SMEsmotivations for CSR: an exploratory study

Purpose – The paper aims to investigate the relationship between firms’ motivation for corporate social responsibility (CSR) and themoderating role of internationalisation. Design/methodology/approach – The authors developed and tested a conceptual model based on a survey of 65 respondents from the Møre and Romsdal (M&R) maritime cluster. The M&R maritime cluster despite being national has strong interconnections to the global maritime industry and as such, presents a suitable context for testing our research model. Findings – The findings show that firms’ intrinsic motivation drives CSR more than extrinsic motivation. Intrinsic motivation is understood as a firm engaging in CSR because it is the right thing to do and done out of one’s free will without compulsion or coercion. Extrinsic motivation relates to an action that is performed to achieve a separate outcome. Intrinsic and extrinsic motivations are found to be related and not mutually exclusive. The impact of intrinsic motivation on CSR was found to be contingent on the extent of the internationalisation of small andmedium-sized enterprises (SMEs). Originality/value – The key contribution of the study is the modelling of firms’ motivation for CSR activities and the contingent effect of internationalisation. In as much as companies perceive CSR activities as the right thing to do, the motive to do so also depends on the business case/profit motive. The study shows that SMEs’ intrinsic motivation is the driving force in CSR implementation and suggests that the urge by firms to give back to society is strengthened under conditions of high economic incentives and the firms’ degree of internationalisation.


Introduction
In the past, corporate social responsibility (CSR) has been optional, but today's global and competitive business environment allows no firm to neglect the importance of social responsibility (Graafland and van de Ven, 2006;Nejati and Amran, 2009;Scherer and Palazzo, 2008;Worthington et al., 2008;Svensson et al., 2010;Khojastehpour, 2015;Eriksson and Svensson, 2017). The accelerating trend of globalisation has highlighted firms' responsibility to society, along with a recognition that the negative externalities have increased because of globalisation (Scherer and Palazzo, 2008).
Major organisations, such as the Organisation for Economic Co-operation and Development, UN and EU, have also made CSR a priority. These organisations are, however, concerned that many firms consider CSR as relevant to big multi-national enterprises (MNEs) only (United Nations Industrial Development Organization (UNIDO), 2002;EC, 2002). Others have also raised this concern, and suggest that the term itself is misguided, and should be replaced by one to which all firms can relate (Jenkins, 2006;Nejati and Amran, 2009). More than 90 per cent of the world's companies are small and medium-sized enterprises (SMEs) (Perrini et al., 2007 in Vázquez-Carrasco andL opez-Pérez, 2013), and it has been estimated that SMEs have a greater environmental impact per unit than large firms and are the largest contributors to pollution, carbon dioxide emissions and commercial waste (Baden et al., 2009, p. 1). Thus, for CSR to be globally useful, it is vital that SMEs engage in this.
An investigation into the motivations driving CSR and CSR activities is important because the social responsibility of a firm depends on the individuals who make decisions, big and small, on a daily basis (Wood, 1991). Previous research has had mixed results. For example, Graafland and van de Ven's (2006) study found that managers were motivated by both intrinsic and extrinsic factors. Nejati and Amran's (2009) study of 10 Malaysian SMEs found that these practiced CSR because of personal beliefs, values and religious thoughts (intrinsic), as well as pressure and encouragement from stakeholders (extrinsic). Brønn and Vidaver-Cohen (2009) found that "improved image", "be recognised for moral leadership" and "serve long-term company interest" were most cited as reasons to engage in CSR. This contradicts Graafland and van de Ven (2006), who found intrinsic motives had a stronger relationship to CSR commitment than extrinsic. However, as Graafland et al. (2010) point out, the Norwegian sample used by Brønn and Vidaver-Cohen (2009) consisted primarily of large companies. Thus, it is reasonable to expect that the relationship between intrinsic motivation and CSR is more profound in smaller companies. The mixed results of previous studies on firms' motivations for CSR activities provide justification for the present study.
An enquiry into managers' motivations for CSR is thus valuable, as it enables a better appreciation of why firms act the way they do. The literature (Baden et al., 2009;Graafland and van de Ven, 2006) affirms that economic theory shows that extrinsic motivation (e.g. strategic motives such as profitability and reputation) can override intrinsic motivation (not for profit). An understanding of motives for CSR, and, more importantly, how motivation relates to CSR activities, provide useful insights for the industry itself, as well as for national and international authorities regarding policy formulation and implementation. We address these issues through the following research questions: RQ1. Does SMEs' intrinsic motivation drive CSR activities more than extrinsic motivation? And, are motivations for CSR activities contingent on the degree of internationalisation of SMEs?
While we know that CSR has become more important as corporations expand abroad because of different cultural and ethical demands (Bondy and Starkey, 2014;Khojastehpour, 2015), the literature is silent on the moderating role of internationalisation in the relationship between firms' motivation and CSR practices in the local environment. This raises important issues about what is considered universal and what needs adaptation to local circumstances. For internationalised firms, the environment becomes more complex and uncertain because of the many different external elements. Thus, the complexity of internationalisation could explain motives for CSR in the domestic environment, and hence, a distinguishing factor between firms that are internationalised or not. This study examines the moderating role of internationalisation to better understand the complex relationship between CSR, firms' motivation and international expansion.
Internationalised firms are shaping CSR through their practices and, therefore, are in a unique position to formulate and implement CSR in ways beneficial to them (Bondy et al., 2012;Khojastehpour, 2015). The purpose of this paper is thus multifold. First, it aims to fill the gap in the extant literature concerning the association between CSR and motivation. Second, most of the research has focused on CSR in MNEs (Jenkins, 2006;Perrini et al., 2007;Bondy and Starkey, 2014) while this study examines CSR within the context of SMEs that are internationalising. This is critical because SMEs lack the resources of MNEs. Larger firms are well endowed, are more likely to have a competitive advantage in international markets and can respond better to challenges (Paul et al., 2017). Finally, the study provides further insight into motivations for CSR practices and implications for theory building and practice.

