The Mediating Effect of Supplier Relationship Management on CSR and Marketing Performance Relationship

The study sought to examine the mediating outcome of supplier relationship management (SRM) on the CSR and marketing performance (MP) relationship. The study was descriptive, quantitative in nature with a questionnaire as its main data collection instrument. A purposive sampling approach was employed to select 317 owners/managers of the manufacturing firms. Data collected was analyzed using Structural Equation Model–Partial Least Square statistical software. The study proposed two models; model 1 without mediation effect and model 2 with mediation effect. The results revealed a direct significant positive relationship between CSR and MP and CSR and SRM; including a direct positive relationship between SRM and MP. However, the direct link between CSR and MP was not supported, when the mediating variable was included; but shows that SRM fully mediates the CSR and MP tie. The effect of SRM explains the effect of CSR on MP. Appearing to be the first study to examine the mediation role of supplier relationship management, this study discovered the novel path through which CSR is linked to a company’s marketing performance, as well as key CSR engagements that can enhance supplier relationship management and marketing performance.


INTRODUCTION
Corporate Social Responsibility (CSR, hereafter), though not a new phenomenon, has come to be a mainstream business factor and growing currency in today's world of business. This is due to the ever growing interest of the general public about the care for the society as well as the new drive for firms to focus on every aspect of their activities including relationship with stakeholders (Kitora & Okuda, 2008). Studies have shown and persist to authenticate the power of collaborative partnerships (Vitezić et al., 2012). This is particularly true when highly of this study as a foothold to enhance the integration of CSR and exploit the synergies between CSR and supplier relationship management. Hence, the outcome of the study can help owners and managers of firms understand how they can enhance their marketing performance and relationship management with suppliers through CSR engagements. It is also envisaged that the result of the study would expose policymakers, educators, and stakeholders to the novel trends in the field for policy direction and development. Thus, the originality and value of our paper is a recognition that it makes this under-researched literature on the mediating effect of supplier relationship management on CSR and marketing performance more accessible for researchers, practitioners, and policymakers.

METHODS
The present study was a descriptive, quantitative in nature with the purpose of developing a comprehensively appreciation of CSR, supplier relationship management, and marketing performance link. A quantitative method was adopted to gather more precise and quantifiable information about the mediating effect of supplier relationship management on CSR and marketing performance. Again a quantitative study was employed as it is more cost-effective, easier to administer and enable the researchers to test statistical relationship among variables of the study. According to Ghana National Commission for UNESCO 2016 report, there are about 25,000 registered manufacturing firms in the country. Of this, the report reveals that 55% are located in Accra/Tema metropolis. Thus, the total population of the study area was 13,750 firms.
However, as Malhotra & Indrayan (2010) noted, it is not possible to cover the whole population in a single survey. Thus, employing a purposive sampling method, 337 owners/managers were selected for the study. Levy & Lemeshow (2008) intimated that this sampling method select respondents based on their familiarity, interaction, understanding, and proficiency of the observable fact regarding a research subject. It was therefore necessary to purposively select owners/managers who engage in CSR endeavours and have experience regarding the impact that it is having on their relationship with suppliers and marketing performance. Out of the 337 sample size, a response rate of 317 (94.06%) was recorded for the study. Thus, 20 questionnaire items were not returned.
A total of 337 questionnaires were self-administered and distributed to top managers or owners of manufacturing firms in the Accra metropolis of Ghana. The questionnaire was administered in English. The data was collected from April-June 2017. The questionnaire items employed a 5-point Likert-type scale rating as highest in agreement 5 (fully agree) to lowest in agreement 1 (fully disagree). The owners/managers were required to specify their level of agreement with the statements provided in the questionnaire. The questionnaire was in four sections: A-D. Section A was designed to solicit for the demographic information of the respondents; B measured the CSR endeavours by the firms; C measured firms' supplier relationship management practices, and D measured the marketing performance of the firms.
