van de Ven & Graafland, 2006). As such it is assumed that the use of formal sustainability management controls, will positively moderate the relationship between managers who have extrinsic motives for engaging in CSR and sustainability performance. Therefore, the following hypothesis was formulated:
Hypothesis 5: The use of formal sustainability management controls positively moderates the relation between having extrinsic motives and sustainability performance
2.5 Theoretical Model
Figure 1 demonstrates the theoretical model. The model shows, as previously mentioned that both intrinsic and extrinsic motives have a positive effect on sustainability performance. In addition, it shows that the positive lines from the independent variables towards sustainability performance, are affected by the use of formal controls. The arrow from the use of formals controls towards the relationship between intrinsic motives and sustainability performance is negative, and the arrow towards the relationship between extrinsic motives and sustainability performance is positive.
Intrinsic motives +
Sustainability performance
Extrinsic motives +
+ -
The use of formal controls
14
3. RESEARCH DESIGN
The research design elucidates the methodologies used to collect and analyse the data. Additionally, this chapter clarifies which measures and criteria were taken into account during these processes.
3.1 Domain
The companies in which the current study is performed are listed on the transparency benchmark list. This list contains organisations that are subjected to an assessment of the quality and content of their social reporting on an annual basis. By participating in the transparency benchmark, organisations gain knowledge about social reporting and manners to improve their social activities. In addition, they enable themselves to observe how they perform on CSR in comparison to other
companies (Ministerie van Economische Zaken, 2016). By using this list as a basis for this research, it can be guaranteed that the companies involved are concerned about CSR. The list exists of large Dutch profit and non-profit organisations. Because non-profit organisations do not consider profit in decision making, these organisations are excluded from this research. By including non-profit organisations, the outcome could become unreliable since could never be a motive for engagement in CSR. The list is ordered by branch which include consumer products, industrial goods, banking and insurance, construction and maritime, universities, real estate, services, energy, oil and gas, trading companies, retail, pharmacy, technology, transport, food and additional (Ministerie van Economische Zaken, 2016). Yet, the companies are randomly selected. This is done because this eliminates the possibility that the outcome of this study could be explained by characteristics of the branch rather than the relation between motivation and sustainability performance.
The choice to conduct the research in the Netherlands is made because of two reasons. First, it was important that organisations included in this research were doing business in a developed country since CSR is a far more accepted phenomenon in developed countries than in developing countries. Conducting this research in countries in which it is not accepted could diminish the internal validity of this study. This difference can be explained by dominant coercive and normative pressures concerning CSR in the organisational field in many developed countries (Abreu, Cunha & Barlow, 2015). Second, the surveys were required to be completed on site. Taking these criteria into account, accessibility was imperative. Thus, due to the researchers` residence in the Netherlands, this country is the most
accessible developed country and, therefore, chosen for this research.
15
employees would be incomparable with the relatively large business units. Furthermore, middle managers are chosen because they are generally individuals with the most discretionary power, who are responsible for the operationalisation of the sustainability strategy and exert the most influence on the way and extent to which management control information is used within business units (Olsen & Stensaker, 2014).
N Minimum Maximum Mean Std. Deviation
BU-employees 30 23 5000 422,47 979,667
Years in current position 30 0,5 25 5,217 5,4531
Table 1 - Descriptive statistics - Amount of employees within business units
When examining table 1 it can be observed that every included middle manager is responsible for a business unit consisting of least 20 employees. However, the difference between the minimum and the maximum and the standard deviation are relatively large. This reflects that the sample consist of a combination of both small and large business units. Such formation of the sample can rule out the possibility that the outcome of this study could be related to merely large or merely small business units. Additionally, table 1 illustrates the descriptive statistics of the amount of years that the middle managers are working on their current position. The mean of this study show that the managers surveyed, have on average more than five years experience. Therefore, there may be assumed that the managers have gained on average enough experience within the business unit to give valid answers.