Theory and hypotheses 2.1 Previous research on corporate social responsibility
In the past two decades, CSR has received increasing attention from scholars, national and international governments, media and the industries themselves (Virakul, 2015;Dabic et al., 2016;Eriksson and Svensson, 2015, 2016Jensen et al., 2018). The Appendix is a summary of some of the important research streams within this field.
There are many definitions of CSR, but a lack of consensus on one (Dahlsrud, 2008). We choose to use one that is simple but relevant for both MNCs and SMEs; "CSR reflects firm actions that further some social good, beyond the interest of the firm and that which is required by law" (McWilliams and Siegel, 2001, p. 117). Vázquez-Carrasco and L opez-Pérez (2013) describe CSR in SMEs to be "silent or sunken CSR" (p. 3211). CSR-related contributions linked to SME contexts have been defined as nonsystematic, unstructured and non-formalised within the global strategy of organisation (Russo and Tencati, 2009). This view is supported by Coppa and Sriramesh's (2012) study of 105 Italian SMEs, that CSR was mainly practiced through informal, internally oriented and relational methods with very little strategic focus. Because of SMEs' focus on internal motivations for CSR and an individual level of analysis (Preuss and Perscke, 2010), Vázquez-Carrasco and L opez-Pérez (2013) strongly recommend that future studies on CSR in SMEs be rooted in the social capital theory.
Vázquez-Carrasco and L opez-Pérez (2013) believe there are significant differences between SMEs and large MNEs in that SMEs are more focused on philanthropy compared to large enterprises, which focus more on economic/strategic CSR. SMEs tend to have closer relationships with stakeholders, attaching more importance to the local community, and have strong ethics and owner values. Large enterprises are characterised by more distant relationships with stakeholders, instrumental/economic objectives and are often more formalised in their approach. Carroll (1979) frames CSR as four different social responsibilities, which are depicted in a pyramid: economic, legal, ethical and philanthropic. However, recent studies frame CSR as three components (economic, social and environmental), which relate to the triple bottom line sustainability perspective (Vilanova et al., 2009;Santos, 2011). For purposes of operationalisation, we choose Carroll's (1979) conceptualisation. Carroll (1979) argues that, although all four elements have always been present in business, ethical and philanthropic responsibilities in recent years have become more important. Aupperle et al. (1985, p. 455), define the four components as:

