Corporate social responsibility and firm performance of Ghanaian SMEs: The role of stakeholder engagement

This paper aims to examine the influence of corporate social responsibility (CSR) on financial performance of smalland medium-sized enterprises (SMEs) in Ghana by using stakeholder engagement as a mediating variable. Primary data were collected from 423 SMEs within the Accra Metropolis. Partial least squares estimation technique was used to analyze the data. The study documented evidence for a mechanism through which CSR results in financial performance of firms: SMEs with improved CSR practices are better positioned to engage more with their stakeholders, which translates into improved financial performance. It was recommended that for SMEs to improve upon their CSR practices, which will eventually result in enhanced financial performance, stakeholder engagement should be a major part of their operations. The paper contributes to our knowledge on how CSR practices lead to financial performance of SMEs in developing countries. In addition, this is the first of its kind to establish the relationship between CSR practices and financial performance of SMEs in Ghana by using stakeholder engagement as a mediating factor. Subjects: Finance; Business, Management and Accounting; Industry & Industrial Studies


PUBLIC INTEREST STATEMENT
This study examined how business organizations' investments in social and philanthropic activities influence their financial performance. Specifically, it investigated the notion that companies can only benefit financially from such investments when they actively involve parties that can affect and/or are affected by the firms' operations. The researcher collected data from 423 SME owner/ managers in Ghana. The findings indicate that engaging groups and persons that can affect and are affected by a firm's operations can translate social and philanthropic commitments to strong financial returns for a business entity. Hence, it was recommended that SMEs should solicit the views of their stakeholders in the area of social and philanthropic practices to enhance their financial performance.

Introduction
Along with the greater attention paid on the important role that foreign direct investments play in developing countries is also the concern that these companies exploit the social and environmental conditions of such countries. Corporate social responsibility (CSR) has generally been employed by companies as a vehicle for offsetting such perceived corporate abuses. In Ghana, CSR has become a front-burner issue due to several developments such as globalization; the liberalization of the economy; a government commitment to making the private sector the "engine" of growth; a promising and an emerging capital market with remarkable performance. Business entities are given the space to survive and thrive and thus have become critical actors in the economic, social, and cultural development of the country. Besides providing goods and services, they are a source of livelihood for many, pay taxes effectively enabling governments to operate, and have an impact on the physical and social environment (Atuguba & Dowuona-Hammond, 2006).
The European Commission's Green Paper of July 2001 defined CSR as "a concept whereby companies integrate social and environmental concerns in their business operations and in their interaction with their stakeholders on a voluntary basis." However, the idea that CSR should be voluntary is currently being challenged as they are many inter sections between CSR and the law, including actual new legislation in some countries (Blowfield & Frynas, 2005). Although, there is no comprehensive or readily available document on CSR in Ghana, nonetheless, there are a variety of policies, laws, practices, and initiatives that together provide the CSR framework in Ghana. In other words, CSR in Ghana is regulated by policies, legislation, and other forms of law. There are many government policies, such as the Ghana Land Policy document that bear directly on CSR (Atuguba & Dowuona-Hammond, 2006).
While society is mostly worried about the devastating effects of business activities, these business organizations are also concerned about the extent to which ethical and responsible treatment of all stakeholders, as advocated by stakeholder theory, impact on firm performance (Freeman, 1984;Hopkins, 2003). In spite of the obvious cost implications that confront companies that attempt to balance the economic interest of stockholders with the ethical, social, and environmental concerns of other stakeholders, a large number of today's leading corporate managers have invested heavily in CSR practices (McPeak & Tooley, 2008). These corporate executives reasoned that the benefits from CSR initiatives such as improved employee morale, customer loyalty, and other social capital far outweighs the costs associated with socially responsible actions (Moskowitz, 1972;Parket & Eibert, 1975;Soloman & Hansen, 1985). However, some scholars still hold the position that firms that invest in CSR activities are likely to be economically disadvantaged relative to less responsible ones because of the perception that such investments do not yield any economic benefits (Aupperle, Carroll, & Hatfield, 1985;Ullmann, 1985;Vance, 1975).
Empirically, several studies have examined the relationship between CSR and corporate financial performance and their results are rather inconclusive (See Walsh, 2003 andOrlitzky, Schmidt, &Rynes, 2003 for an extensive review on this subject). McWilliams and Siegel (2000) contend that these conflicting results could be due to a number of empirical and theoretical drawbacks associated with the works. Ullman (1985) advanced that the disregard for "contingency elements" provides an apt explanation for the inconsistent findings on this relationship. Similarly, Carroll and Shabana (2010, p. 94) argue that "the broad view of the business case for CSR suggests that the relationship between financial performance is better depicted when the role of mediating variables and situational contingencies are accounted for." It has been acknowledged that one of the factors that could influence the kind of impact CSR could have on firm financial performance is stakeholder engagement (Barnett, 2007). In CSR discourse, the success of CSR initiatives is often connected to stakeholder engagement because such engagements assemble representatives of business, non-governmental, regulatory bodies, and other public sectors in order to identify and address aspects of CSR most relevant to society and less inimical to business survival. Such engagements also assist firms gain legitimacy among business and development practitioners and society as a whole (Blowfield & Frynas, 2005). The management of firms' relationship with its key stakeholders can enhance firm performance since the creation and nurturing of such relationships can help firms access external resources (Kapstein, 2001), and establish good reputation (Agyemang & Ansong, 2017;Ansong & Agyemang, 2016;Fombrun, 2005;Fombrun & Shanley, 1990;Freeman, Harrison, & Wicks, 2007). Waddock and Graves (1997) contend that CSR initiatives that do not involve stakeholders are less likely to have any reward for firms. The authors seem to suggest that firms that involve stakeholders such as community, customers, and strategic partners (such as capital providers) in their CSR efforts are those that are likely to improve their financial performance. In contrast, in firms with lower involvement of stakeholders in CSR initiatives, stakeholders are more likely to be indifferent toward firms' operations. This will, therefore, decrease their financial performance.
Thus, the thrust of this paper is to examine the relationship between CSR and corporate financial performance by focusing on stakeholder engagement as the mechanism through which CSR activities translate into improved financial performance. Specifically, it is argued that the benefits accruing from CSR efforts are only possible when stakeholders are actively engaged in such activities in the context of a developing country such as Ghana. While there are far more empirical studies on CSR in developed countries relative to developing ones, the need of CSR is more pronounced in the latter since they are characterized by weaker governance structures to ensure adequate provision of social goods. Under these circumstances, companies are expected to pay more attention to their social responsibilities to fill these gaps (Baughn & McIntosh, 2007).
The remainder of the paper is structured as follows. The next section presents the paper's literature review. Ensuing is the methodology of the paper. Finally, the analyses, discussion, recommendations, suggestions for future research, and conclusion are presented.