3.2 Data Collection
The data was collected by making use of a survey. This method was chosen because the aim was to identify features of a relatively large group of individuals. In addition, the aim was to obtain general information in a relatively short amount of time. Therefore, survey research seemed to be the most suitable method (Song, Im, van der Bij & Song, 2011).
A team of five master students of the University of Groningen collected the data. This was done to increase the possibility to recruit 30 respondents who were approached by making use of the students` networks, cold calling, and emailing, within five weeks. The five students used the data in their own way, because they all had different propositions that they wanted to test.
The surveys were distributed and completed in person to guarantee that the right person completed the survey properly. This optimized the internal validity of the outcome of this study and the reliability of the answers ( Song, Im, van der Bij & Song, 2011). In addition, this created the possibility for the students to ask additional questions to clarify the respondents` answers to the survey.
The choice to only collect data through a survey was because this research had to be
16
spent valuable time on collecting data, which were not relevant for the purpose of their own study. Nonetheless, a limitation of a survey is that it provides perceptual data instead of factual data. The use of perceptual data, even when the surveys are completed on site, can diminish the reliability of the results.
3.3 Study Measures
3.3.1 Independent variables
The independent variables in this study were intrinsic motives and extrinsic motives. These two variables consist of 16 items that Bron and Vidaver-Cohen (2008) categorised as sustainability, profitability, and legitimacy. The items in the category sustainability were assigned to intrinsic motives and the items in the category profitability and legitimacy were assigned to extrinsic motives. The responses on these items were subjected to a Likert-scale of 1 to 7. A score of 1 on an item means that a manager does not take that consideration into account at all, while a score of 7 entails that a middle manager finds that item important. The factor analysis concerning the independent variables was the most complex. In contrast to the other two analyses, the aim was to identify two factors instead of one. The analysis resulted in five factors with an unclear character. Therefore, a
confirmatory factor analysis had to be performed. This is achieved by manually excluding items until two factors remain. One factor consisting of questions which Bron and Vidaver-Cohen (2009) characterised as sustainability and one factor consisting of questions that they labelled as profitability of legitimacy. Furthermore, various items were excluded because they scored below 0.50 on both remaining factors. In all factor analyses, a minimum factor loading of 0.50 is maintained. The excluded items are questions 3.1b improve image (ex), 3.1c remain competitive (ex), 3.1d regulation (ex), 3.1g problem solving (ex), 3.1m no good reasons (in), and 3.1n foreign networks (in). The result of the confirmatory factor analysis is presented in table 2.
Question Motive label Intrinsic motives
(in)
Extrinsic motives (ex)
3.1a Serve long-term interests X
3.1e Fulfill expectations X
3.1f Meet shareholder demand X
3.1h Share recourses with society X
3.1i Create financial opportunity X
3.1j Prevent future business
problems
X
3.1k Concern for society's future X
3.1l Personal satisfaction X
3.1o Be recognised for moral
leadership
X
3.1p Learn from social agencies X
17
The items included in the factor which forms the variable extrinsic motives have a Cronbach's-alpha of 0.806 and the items included in the factor which forms the variable intrinsic motives have a Cronbach's-aplha of 0.752. When taking a minimum Cronbach`s-Alpha of 0,700 into account, both factors are sufficient in terms of internal consistency.
3.3.2 Dependent variable
The dependent variable concerns sustainability performance. This variable is built out of eight items of the survey which cover aspects such as achievement of CSR goals, customer satisfaction, employee satisfaction, cooperation with other business units, compliance with guidelines, and overall sustainability performance. These items were subjected to a Likert-scale of 1 (very poor performance) to 7 ( very good performance). The second factor analysis assessed the items concerning sustainability performance. The survey contained eight questions related to the sustainability performance of a business unit. All the questions were relevant because they all indicate value judgements concerning sustainability performance. Nevertheless, two items scored lower than 0.50. Therefore, excluded from the study. The excluded items contain questions 4.1b and 4.1g. The sum of the remaining items had a Cronbach's-Alpha of 0.834. As a result, the factor which forms the variable sustainability performance is also sufficient in terms of internal consistency.