SMEs motivations
for CSR (1) Economic responsibilities of business reflect the belief that business has an obligation to be productive and profitable and meet the consumer needs of society.
(2) Legal responsibilities of business indicate a concern that economic responsibilities are approached within the confines of written law. (3) Ethical responsibilities of business reflect unwritten codes, norms and values implicitly derived from society; ethical responsibilities go beyond mere legal frameworks and can be both strenuously undertaken and nebulously and ambiguously stated. (4) Discretionary responsibilities of business are volitional or philanthropic in nature, and, as such, also difficult to ascertain and value.
Although the four areas seem straightforward in theory, in practice, they may not always be clear. For example, philanthropy necessitates a reciprocal relationship between the philanthropist and the beneficiary where both, in some way, invest (Acs, 2013). While a philanthropist gives money, the recipient is expected to give time and energy to benefit from the largesse (Acs, 2013, p. 3). Further, the legal aspect is, perhaps, the clearest in theory: obey the law. The question remains, however: which law? Is it acceptable for firms to 'do in Rome as the Romans do'? The legal aspect is also somewhat problematic regarding the voluntary aspect of several CSR definitions (Gjølberg, 2009). With increasing internationalisation, firms often face situations where their operations are located in countries with other social standards than those at home. According to Carroll (1991), there is a natural fit between CSR and organisational stakeholders. Thus, stakeholder theory (Freeman, 1984) is an important contribution to the literature for clarifying to whom the firm is responsible, which is less clear in the CSR pyramid. Stakeholders are also considered a key component of SMEs' understanding of CSR (Jenkins, 2006). Additionally, legitimacy (Murillo and Lozano, 2009) has been an important reason for engaging in CSR as society grants legitimacy and power to business. In the long run, those who do not use power in a manner, which society considers responsible will tend to lose it (Davis, 1973, p. 314). We thus conceptualise CSR as a multidimensional construct measured by the firm's economic, legal, ethical and philanthropic responsibilities. Lynes and Andrachuk (2008) listed frequently cited motivations in the literature; long-term financial strategy, eco-efficiencies, competitive advantage, good corporate citizenship, image, delay or avoidance of regulatory action and pressure from stakeholders. These different motivations can be separated into two categories (Ryan and Deci, 2000). Hence, in line with Ryan and Deci (2000), we conceptualise motivation into intrinsic and extrinsic motivation. The intrinsic motivation here is understood as a firm engaging in CSR because it is the right thing to do (Ryan and Deci, 2000). Extrinsic motivation is perceived as the belief that CSR engagement leads to a separable positive outcome for the firm or in other words, the belief that there is a business case for CSR (Ryan and Deci, 2000). Graafland and van de Ven's (2006) study of 111 Dutch companies found that, although extrinsic or strategic motives and intrinsic motives were related, intrinsic motives had a stronger correlation with SMEs involvement in CSR. The same found that larger firms were more motivated by extrinsic factors. This is supported by the Norwegian study of Brønn and Vidaver-Cohen (2009) that also found significant differences regarding motivation between small and large firms, larger firms being more motivated by extrinsic motivation than intrinsic motivation. We can conclude that an individual may be both intrinsically motivated and extrinsically motivated for performing CSR activities (Bénabou and Tirole, 2010;Ryan and Deci, 2000). Based on the above arguments, we hypothesise that:

corporate social responsibility and motivation
H1. Intrinsic motivation (H1a) and extrinsic motivation (H1b) are positively associated with CSR activities.
Although Ryan and Deci (2000) argue that the two types of motivations are contrasting, they stress that they are interdependent, not mutually exclusive. Recent studies suggest that SMEs are not only internally motivated to implement CSR but also by the strategic/ economic benefits (Vázquez-Carrasco and L opez-Pérez, 2013; Nybakk and Panwar, 2015). Thus, an individual may be intrinsically-and extrinsically-motivated (Bénabou and Tirole, 2010;Ryan and Deci, 2000). As mentioned above, Graafland and Van de Ven (2006) also found in their Dutch study that managers were both motivated by intrinsic and extrinsic factors. They found that Intrinsic Motivation had a stronger relationship to CSR commitment than extrinsic motivation. Frey and Jegen (2001), however, found that emphasis on extrinsic rewards for individuals that are intrinsically motivated, may actually crowd out their intrinsic motivation. Deci, Koestner and Ryan (1999) also found in their meta-analysis of 128 studies, that extrinsic motivation undermined free-choice intrinsic motivation, and as such extrinsic motivation may weaken the relationship between intrinsic motivation and CSR activities. Based on these and earlier arguments, we hypothesise that: H2a. Intrinsic motivation and extrinsic motivation are correlated.
H2b. The association between intrinsic motivation and CSR activities is moderated by extrinsic motivation.