The variables used to measure the CSR efforts of the manufacturing firms consisted of seven items and were adapted and modified from prior research studies (Carroll, 1991;Alrubaiee et al., 2017). These included economic (being profitable), legal (obeying rules and regulations), ethical (being fair and just), and philanthropic duties (being good citizen by altruistic business giving). The premier study that employed these constructs confirmed the construct's content validity (Aupperle et al., 1985). The factor analysis carried out in their study concluded that these four variables are empirically interrelated but conceptually independent constructs of CSR. Later studies also confirmed that measuring CSR through the use of Carroll's four dimensions were valid and useful (Pinkston & Carroll, 1996;Edmondson & Carroll, 1999). Thus, the theoretical usefulness and distinctiveness of these four constructs have been confirmed through a number of empirical studies. The other three measurements of CSR that have been tested empirically by Mahadevappa (2010) and subsequently by Alrubaiee et al. (2017), which were adapted and modified in the present study, included: 1) the extent to which the firms reports on social service, 2) the extent to which the firms consider the financial characteristics of citizens (customers), and finally, 3) the extent to which the firms consider the social characteristics of citizens (customers) in their programmes and marketing campaigns. Thus, these constructs are envisaged to capture comprehensively the CSR disclosure of the manufacturing firms in the study area of the present study.
Five supplier relationship management-related dimensions, recognised as potentially antecedents of marketing performance, were measured through the use of the existing scales, namely trust, commitment, collaboration, supplier partnership, and information sharing (van Echtelt et al., 2008;Kim, 2011;Al-Abdallah et al., 2014). These constructs have been theoretically, conceptually, and empirically identified as essential drivers of relationship quality, which positively enhances firms' marketing and financial performance (Fynes et al., 2005;N'Goala, 2007;van Echtelt et al., 2008). We measured trust using three items that explore the extent to which the manufacturing firms believe the reliability and credibility of their suppliers (Dyer & Chu, 2000). The measurement of the collaboration comprises five items that assess the extent to which the manufacturing firms are able to collaborate jointly with suppliers to improve suppliers' efficiency in terms of cost, delivery etc. (Su et al., 2008). We assessed information sharing using four items that explore whether there is a frequent flow of information between the manufacturing firms and their channel members for enhanced supplier relationship management (Cooper & Gardner, 1993). We measured commitment using three items, which assess the manufacturing firms' dedication in their relationship with suppliers (Langfield-Smith & Greenwood, 1998;van Zyl, 2003). Finally, we measured the partnership construct using four items based on the assessment of the manufacturing firms' efforts to expend on their suppliers to increase the capabilities and performance of the suppliers to meet their short and long-term supply needs (Krause & Ellram, 1997;Krause et al., 2000).
The marketing performance also comprised 6 items that were adapted and modified from Kotler & Keller (2016); Alrubaiee (2013); Alrubaiee et al. (2017). These included customer service, competitive position, customer loyalty, customer satisfaction, and sales, profitability. These measurement consist of 13 items in total (customer service = 2; competitive position = 2; customer satisfaction = 3; customer loyalty = 2; sales growth = 2; profitability = 2) that were assessed through the manufacturing firms' subjective assessment of their companies' marketing campaigns effectiveness and efficiency with regard to whether or not it has accomplished its predetermined, market-based goals (Pertusa-Ortega et al., 2009;Alrubaiee, 2013). This method of measuring firm marketing performance was also to avoid the problem of the reluctance of owners/managers to disclose objective financial information about their organisation. In all, there were 28 questionnaire items.
In order to test the hypotheses formulated for the study, Structural Equation Model-Partial Least Square (SEM-PLS) was engaged to examine the significance level and weight of each path in the structural model. Concerning the testing of the mediation effect, though there are many methods such as Baron & Kenny (1986) mediation analysis (Sobel, 1982) and Bootstrap method (Preacher & Hayes, 2008), this current study adopted the bootstrapping approach and analysed through the use of Variance Accounted For (VAF), a view shared by (Hair Jr et al., 2014). According to Hadi et al. (2016), the bootstrapping approach has an advantage over Sobel test as the former helps to determine mediation effect with certainty. The outcome of the analysis is discussed in the succeeding paragraph.