3.3.3 Moderating variable
The moderating variable is the use of formal management controls. The use of formal
management controls consists out of nine items concerning planning systems, performance indicators, results comparison, environmental and social information on employees, investment decisions, rules and procedures, description of functions and external communication. Similar to the other variables, this variable is subjected to a Likert-scale of 1 to 7. A score of 1 means that a middle manager uses it to a very small extent, while a score of 7 means that a middle manager uses it to a very large extent.
The final factor concerned the use of formal controls. The part of the survey covering
questions related the use of formal controls existed of 11 items. The final two were not included in the factor analysis since they did not reflect anything about the extent to which a formal control is used, making them irrelevant for the purpose of this study. Furthermore, no item scored lower than 0.50 which entails that no items were excluded from the study. The Cronbach's-alpha of this factor is 0.904. Therefore, the factor which forms the variable use of formal controls is also sufficient in terms of internal consistency.
18
3.3.4 Control variables
This study used two control variables to eliminate the possibility that the outcomes are dependent of non-used variables. Starting with the size of business unit, it can be argued that the rationale of a middle manager can be explained by size. The more FTEs in a business unit, the more difficult it could be to manage a business unit (Chrisman, Bauerschmidt & Hofer, 1998). Therefore, it can be likely that managers of business units with relatively many FTEs could use formal
management controls more often, than the managers of business units with relatively few FTE's, since formal controls provide structure and coordination within business units. Thus, the size of the business units can be significantly related to sustainability performance, seeing the positive relationship
between the use of formal controls and performance (Miller, Saldanha, Hunt &Mello, 2013. The second control variable is years working in current position. It could take some time before a manager becomes familiar with coordinating a business unit. Furthermore, standardisation can be solution to a coordination problem FBy taking this into consideration, it is possible that managers who are working in their current position for a relatively short time use formal controls more intensively than managers who have more experience since formal control information contributes to standardisation. Therefore, years working in current position can be significantly related to sustainability performance.
3.4 Data analysis
This research identified whether and how two independent variables, intrinsic and extrinsic motives, affect the dependent variable sustainability performance. Moreover, it was also ascertained if and how this relationship interacts with the use of formal controls. The analysis of these aspects was performed using a multiple regression analysis. This specific type of analysis was used because it is a tool that aids to identify whether multiple independent variables correlate with a dependent variable (Lewis, 2007). Additionally, it helps to determine if and how correlating variables interact with a moderating variable. Before the regression analysis was applied, an assessment was made whether the data met the requirements for a multiple regression analysis. It valuation was made on the basis on the following criteria: the data should be normally distributed and the relationship between the
independent and dependent variables must be linear (Lewis, 2007). The content of this study met all these requirements. To exemplify, normal distribution was confirmed by conducting a skewness and kurtosis test and linearity was tested by analysing a scatter-plot concerning the relation between motives and sustainability performance.
The multiple regression analysis was conducted using SPSS. The survey existed of the following components: general information, management information on sustainability, sustainability considerations, performance and conclusions. Subsequently, a subset of items was used of each component. Before the data was implemented in the multiple regression analysis, a factor analysis was performed to aggregate the items from the survey into four workable variables. This is done to
19
tested for the construct validity. Yet, the factor analysis excluded various items because they did not correspond with the factors. The exclusion of items was minimised to ensure that the factors remained valid. As previously mentioned, in case of sustainability considerations, the factor analysis resulted in five factors, while the assumption was that it would result in three factors. Additionally, the
distribution of the items was not in line with the Bron and Vidaver-Cohen‟s factor analysis. Therefore, confirmatory factor analysis had to be performed.