Small and medium-sized enterprises' internationalisation and motivations for corporate social responsibility
Firm internationalisation can be described as the process "through which a firm expands the sales of its goods or services across the borders of global regions and countries" (Hitt et al., 2007, p. 251). Solberg and Askeland (2006) state that the economic theories (Williamson, 1971(Williamson, , 1981Dunning, 1980Dunning, , 1988 appear too narrow to cope with the complexity of global competition. Accordingly, theories of networks and alliances (Snehota and Håkansson, 1995;Vahlne and Johanson, 2013) are most useful in explaining firms' internationalisation.
In the 1990s, the first signs of "internationalising" the CSR concept appeared according to Eteokleous et al. (2016). Examples were the first auditable certification standard (dealing with labour issues across countries) issued by social accountability international and initiating the development of the global reporting initiative. In a study of 50 companies in seven Asian countries, Chapple and Moon (2005) found a relationship between MNEs with international sales and/or foreign ownership and "level of CSR". Laudal (2011) believes that one reason for this may be that firms exposed to international competition will, in most cases, raise CSR standards. As cited in Eteokleous et al. (2016) socially responsible behaviour may mean different things in different places to different people and at different times (Campbell, 2007, p. 950), which creates difficulties in applying the concept when transcending national boundaries (Bondy and Starkey, 2014). Laudal (2011, p. 240) also states that if a firm is exposed to international conditions, it is likely that it is also exposed to different norms and ideas on workplace conditions and environmental protection.
The extant literature has been criticised for neither providing a clear guiding framework for implementing CSR in international markets (Zeriti et al., 2014) nor addressing crucial CSR issues within the overall international business field that could help improve theoretical SMEs motivations for CSR knowledge on the subject (Bondy and Starkey, 2014;Campbell et al., 2012;Eteokleous et al., 2016). Vázquez-Carrasco and L opez-Pérez (2013) argue that the focus should be more on social capital theory than on stakeholder theory to better understand SMEs' commitment and motivation regarding certain social aspects. The stakeholder's focus should be more implicit (Vázquez-Carrasco and L opez-Pérez, 2013). The way in which relationships with stakeholders are managed legitimises business activity, which, in turn, is closely tied to a series of moral and legal obligations with respect to the local community (Dunham et al., 2006 in Vázquez-Carrasco andL opez-Pérez, 2013). Attig et al.'s (2016) study of over 3,000 US firms in the period 1991-2010, found strong evidence that firm internationalisation is positively related to CSR rating. Although this study focused on large MNCs and not SMEs, we might assume that increased pressure from an extended set of stakeholders affects SMEs that are expanding internationally too, and may induce them to increase their CSR activities to "demonstrate their responsiveness to a wider range of stakeholders" (Brammer et al., 2009, p. 575). Scherer and Palazzo (2008) discuss the forces of globalisation on CSR, claiming that the complexity of global businesses urges every internationalised firm to pay critical attention to social issues. This implies that companies with a higher degree of internationalisation have a stronger relationship with CSR activities than lessinternationalised firms. Based on the theoretical review presented above, H3a and H3b are proposed: H3a. The association between intrinsic motivation and CSR activities is moderated by the level of internationalisation of the SME.
H3b. The association between extrinsic motivation and CSR activities is moderated by the level of internationalisation of the SME.

Research setting
The context of this study is the maritime cluster in the Møre and Romsdal (M&R) county, located on the west coast of Norway. The scope of CSR engagement in the majority of maritime companies in the cluster is uncertain, and scientific papers on CSR and shipping, in general, are limited (Skovgaard, 2014;Dabic et al., 2016). The M&R cluster has many actors with strong internal relationships both vertically and horizontally, characterised by low cultural and geographical distance (Jakobsen, 2008). The term cluster has different meanings in different sectors, ranging from geographical perspectives to socio-cultural factors and may even involve territorial dimensions (Pinto and Cruz, 2012). Three main definitive features of a cluster are identified: (1) an input-output model of an industrial complex with inter-industry links and transactions; (2) an agglomeration of interlinked industries, showcasing a nation's innovativeness and competitiveness; and (3) a community-based network with emphasis on the "regional innovative environment" focusing on firms' knowledge, policy support infrastructure and an institutional environment that stimulates innovation and entrepreneurial activity (Doloreux, 2017).
The vertical structure is because of the substantial number of customer-supplier relationships and the horizontal structure, the local competition in every part of the value chain. The cluster has links nationally through connections to a strong technological environment in Trondheim, and to national oil companies, and internationally: a substantial amount of total turnover, both among shipping, shipyards and suppliers comes from export (Jakobsen, 2008).

Measures, data collection and sample characteristics
The constructs and sources are shown in Table I. We used measures from multiple sources: some of the measures were newly formulated based on concepts discussed in the respective sources. For example, we formulated the legal measures based on insights from Maignan and Ferrell (2000), and the CSR dimensions are based on Carrol's 1991 CSR pyramid. All the constructs were measured on a seven-point Likert scale ranging from 1-strongly disagree to 7-strongly agree. The data source comprised 65 survey respondents from the M&R Norwegian maritime cluster (Table II) based on the key informant approach (John and Reve, 1982;Bryman and Bell, 2015). The list of relevant companies were extracted from a list provided by maritime forum in the north-west of Norway compared with a list used by Møre research institute who have used the list in previous research on the maritime industry. A total of 185 companies were targeted for the survey, with 65 completed responses received. SMEs are categorised as firms with fewer than 100 employees, in line with the definition by the Norwegian Government, and forms 81.5 per cent of the sample while large companies (over 100 employees) make up only 18.5 per cent. We used a dummy for turnover: firms with annual turnover below or above 169 mill. NOK are coded as 0 or 1, respectively. Export share is also a dummy variable of firms below or above 30 per cent export share with 0 or 1, respectively. International sourcing is coded as a dummy where firms below or above 30 per cent are coded as 0 or 1, respectively.