RESULTS AND DISCUSSION
Of the 337 questionnaires sampled for the study, a response rate of 317 (94.06%) was recorded. The respondents' demographic information considered in the study was gender, age, educational background and experience in the position. The outcome of the result is made known in Table 1. As clearly divulged in Table 1, regarding the gender of the respondents, 212 (66.88%) were males while 105 (33.12%) were females. This shows that the majority of the owners/managers of the manufacturing firms in the study area are males, highlighting that the majority of the opinions revealed in the present study came from men. Again, it is shown that 27 (8.52%) were between 25-30 years, 53 (16.72) were between 31-35 years, 64 (20.19%) were between 36-40 years, and 111 (35.02%), 41 (12.93%), 21 (6.62%) were between 41-45 years, 46-50 years, and above 50 years correspondingly. The majority of the owners/managers are between the ages of 41-45 years. Further, regarding the educational background, the majority of the respondents were postgraduates (Master degree holders), constituting 171 (53.94%). This was followed by first-degree holders 103 (32.49%), then Ph.D. holders 26 (8.20%), and Diploma/HND holders scoring least 17 (5.36%). This, by inference, shows that the respondents have acquired the needed educational qualification required to understand the phenomenon under study.
The practical experience of a manager or owner is very important to understanding a phenomenon. Thus, the study sought for the number of years the owners/managers have assumed their own positions. Again, Table  1 shows that the majority of the respondents 123 (38.88%) have held their positions for 7-10 years, followed by 6 Nyarku and Oduro Indonesian Journal of Sustainability Accounting and Management, 2019, 3(1), 1-13 those with experience for 11-15 years 87 (27.44%). Those who have been in position for 15 years and above, 3-6 years, and below 2 years were 41 (12.93%), 37 (11.67%), and 29 (9.15%) respectively. This implies that the respondents were experienced enough to express their opinions on the phenomenon in question.
As Preacher & Hayes (2008) intimated, any study must carry out the bootstrapping exercise twice, if it aims to test for the presence of mediation effect. The initial test should not consider the presence of mediation, but the second test should consider the presence of the meditation. As hinted by Hair Jr et al. (2014), there is no mediating effect if the direct path is not significant. To test for mediation effect, there must be a significant independent and dependent variable relationship. If significant, then the mediating variable is introduced. When the introduction of the mediating variable annuls the direct effect, then there is a full mediation or otherwise, mediation is partial or absent. The result of the reflective outer model is presented in Table 2.  Table 2 shows the reflective measurement model employed to analyse the reliability and validity of our model. According to Carmines & Zeller (1979), when all the standardised cross loadings are greater than 0.7, then the model's reliability is sufficient. As clearly observed in Table 2, all our individual items meet the reliability criterion. Similarly, our model meets the requirement for construct reliability, as the composite reliabilities have greater than 0.7 values (Nunnally & Bernstein, 1994). Fornell & Larcker (1981) criterion was employed to check for convergent validity, as according to them, when all the latent variables Average Variance Extracted (AVE) are greater than 0.5, convergent validity is achieved. Again, our model satisfies this criterion. Table 3 shows the result from the discriminant validity. To confirm the discriminant validity of a model, Sobel (1982) postulated that the diagonal elements should be greater than the off-diagonal elements in the respective rows and columns. Again, our variables meet this condition as the diagonal elements (bold) have values exceeding the corresponding off-diagonal elements values.  