The analysis was performed in three steps. First, a model was provided containing the control variables to rule out any significant relationship between these variables and the dependent variable. Second, a model which illustrates the relationship between motives and sustainability performance was provided, to test hypothesis one, two and three. Finally, the interaction effect between the moderating and the independent variables were added to model 2 to test hypothesis four and five, resulting in model 3.
4. RESULTS
In this chapter, the outcomes of the analyses are examined. This chapter starts by highlighting the correlation matrix, followed by the descriptive statistics. The chapter concludes with an explanation concerning the outcome of the multi-regression analysis which contains three models.
4.1 Correlation matrix
Before the multi regression analysis was performed, the Spearman correlation coefficients were checked. The Spearman correlation coefficient reflects if and how the variables correlate with each other. This is done to statistically support that the assumptions made concerning the relation between the dependent, independent and moderating variables are sequacious. If the independent and moderating variables do not correlate with the dependent variable, testing the hypotheses would be useless in advance. The coefficients provide a first indication of the possible outcomes of the
hypotheses. However, this does not reveal anything about the significance of the relationships that are tested. Spearman correlation coefficient can lie in between -1 and 1. Here -1 means that the
relationships between the variables are completely negative, while 1 means that the relationships between the variables are completely positive.
20 Sustainability performance Intrinsic motives Extrinsic motives Formal controls Sustainability performance 1 0.588** 0.434** 0.417* Intrinsic motives 0.588** 1 0.689** 0.695** Extrinsic motives 0.434** 0.689** 0.507** 0.695** Formals controls 0.417* 0.695** 0.507** 1
** Significant at the level 0.01 *Significant at the level 0.05
Table 3 - Correlation matrix independent dependent and moderating variables
Table 3 also reveals that all non-dependent variables correlate with each other. This could be critical when they correlate with each other in a certain extent, which makes it impossible to assess the individual effect of a variable on the dependent variable. To rule that this is the case, all the
independent variables are tested on multicollinearity by checking the variance inflation factor (VIF). The VIF indicates how much change in one variable is caused by correlating with other variables. A VIF score above 3 indicates that there might be a multicollinearity issue, a score above 5 designates that it is very likely that there is a multicollinearity issue, and a score above 10 signifies a
multicollinearity issue exists. Taking this into account, VIF scores <10 are required to make valid claims of the individual effect of the independent and moderating variables. In this study, all scores in are below 3. The relationship between intrinsic motives and extrinsic motives scores the highest (VIF=2.443).
4.2 Descriptive statistics
The previous section highlighted the correlation issues between the variables used in this study. In this section the descriptive statistics are discussed. Table 4 presents the descriptive statistics of all variables based on 30 middle managers. The second column illustrates the range in which the respondents co
A significant figure in the table 4 below concerns the mean of the sustainability performance 4.2822. This is a moderated score reflecting that on average the sample does not excel, but also does not underperform in terms of sustainability.
uld assign values to the variables. As mentioned, a Likert-scale of 1 to 7 was employed for all variables.
21
variable and managers who assigned a low score to that specific independent variable. Furthermore, the mean of having intrinsic motives is 4.7067, while the mean of having extrinsic motives is 5.1017. It suggests that on average the sample scores higher on having extrinsic motives than having intrinsic motives.
Range Minimum Maximum Mean Std.
Deviation Intrinsic motives 1-7 1.60 6.60 4.7067 1.11632 Extrinsic motives 1-7 1.60 6.80 5.1017 1.09768 Formal controls 1-7 1.00 6.44 3.5444 1.42332 Sustainability performance 1-7 2.67 6.00 4.2822 0.94637
Table 5 - Descriptive statistics independent, dependent and moderating variables
4.3 Multiple regression analysis
In this research, the five formulated hypothesis were tested by conducting a multiple regression analysis. Before the multiple regression analysis was conducted, factor analyses were performed to discern four reliable variables. After the factor analyses, the data were tested on their distribution and the variables were tested on multicollinearity issues. As the outcome of these tests was sufficient the multiple regression analysis was performed.