Common method variance
First, we used Harman's (1976) single-factor test. CMV is assumed to be present if a single factor emerges from the unrotated factor solution or one element explains most of the variance in the variables (Podsakoff and Organ, 1986). However, we found that a one-factor solution accounted for only 26 per cent of the overall variance. Additionally, we used the marker variable method (Kemery and Dunlap, 1986;Lindell and Whitney, 2001;Malhotra et al., 2006), a theoretically distinct variable, unrelated to at least one other variable in the study. We used several marker variables (e.g. age of key informant) to estimate the loadings on every item in the partial least square (PLS) path model and observed each item's loadings on its theoretical construct. We compared the estimated path model relationships with and without the marker. All theorised paths maintained their level of statistical significance. This shows that CMV bias is unlikely to influence the findings of the study.

Measurement model evaluation
The analysis is based on 24 items (indicators), using variance-based PLS technique and SmartPLS . Our choice of an analytical procedure is determined by the exploratory nature of the study and the small sample size (65). One other significant advantage is our ability to model the CSR latent construct as a reflective-reflective and formative multidimensional construct. PLS also makes fewer assumptions about the distribution of data (Chin and Newsted, 1999) and has the capacity to deal with complex models with a high number of constructs, indicators and relationships (Hair et al., 2016;Tompson et al., 1995). We evaluated the measurement model in terms of factor loadings, item reliability, convergent and Notes: #Based on 1,000 bootstrapping samples. CR = composite reliability; AVE = average variance extracted; M = mean; and SD = standard deviation; ***p < 0.001 and ** p < 0.01 (two-tailed) discriminant validity. All factor loadings were higher than 0.7 and significant (p < 0.001), except CSRER1 with a loading of 0.695 and EXMO4 (loading of 0.783, significant at p < 0.01). Loadings of at least 0.5 are considered acceptable (Hair et al., 2016) (Table I).
We assessed item reliability using composite reliability (Fornell and Larcker, 1981); all constructs exceeded the acceptable value of 0.7 (Hair et al., 2016), with the economic responsibility construct presenting the lowest (0.78) and extrinsic motivation the highest (0.93). We also assessed convergent validity using AVEs where all the constructs had values above 0.50, demonstrating convergent validity (Fornell and Larcker, 1981). Discriminant validity shows the extent to which a given construct is dissimilar from other latent constructs. We used the heterotrait-monotrait ratio of the correlations approach, where all the values were under 0.85, demonstrating discriminant validity (Henseler et al., 2015).
An inspection of loadings and cross-loadings also confirmed that all constructs were more strongly related to their own measures than to other constructs, signifying good convergent and discriminant validity. As we modelled CSR as a reflective-reflective and formative latent construct, we also assessed the formative factorial structure of the measurement model in terms of outer weights, t-values and collinearity. Regarding multicollinearity all variance inflation factor values were less than the more conservative rule of thumb value of 5 (Hair et al., 2016). Figure 1 shows the results of the path analysis estimated using a variance-based SmartPLS technique. We evaluated the structural relationships to confirm or disconfirm our hypotheses. Our first hypothesis states that intrinsic motivation (H1a) and extrinsic motivation (H1b) are positively associated with CSR activities. The association between intrinsic motivation and CSR was supported (b = 0.78, p < 0.001, two-tailed), but no support for H1b (b = À0.08, p > 0.05). We found "mixed" support for H1 with intrinsic motivation strongly related to CSR and extrinsic motivation showing a negative relationship with CSR though not a significant association.

Structural model assessment
To test our second hypothesis (H2a), we modelled "motivation" as a separate secondorder construct with structural links to intrinsic and extrinsic motivation. Intrinsic and extrinsic motivation were significantly related to "motivation" with loadings of 0.909 (t = 35.17) and 0.898 (t = 37.26), respectively. We inspected the latent variable correlation between intrinsic and extrinsic motivation and found these to be highly correlated (r = 0.633). Hence, we conclude that both intrinsic motivation and extrinsic motivation are interrelated. To test H2b, we relied on the main structural model result  Figure 1). The interaction between extrinsic and intrinsic motivation, as hypothesised in (H2b), was found to have a positive moderating effect on CSR activities (b = 0.20, p < 0.05, one-tailed). The moderating roles of the extent of internationalisation of the firm between intrinsic motivation and CSR (H3a) and between extrinsic motivation and CSR (H3b) were supported (H3a: b = 0.27, p < 0.05, one-tailed) and not supported (H3b: b = À0.09, p > 0.05), respectively. Figure 1 shows the results of the structural relationships and the second-order standardised parameter estimate loading between CSR and its four dimensions. Economic responsibility has the strongest measurement relationship with CSR (r = 0.84, R 2 = 0.71 and t = 18.56), followed by ethical responsibility (r = 0.72, R 2 =0.52 and t = 11.47). CSR explained the least variation in the sub-construct, philanthropic responsibility (r = 0.52, R 2 = 0.27 and t = 3.73). However, the variation in CSR itself, explained by intrinsic and extrinsic motivation, interaction effects and controls is 47 per cent. We also conducted a complementary analysis to test the relationships between intrinsic and extrinsic motivation and each of the four CSR dimensions. We found that intrinsic motivation was strongly related to CSR dimensions of economic (b = 0.61, t = 4.25 and p < 0.001), legal (b = 0.45, t = 2.93 and p < 0.01), and philanthropy (b = 0.56, t = 4.99 and p < 0.001) and less related to ethical (b = 0.12, t = 0.36 and p > 0.05). Extrinsic motivation was negatively related to legal (b = À0.39, t = 2.48 and p < 0.05) and had no significant association with economic (b = À0.04, t = 0.24 and p > 0.05), ethical (b = 0.05, t = 0.21 and p > 0.05) or philanthropy (b = 0.03, t = 0.16 and p > 0.05). We controlled for firm size, turnover and extent of internationalisation in each instant, regarding the effect of intrinsic and extrinsic motivation on the four dimensions of CSR. Regarding, explanatory powers, we found that intrinsic motivation, extrinsic motivation and the control variables explained 45 per cent variability in philanthropy, 33 per cent in economics, 13 per cent in legal and 9 per cent in ethical dimensions of CSR.