Table 4 shows the result of the hypotheses test and path coefficients. As clearly displayed, Table 4 includes the direct path for both models 1 and 2. However, only model 1 consists of the CSR-MKP direct relationship. Thus, in the direct model (model 1), CSR explained 34.3% % of the variance in marketing performance. This is shown by the R 2 or coefficient of determination, which measures a model's overall goodness of fit. Scholars have indicated that R 2 of 0.75 is substantial, 0.5 is moderate, and 0.25 is weak. This way, it can be argued that CSR moderately explains the variance in marketing performance, denoting that there are other variables that may account for variance in marketing performance. But in model 2, it is seen that when the mediating variable, supplier relationship management (SRM) was included, the variance explained in marketing performance increased to 42.3%. Thus, CSR and SRM moderately explain the variance in marketing performance. But is seen that CSR sufficiently explain the variance (76.9%) in SRM, denoting how important CSR endeavours is for building relationship with suppliers. From Table 4, we find support for H1, that describes the direct CSR and marketing performance relationship (t-value = 15.524; p-value = 4.73828E-40). Moreover, model 2 comprises the other direct relationship when SRM was included within the model. Results reveal that CSR no longer shows significant relationship with marketing performance (t-value = 0.7119; p-value = 0.477050281). However, we find support for H2a, which describes the direct link between CSR and SRM (tvalue = 80.4863; p-value = 4.751E-213). As well, our results support H2b that SRM relates positively to marketing performance (t-value = 6.4172; p-value = 5.05361E-10). Table 5 displays the strength and significance of the mediating effect. According to the bootstrapping approach of testing mediation, we have to evaluate the direct effect first, which in this case include CSR on marketing performance, CSR on supplier relationship management, and supplier relationship management (SRM) on marketing performance (MKP). There is no mediation if the direct effect is insignificant; however, we include the mediating effects, if the direct effect is significant. Our model showed significance in all the direct effects. If the indirect effect after bootstrapping is insignificant, then there is no mediation . Hair Jr et al. (2014) highlighted that if the indirect effect is significant, we can calculate the variance accounted for (VAF). Since our model met the above condition, we proceeded to calculate the VAF and its significance as shown in Table 5. Through this, the variance of the dependent variable that is indirectly 8 Nyarku and Oduro Indonesian Journal of Sustainability Accounting and Management, 2019, 3(1), 1-13 explained by the mediating factors was determined (Chin, 2010). VAF = (b1*c1)/b1*c1 +a), that is, indirect effect divided by total effect (indirect effect/total effect * 100). From model 2 (b1 = CSR-SRM = 0.877; c1 = SRM-MKP = 0.588; a = CSR-MKP = 0.070). The computations achieve a VAF value of 88.05%. According to Hair Jr et al. (2014), a value less than 20% shows no mediation, a value between 20% and 80% demonstrate partial mediation, and a VAF greater than 80% is a full mediation. Thus, SRM fully mediates the CSR-MKP tie. The discussions of these findings in relations prior related are shown in succeeding section. The literature highlights the impacts of CSR on company performance and the role of relationship management with key stakeholders. Despite the fact that earlier studies have highlighted the significance of CSR in managing the relationship with stakeholders such as customers, investors, employees, it appears that none considered the mediating role of supplier relationship management. Thus, this study contributes to enriching the current research on corporation's strategic endeavours on CSR and relationship management, in their quest to improve marketing performance.
First and foremost, we find support for H1 that CSR positively relates to marketing performance. It was discovered that CSR activities enhance marketing performance specifically in terms of sales, customer value/satisfaction, competitive positioning, and trade/channel relationship. The results divulged that the manufacturing firms who engage in CSR activities are enjoying sales increase, better customer relationship, better positioning in relation to competing brands and better trade or channel relationship with wholesalers, retailers, distributors. It goes without saying that this finding is in line with prior related studies (Slater & Narver, 2000;Barone et al., 2000;Khalifa, 2004), that hinted a positive association between CSR and customer value. Customers are increasingly passionate to buy from firms that engage in CSR initiatives. Nyaga (2012) also noted in his study that customers obtained prestige/value from socially responsible brands and that CSR efforts can influence customer satisfaction (Chung et al., 2015). Again, the finding corroborates the results of Vitezić et al. (2012) that argue that being socially responsible can maximise the sales and market share of firms, and a vital antecedent of marketing performance (Alrubaiee et al., 2017).