The analysis was done by means of three steps. The first step is not taken to test a hypothesis, but to rule out the possibility that sustainability performance is significantly related to the size of the business unit and the amount of years the middle managers have been in their current position. This step provides a model (Model 1) in which the relationship between the control variables and the dependent variable is demonstrated. Model 1 reveals that the effect of both the size of the business unit and the number of years in the current position are insignificantly related to sustainability
performance. As a result, it can be concluded that the sustainability performance of the business unit of the included organisations does not depend on the size of the business unit or on the number of years that a middle manager works in his or her current position. The R-square of Model 1 is 0.091, which means that 9.1% of the total variance of sustainability performance can be explained by the two control variables. This is a relatively low determination coefficient.
The second step of the analysis was providing a model (Model 2) which presents the
22
total variance of sustainability performance can be explained by the variables included in Model 2. With respect to Model 1, this is an improvement (30.5%). Nonetheless, it is still a moderate determination coefficient. The outcome of this analysis was essential because an insignificant
relationship between one of the two motives and the dependent variable could make testing hypothesis four and five irrelevant. This model was created by adding the independent and moderating variables to Model 1. As such, it can be observed that both having extrinsic motives for engaging in CSR and the use of formal controls are insignificantly related to sustainability performance. Thus, no support is found for hypothesis 2 and therefore, hypothesis 2 is rejected.
In contrast to hypothesis 2, support is found for hypothesis 1. As is evident in the outcome of Model 2, having intrinsic motives (B = .452, p <.10) is positively related to sustainability performance. This means that the better middle managers score on having intrinsic motives for engaging in CSR, the better they perform on sustainability. This is consistent with the expectations formulated by using the papers Dare (2016) and van de Ven and Graafland (2009). Yet, this does not mean that having intrinsic motives generates a good sustainability performance. As it merely denotes that sustainability performance is positively related to intrinsic motives for CSR activities. Because hypothesis 2 is rejected, hypothesis 3 is rejected as well. In contrast to having extrinsic motives, having intrinsic motives for engaging in CSR is significantly related to sustainability performance. Nevertheless, it does not reveal anything about the strength of having intrinsic motives with respect to extrinsic motives.
The final step was providing a model to test hypothesis 4 (The use of formal sustainability management controls negatively moderates the relation between having intrinsic motives and sustainability performance) and hypothesis 5 (The use of formal sustainability management controls positively moderates the relation between having extrinsic motives and sustainability performance). This model was developed by adding the interaction effects between the independent variables and the moderating variables to Model 2. Before the interaction effects could be implemented in Model 3, the variables were mean centred. By mean centring the zero points are redefined, which increases the interpretability of the results. The R-square of Model 3 is .410, reflecting that 41% of the total variance of sustainability performance can be explained by the variables included in Model 3. Although this is a slight improvement of Model 2, it is still insufficient to label the determination coefficient as strong.
Additionally, Model 3 only provides one significant relationship which is the relationship between having intrinsic motives and sustainability performance (.466, p <.10), revealing that hypothesis one is supported by both model 2 and 3. Furthermore, as no support is found for
23
Model 1 Model 2 Model 3 Hypothesis
Step and variables B SE B SE B SE Intercept 4.288*** .215 4.288** .215 4.265* .227 Control variables Business unit employees .000 .000 .000 .000 .000 .000 Years in current position -.020 .032 -.018 .031 -.018 -.033 Main effects Intrinsic motives .452* .233 .466* .236 Hypothesis 1,3 Extrinsic motives .093 .194 .045 .228 Hypothesis 2,3 Use of formal controls -.015 .150 -.012 .162 Interaction effects Intrinsic motives * Use of formal controls . .182 .259 Hypothesis 4 Extrinsic motives * Use of formal controls -.186 .268 Hypothesis 5 R square .091 .396 .410 Change in R square .305 .014
Dependent variable: Sustainability performance Valid N = 30
*p<0.10 **p<0.05 ***p<0.01
B= Beta
SE= Standard error
Figure 2 - Results multiple regression analysis
4.4 Robustness tests
24
reduces the reliability of survey results. Furthermore, the unreliable responses of managers can create correlated errors, which mean that motives and sustainability performance may correlate with each other merely due to the unreliable responses of middle managers. Correlated errors can result in false positive and negative relationships (Meier & O`Toole, 2013).