Discussion
In this study, we sought to find answers to the questions: do SMEs' intrinsic motivation drive CSR activities more than extrinsic motivation? Are motivations for CSR activities contingent on the extent of internationalisation of SMEs? In answering the first research question, we found support for the hypothesised relationship between intrinsic motivation and CSR and not for extrinsic motivation. Our findings suggest that SMEs' intrinsic motivation drives CSR more than extrinsic motivation. This is in line with Graafland and van de Ven (2006), who showed that intrinsic motives had a stronger association with a firm's involvement in CSR than extrinsic motives. The analysis also shows that intrinsic and extrinsic motivation are highly correlated, which is consistent with the literature (Graafland and van de Ven, 2006;Graafland et al., 2010). However, the interesting findings relating to the moderating role of extrinsic motivation can be interpreted as follows: at high levels of extrinsic motivation and intrinsic motivation enhances CSR activities. This means that the effect of intrinsic motivation on CSR is dependent on extrinsic motivation. The positive coefficient of the interaction effect shows that both forms of motivation enhance each other such that the association between intrinsic motivation and CSR is enhanced by extrinsic motivation or the relationship between extrinsic motivation and CSR is enhanced by intrinsic motivation. Both forms of motivation are not mutually exclusive. In as much as companies perceive CSR activities as the right thing to do, the will to do so is not in a void but depends on the business case for engaging in CSR (Fontana, 2017).
Our second research question asks whether motivations for CSR activities are contingent on the extent of internationalisation. The answer is inconclusive, as we had support for the moderating role of internationalisation between intrinsic motivation and CSR, but not for extrinsic motivation and CSR. This means the extent of internationalisation of the SME plays some role in motivations for CSR. Thus, for highly-internationalised SMEs, intrinsic motivation drives CSR more than less-internationalised firms. How can this be explained? Previous research (Laudal, 2011) shows that this may be because firms exposed to international competition will usually acquire higher CSR standards, while other researchers agree that companies engaged in global sales or that have subsidiaries in foreign markets exhibit some level of CSR (Chapple and Moon, 2005;Eteokleous et al., 2016;Jenkins, 2006;Laudal, 2011).
6. Implications 6.1 Contributions and implications for theory/research The mechanisms of CSR are not yet fully explained, which means that our knowledge of this phenomenon is still insufficient. One essential contribution of this paper is the modelling of the CSR concept as a multidimensional reflective-reflective and formative construct using variance-based techniques to achieve factorial validity. This can be used when conducting future studies on how to operationalise the CSR construct. In addition, the modelling of firms' motivations for CSR activities and the contingent role of internationalisation provides increased understanding of the mechanisms, which highly internationalising firms go through in their domestic environment as a result of experiences abroad. This is critical in helping to improve theoretical knowledge on the subject (Bondy and Starkey, 2014;Eteokleous et al., 2016). The study also provides evidence for the strong interrelationship between intrinsic and extrinsic motivation, and support for the positive effect of intrinsic motivation on CSR compared to extrinsic motivation. This paper provides further confirmation, establishing robustness for these findings in the extant literature.
The moderating relationships found in the present study also highlight the contingency relationships between motivation for CSR and firm-level characteristics such as the extent of internationalisation. Previous studies neither have examined this link without uncovering the underlying mechanisms (Laudal, 2011;Eteokleous et al., 2016) nor have they related this to the different types of motivation. The present study has conceptualised motivation into two broad categories (intrinsic and extrinsic) and has shown that the contingency effect of internationalisation relates more to CSR and intrinsic motivation. Additionally, for theorybuilding purposes, this paper has also shown that the effect of one dimension of motivation on CSR is dependent on the other. Thus, intrinsic and extrinsic motivation are highlyinterrelated, though each can be distinguished from the other.