Secondly, our results support H2a, which describes CSR as an antecedent of supplier relationship management. This result supports previous findings (Orlitzky et al., 2003;Kim, 2011) that stressed that corporations CSR activities can enhance their relationship with key suppliers in terms of commitment, trust, partnership, information sharing. In line with the above sentiments, Spicer (1978) also disclosed that firms' relationship with key stakeholders such as suppliers would improve when they engage in social performance endeavours. These findings show that CSR is one of the cornerstones for building a relationship with suppliers. Similarly, we find support for H2b that supplier relationship management positively influences marketing performance, denoting that supplier relationship management is an antecedent of marketing performance. This is in line with the prior related studies Wisner & Tan (2000); Williams (2006); Al-Abdallah et al. (2014) that argue that organisations, in order to enhance their performance in terms higher sales, market share, profits, better customer service, must maintain a good and lasting relationship with suppliers. This finding indicates that through effective supplier relationship management, firms can increase their sales, market share, customer value since they will receive timely delivery of raw materials that are of best quality.
Finally, H3, which describes the mediating effect of supplier relationship management on CSR and marketing performance, is also supported. The analysis from the bootstrapping exercise disclosed that supplier relationship management fully mediates CSR and marketing performance relationship, which is supported at a p-value of 3.99679E-08. The findings have shown that marketing performance is primarily driven by supplier relationship management and supplier relationship management is driven by CSR. This finding is very important for firms involved in CSR endeavours. The findings show that firms can use CSR activities to develop and harness their relationship with their suppliers, which in turn, may improve their marketing performance in terms sales increase, better customer service, better channel/trade relationship, and customer value.
This finding supports previous studies Spicer (1978); Orlitzky et al. (2003) that discovered that organisations' relationship with suppliers, improves if they undertake CSR activities, which may allow them to secure timely and quality deliveries, and better contractual terms. Again, the finding supports the results of Kim (2011) that CSR efforts can ameliorate customers' attitudes, increase purchase intention, in a B2B relationship. Thus, as the stakeholder theory hints, socially responsible corporations can secure stronger support from stakeholders (Maignan & Ferrell, 2004;Mukherjee & He, 2008). Thereby, the findings show that for CSR to have the desired impact on marketing performance, firms need effective supplier relationship management. In fact, it has been argued that strong relationship with suppliers is associated with the competitiveness of firms (Al-Abdallah et al., 2014). Thus, the findings have disclosed CSR as one of the elements that the manufacturing can use to enhance their relationship with suppliers, and thereby achieve their marketing objectives.

CONCLUSION
The study sought to investigate the effect of CSR on marketing performance, by particularly examining the mediating role of supplier relationship management, an area which has received little attention by researchers. Based on the discussions and findings revealed above, below conclusions, managerial implications, and recommendations for managers and future researchers are drawn. The outcomes of the study have shown that CSR has a momentous direct effect on marketing performance in absence of mediating variables. It was further revealed that CSR significantly influences supplier relationship management, and supplier relationship management significantly influences marketing performance. However, the mediation analysis has shown that CSR impact on marketing performance is fully mediated by supplier relationship management. Thus, it is concluded that the effect of CSR on marketing performance is as result of the effect of supplier relationship management, because when this mediating variable was introduced, the direct CSR and Marketing performance link was no more supported.
The implications for managers of manufacturing firms in Ghana and beyond are relevant. To achieve marketing objectives, supplier relationship management should be considered as an indispensable piece. The firms should understand supplier relationship practices such as trust-based relationship; supplier leads time cuts, supplier quality enhancement, supplier collaboration, and develop programmes required to achieve them. However, as the findings have shown, there is a significant CSR and supplier relationship management