The risks of common source bias could be nullified by using archival data instead of perceptual data deriving from the same source employed to establish the motives for engagement in CSR. To verify whether the results of this research are affected by common source bias, Harman`s (1976) single-factor test was applied using SPSS. In this test, all of the latent measuring items are put together in one common factor and tested for variance. If the variance for that factor is below 50%, it may be assumed that common source bias does not affect the results. Since the total variance of the common factor in this study is below 50% (i.e. 33.12%), it is assumed that common source bias does not affect the results. Nevertheless, this method contains some limitations that reduce the reliability of its outcomes (Kemery & Dunlap, 1986). As such, the results were also checked for robustness after the multiple regression analysis was performed. This is done using SPSS as well. Robust results optimise the structural validity and minimise the possibility that the relationship between motives and
sustainability performance can be explained by common source bias. It was preferred to check the robustness by replacing the perceptual performance data by archival data, to see whether perceptual data is consistent with the real data. In case it was; the results are robust and are not affected by common source bias. Due to insufficient time for gathering archival data, the robustness of the results was tested by taking a random selection of 20 observations and checking whether the results of the multiple regression analysis remained the same. A random selection of 20 observations and check on the results, were completed 10 times. If the results remain the same after the sample is reduced and frequently changed, it may be stated that the results are robust (Lu & White, 2014).
25
5. DISCUSSION
This chapter elaborates on the previous chapter. Rather than only explaining the outcomes of the multiple regression analysis, this section highlights a critical interpretation of interesting results. This is done per model.
Model 1
Model 1 only contains the control variables. Although the relationship between the size of a business unit and sustainability performance is insignificant in this study, its beta reveals an interesting outcome (b =.000). It suggests that there is absolutely no relation at all between the size of a business unit and sustainability performance. It could be interesting to find out if the beta remains .000 in case the sample size is large enough, because currently it is not. Additionally, it can be interesting to find out if this also holds for the relationship between the size of the whole organisation and sustainability performance. Tari and Abdullah (2017) claim that larger firms are more associated with total quality management (TQM) and TQM is significantly positive related to performance. If this also holds for sustainability performance, it would mean that the larger the firms, the better the sustainability performance.
Model 2
26
Model 3
Model 3 reveals that by adding the interaction effects to model 2 all the included relationships change. Starting with the beta concerning intrinsic motives and sustainability performance. It
decreased from a value of 0.519 to a value of 0.223. This supports the assumption made concerning hypothesis four. It reveals that the relationship between the two variables is negatively moderated by the use of formal controls. This means that the more middle managers who have intrinsic motives for engaging in CSR make use of formals controls, the weaker the effect of having intrinsic motives on sustainability performance. Various middle managers surveyed in this research, had a clear
explanation for this. They claimed that managers with intrinsic motives for engaging in CSR often strive to meet their moral interest as optimal as possible, regardless what formal controls impose them to. By making intensively use of formal controls their focus can shift from meeting their moral ambitions to complying with formal controls. In many cases this could lead to a situation in which they expect their employees to produce slower, more costly and less environmental friendly because the formal controls allows them. In such situations, the moral ambitions of middle managers are more progressive than the content of formal controls. Despite the fact that this explanation makes sense, the p-value concerning the relationship between intrinsic motives and sustainability performance changes from significant to insignificant when the interaction effect with use of formal controls is taken into account. Hereby no meaning could be given to the moderating effect of formal controls on this relationship. Therefore, as previously mentioned, the outcome of this study does not support hypothesis four.