Implications for policymakers/management
A key contribution of the paper is that the findings are accessible to policymakers and practitioners responsible for the formulation and implementation of policies for the maritime industry, at the company, national/sub-regional and international levels. Exploring motivation is essential for policymakers and societal groups that want to stimulate CSR. According to Graafland et al. (2010), if extrinsic motives drive CSR, policymakers should implement institutional reforms that increase financial incentives. However, if executives are motivated to conduct CSR-activities based on intrinsic motives, policymakers should be careful of providing financial incentives, because extrinsic motives may override intrinsic [Frey andJegen, 2001, cited in Graafland et al. (2010)]. An extensive content analysis of published EU documents from 2000 to 2011 regarding CSR indicated that the primary driver of EU policy was performance motivation, focusing on the business case for CSR (Skovgaard, 2014). However, with a new definition by the EU, ''the responsibility of enterprises for their impacts on society'', the focus is not only on the business case but also on the need for enterprises to integrate social, environment, ethical, human rights and consumer concerns into their business operations and core strategy in close collaboration with stakeholders (EU, 2019). The EU recognises that large multinational firms have influenced the CSR debate; that the term itself should be adjusted when addressing SMEs; and that personal values and ethics are essential so that support for CSR in SMEs should seek to build on and respond to these value-based motivations and not to replace or ignore them.
SMEs from the same sector often face similar social and environmental issues. Addressing these problems collectively can reduce the costs of action and result in improvements that an individual SME acting alone cannot achieve. Scholars also raise concerns that, for some SMEs, the pressure to engage in CSR may be counterproductive (Baden et al., 2009). They argue that CSR regulation may overshadow the voluntary aspect of CSR, thus reducing SMEs' innovativeness and creativity (Baden et al., 2009). The maritime industry is capital-intensive, but SMEs in M&R have been found to be highly innovative. Size is no hindrance to innovativeness, as it has been shown that [. . .] large firms [e.g. MNEs], in fact, are not more innovative than their smaller counterparts in every industry (Acs and Audretsch, 1987, p. 567). This is very important, especially for the maritime cluster in Norway, which is solely dependent on the innovative and entrepreneurial capability of actors within the cluster for its competitiveness and sustainability. One other necessary implication is that government policies should be implemented that help firms within the cluster (and the nation) to internationalise. This can be achieved by removing barriers and challenges that SMEs face in their efforts to become international.

Conclusion and suggestions for further studies
First, this paper used a cross-sectional correlational design within the context of a single industrial cluster in a developed country. Hence, the findings of this study may have limited generalisability to other settings. Further studies are needed using other contexts (e.g. developing countries, other developed countries and different clusters). Second, the sample size is small, hence, there is a need for larger samples to be used in future studies.
Third, implementation of CSR strategies can be challenging for SMEs (Lee et al., 2017) because of the limitations that size and lack of resources place on them. Hence, there is a need for more research on the mechanisms, which SMEs go through in formulating and implementing CSR: studies on CSR best practices would be useful to SMEs. Finally, longitudinal studies that follow-up on how SMEs initiate CSR, maintain and sometimes terminate or disinvest in CSR practices could also provide further insight into the changing dynamics of CSR practices. This is equally important as motivations for CSR could change over time depending on the industry, regulations and economic conditions in the external environment.