In contrast with the expectations, the relationship between having extrinsic motives and sustainability seemed to be positive in Model 2 and 3, but turned negative by implementing the interaction effect with the use of formal control. Seeing the explanations of several middle managers who were surveyed in this study, this is not surprising. They note that managers who have extrinsic motives for engaging in CSR are in principle managers that are focusing too much on formal controls. By focusing too much on formal controls middle managers overlook important aspects which actual benefit sustainability performance, such as optimizing the social environment for their employees or acting like a role model in order to stimulate employees to be as sustainable as possible. Because these kinds of managers often neglect those crucial aspects, it is not illogical that their business units often do not meet their targets and perform poor on sustainability level. By means of these explanations of various middle managers and the beta there can be concluded that the more managers with extrinsic motives for engaging in CSR make use of formal controls, the poorer their sustainability performance. Nevertheless, in this case there should again not be given any meaning to the relationship and its moderator because of a p-value of 0,477. Therefore, as previously mentioned, the outcome of this study neither support hypothesis five.
27
interaction effects, the relationships between the independent and the dependent variables are evidently influenced by the moderating variable. Nevertheless, its insignificance can probably be explained by the same reasons by which the insignificant results in model 2 are explained. Therefore, future research is also required to confirm this.
The first part of this chapter appointed that it would be interesting to find out if and why the size of companies is positively related to sustainability performance. Additionally, the last two sections revealed that future research is also required to retest hypotheses 2,3,4, and 5. The need for further research concerning these hypotheses, derives from the insignificance of the results. The insignificant results can probably be explained by some limitations in this study. These limitations are already briefly implied in the last two sections of this chapter. The next chapter will elaborate on these limitations and appoint other important limitations which justify future research.
6. LIMITATIONS
As mentioned in the research design various measures were taken to optimize the quality of this research. Although these measures were taken, it still was not free from limitations. This chapter highlights the most important ones. All the remarks presented in this chapter have to be taken into account while assessing the quality of this research. In addition, all the limitations appointed in this section justify future research, which elaborates on this study.
The first limitation which is already mentioned in the discussion, was the sample size maintained in this study. The sample consisted of 30 respondents. This is too small. In first instance, this does not have to be an issue in case it was a case study and the purpose of this study was creating depth concerning a certain phenomenon. However, it was a quantitative study so in this case it is an issue. A small sample size can reduce the external validity of the outcome and even influence the significance level of the outcome of the multiple regression analysis. Four out of five hypotheses were rejected. Insufficient data to test the hypotheses can be a serious explanation for it (van Aken et al., 2012). Furthermore, due to insufficient data, it cannot be confirmed if and how common source bias affect the results of the multiple regression analysis.
The second limitation was the exclusion of six items due to the factor analysis, because they did not correspond with one of the two factors. Four items were extrinsic motives and the remaining two items concerned intrinsic motives for engaging in social activities. According to Bron and
28
while improve image, proved to be the most important one. The exclusion of these items can be a consequence of insufficient data and an explanation for the insignificant results (Lewis, 2007).
The third important limitation is the fact that some of the respondents were not responsible for the whole business unit, but merely for a part of it. This can reduce the internal validity, because the sample contains managers who do not have the responsibility which is required to give valid answers to the survey (van Aken et al., 2012). The managers were required to have full profit and loss
responsibility. The fact that managers with deviating responsibilities were not excluded from this study, has especially to do with the sample size. By excluding managers with deviating
responsibilities, the sample size would become that small, that doing a proper quantitative study would be unrealistic in advance.
The fourth limitation concerned the fact that the branches were not taken into account. This deprives the possibility to verify if managers in one branch scores higher on one of the variables than managers operating in one of the other branches. This is quite surprising, because in not unthinkable that the extend in which middle managers can exert influence on for example task performance of employees, vary per branch. Consequently, nothing can be said about how much for example
sustainability performance depends on the branches in which middle managers operate, with respect to motives of middle managers for engaging in CSR.
The fifth limitation concerns the fact that a survey is based on perceptual data. This can be critically in case of sustainability performance, because the score on sustainability performance is merely based on the perception of the middle managers. Basing the score of sustainability performance merely on the perception of middle managers, may result in an output which basically is an
unrepresentative reflection of the reality. This limitation could be discarded by checking how the perception of middle managers was in accordance with for example key performance indicators (KPI's) and/or the output of a balanced scorecard (BSC).
The last remark is not specifically a limitation, but definitely worth mentioning. By focusing on the individual effect of intrinsic motives or extrinsic motives on sustainability
29
7. CONCLUSION
This study began by understanding that CSR is a phenomenon developed because Western companies` stakeholders became increasingly concerned with the sustainability level of businesses after industrialisation. Because it derives from external pressure, academic researchers have for years seen engagement in CSR as being extrinsically driven. Over the last decade various researchers have come to identify motives for engaging in CSR that go beyond extrinsically driven motives. For instance, Van de Ven and Graafland (2006) have identified that managers can have both intrinsic motives and extrinsic one to engage in CSR. This is also emphasised by Graafland and Mazereeuw-Van der Duijn Schouten (2012) who identified both intrinsic motives and extrinsic motives as well. Furthermore, Bron and Vidaver-Cohen (2009), who examined the motives for engaging in social activities, identified both strategic (extrinsic) and moral (intrinsic) motives. Remarkably, none of the papers concerning intrinsic and extrinsic motives for engaging in CSR examines the individual effect of the motives on sustainability performance nor the respective differences of the two kinds of motives, despite support for the assumption that differences exist. On the one hand a manager`s commitment to CSR seems to be the most important driver of sustainability performance; on the other hand, managers with intrinsic motives appear to be more committed to their CSR responsibilities than managers with extrinsic motives. Therefore, it was assumed that the effect of having intrinsic motives on sustainability performance is stronger than that of having extrinsic motives. It was also assumed that the use of formal controls moderates these effects. These assumptions were tested through the formulation of five hypotheses. The data used to test the hypotheses was collected through a survey of 30 middle managers from 30 different companies. These hypotheses were tested to answer the
following research question:
What are the differences in effect on sustainability performance between middle managers with intrinsic motives to engage in CSR and those with extrinsic motives, and how are these effects influenced by the use of formal controls?
The data is analysed with the use of a multiple regression analysis. Support was found only for Hypothesis 1 (Having intrinsic motives to engage in CSR is positively related to sustainability
performance). Because the other four hypotheses were rejected, the research question in this study remains unanswered.
30
than managers with extrinsic motives. By identifying a difference in effect, this research tried to link these two conclusions to each other. Further research should be needed to verify whether the
difference in effect really derived from a different level of commitment, since the absence or presence of other managerial characteristics could also explain why having intrinsic motives is more strongly related to sustainability performance (Muller & Kolk, 2010), than having extrinsic motives. However, the current research could be considered an elaboration on earlier studies regarding intrinsic and extrinsic motives for engagement in CSR. Previous studies aimed at identifying differences in characteristics and importance; this study adds to the differences in effect.
In addition to its theoretical relevance, this study could have a managerial relevance as well. By reading this report, managers could become aware of the different effects on sustainability performance that stem from having intrinsic or extrinsic motives. It is important for middle managers to note that this difference does not derive from the motives themselves, but from the behaviour that unconsciously arises from having intrinsic or extrinsic motives. Nevertheless, this study could only confirm that a difference exists; it could not identify which characteristics could explain the difference. Future research should thus explore the explanatory characteristics. If middle managers are made aware of behaviour that drives sustainability performance, they can adapt changes to optimise their performance. These changes could, for example, include the adaption of behaviour of highly committed managers by managers with extrinsic motives, since literature suggest that the
characteristics of committed managers are key success factors of the implementation and execution of CSR (Rajapakse & Fernando, 2014). Nonetheless, in order to be theoretically or managerially
relevant, the results must be significant.