Article
Purpose of the study Major findings Walker et al. (2016) This study examine how increases in CSR and corporate social irresponsibility (CSiR) relate to firm performance (profitability, management efficiency and market valuation) CSR significantly relates to increased firm performance in all three measures, and that increased CSiR significantly relates to decreased profitability only. The results suggest that CSR dominates the relationship to firm performance, as it was positively related to all three measures of firm performance, and when CSR and CSiR exist simultaneously, CSR has a dominant positive effect Nybakk and Panwar (2015) This study focuses on understanding firms' instrumental motivations for engaging in socially responsible activities based on 230 micro firms The findings indicate that market orientation, learning orientation and risk-taking attitudes affect social responsibility towards different stakeholder groups in different ways. In some cases, the size and age of firms also affect these relationships Avram et al. (2018) This study reviewed 118 published papers/articles using systematic literature review methodology. The purpose is to uncover a growing conceptual and terminological fragmentation of the CSR literature concerning SMEs The review shows that research on the integration of CSR into strategies of SMEs is highly fragmented. 56 different terms are used in this academic literature as an alternative to CSR. The review identified 19 individual issues categorised into 4 overarching topics Balabanis et al. (1998) This study investigates the relationship between CSR and the economic performance of corporations The results supported the conclusion that (past, concurrent and subsequent) economic performance is related to both CSR performance and disclosure. However, the relationships were weak and lacked an overall consistency. For example, past economic performance was found to partly explain variations in firms' involvement in philanthropic activities. Environmental protection activities was found to be negatively correlated with subsequent financial performance Jensen et al.
The aim of this study of was to develop a framework that describes and explains how CSR initiatives are perceived by customers, and it links customer perception to the notion of customer value perception The results suggest that CSR initiatives, when communicated efficiently and considered as relevant by customers, will enhance two customer value categories: the extrinsic selforiented value defined as efficiency and excellence and the intrinsic other-oriented value pertaining to ethics or spirituality Fatma et al. (2016) This study aims to examine the interconnection between the consumer perceived CSR and its effect on the post-recovery satisfaction and loyalty after service failures The finding of the study provides the empirical evidence of the existence of a relationship between perceived CSR and customer postrecovery satisfaction and loyalty influenced by trust (continued) The purpose of this study is to analyse the literature on industry-specific CSR practices The findings indicate CSR studies are very unevenly distributed and that the issues studied and the methods used vary widely across industries. The authors also map this field of study and propose suggestions on where research on industry-specific CSR should go in the future Papasolomou-Doukakis et al.
The paper seeks to consider the findings that emerged from a preliminary study into Cypriot businesses and their attitudes and behaviour towards CSR Local corporations emphasise the importance of meeting their responsibilities and obligations towards their employees and customers. The majority of respondents postulate that their philanthropic donations are made to approved organisations for tax deduction. So, financial gain appears to be a key motive for the adoption of social responsibility by the business sector in Cyprus Fontana (2017) This study investigates cognitive antecedents and behavioural consequences of corporate executives towards investing in strategic CSR showing that profit-driven CSR diffuses at the expense of altruism The study suggest altruism and performance as being cognitively and theoretically espoused in strategic CSR; yet, one appears to oust the other.
The study highlights CSR's strategic role as necessary but not sufficient for competitive advantage, delivering insights on suppliers' future posture vis-à-vis CSR in the Bangladeshi RMG supply chains Svensson et al. (2010) This study considers contemporary business practice and its sustainable performance from the view of stakeholders and their perceived value The study argues that a company has responsibilities and commitments to many different internal and external stakeholders in the marketplace and society. The authors presented a model based on five, separate but interconnected elements. The model is iterative and acknowledges its elementary state, suggesting further development and refinement in the field of sustainable business practices from an ethical perspective Khojastehpour and Johns (2014) This study investigates the effect of environmental CSR (climate responsibility and natural resource utilisation) on corporate/brand reputation and corporate profitability The study highlights that environmental CSR has a positive effect on corporate/brand reputation and corporate profitability Vázquez-Carrasco and L opez-Pérez (2013) Through a systematic review of the literature, the authors analysed the state of the art literature and identified the theoretical framework of reference, and CSR antecedents and consequences in SME contexts The authors highlight the uniqueness of SMEs in relation to CSR. The review findings shed light on key themes including idiosyncrasies of CSR management, aspects such as terminology and language, the recommended theoretical framework, and antecedents, barriers and potential impact of CSR in the SME context Ralston (2010) The study examine and reviewed the body of cross-disciplinary literature on CSR, organisational structure and organisational culture The author suggests that the issue of social responsibility is a phenomenon both external and internal to organisations, as it lies on the cusp of organisational culture and social expectations (continued)  Eriksson and Svensson (2016) This study assess elements that affect social responsibility in supply chains and beyond. The elements were classified into drivers, facilitators and inhibitors In total, 16 elements were identified and presented in a framework along with their proposed constituents. The elements capture structures and management principles of supply chains that are important for social responsibility. The elements provide a basis to better understand how social responsibility in supply chains is related to contextual factors Eriksson (2016) The study explains the role of moral disengagement in supply chain management (SCM) research and the challenges that arise if the theory is used beyond its inherent limitations The paper suggest that moral disengagement (an important element of CSR) can be used validly in SCM research. The paper suggests a new theory for a better understanding of business ethics, CSR and sustainability in SCM. Furthermore, the paper outlines how the theory should be used and some challenges that remain Bondy and Starkey (2014) This study investigates the extent to which foreign national culture and related local issues are incorporated into the CSR policy of 37 multinational corporations, examining strategy development and implementation across global locations This study suggests that integrated internationalisation strategies do not resolve global and local CSR issues. In fact, they reinforce outcomes similar to global strategies, where core issues identified by headquarters are legitimated and local issues are marginalised, an outcome that appears somewhat at odds with the spirit of local responsiveness embedded in CSR thinking Note: RMG = Ready-made garment