Literature review
The CSR theories, concepts, and ideas are often considered as western phenomenon due to the fact that these countries in this region have relatively strong institutional environments in which regulation is efficient and fairly enforced. Yet, the tenets of CSR are more needed in emerging economies and developing countries characterized by weak institutional environments underlined by arbitrary enforcement of law, bureaucratic inconsistency, insecurity of property rights, and corruption (Jamali & Mirshak, 2007;Kuznetsov, Kuznetsova, & Warren, 2009). The absence of strong standards poses considerable challenges for firms that desire to practice CSR in developing countries in spite of the prevailing notion that business entities must play a wider role in society bedsides providing quality goods and services at reasonable prices and making returns for their shareholders (Mishra & Suar, 2010).
At the genesis of the discussions on CSR during the 1950s, little or no attention was drawn to the possible benefits that could accrue to businesses that engage in CSR initiatives. The impetus of the discourse centered on businesses' responsibilities to society and the need to pursue philanthropic activities for society. Theodore Levitt concluded these earlier narrations of the 1950s by cautioning the business world about the threat CSR initiatives pose to its very existence (Levitt, 1958). In spite of Levitt's warnings, the concept developed tremendously in the 1960s through the efforts of academics and researchers that articulated the business case for CSR (Shabana & Carroll, 2010). For instance, the proponents of instrumental theories of CSR have advanced that the sustenance of CSR practice is contingent upon the extent to which the interaction between business and society leads to some economic benefits to business organizations because wealth creation is their sole social responsibility (Garriga & Melé, 2004). Arguably, one of the medium through which CSR could have a significant influence on shareholders' wealth creation objective is through the engagement of relevant stakeholders. Jones (1995) posited that certain types of CSR activities are manifestations of attempts to establish trusting, cooperative firm-stakeholder relationships and should be positively linked to a company's financial performance. Jamali and Mishark (2007) have advanced that, in the context of developing countries, economic factors deserve attention in the pursuit of CSR. They opined that while many companies in developing countries are still grappling with the bottom line implications of their CSR performance, few of these companies in practice engage their stakeholders.
The study discovered that, even though, most Lebanese companies emphasized the relevance of stakeholder engagement in the planning stages of CSR interventions, they rather prefer to practice the so called "silent CSR" which tend not to have substantial influence on firm performance.

The link between CSR and financial performance
Based on neoclassical economics, some researchers have argued that CSR unnecessarily raises a firm's costs in areas that do not yield any direct economic benefits thereby placing the organization at a competitive disadvantage (Aupperle et al., 1985;Friedman, 1970;Jensen, 2002;McWilliams & Siegel, 2000). A classic example is when a firm decides to invest in pollution-control technology when other competitors do not consider such investment as necessary. Agency theory also suggests that CSR initiatives could serve as avenues through which managers can "legally" expend valuable firm resources at the expense of firm's shareholders wealth maximization objective. Hence, some have concluded that CSR only results in colossal managerial benefits rather than yielding any financial gains to corporate shareholders (Brammer & Millington, 2008).
Some scholars have also attributed the negative effects of CSR on financial performance in the context of emerging economies to poor stakeholder engagement. It has been established that the ways firms relate with their stakeholders and communicate CSR activities to stakeholders can affect their financial performance (Rettab, Brik, & Mellahi, 2009). For instance, Lima Crisóstomo, de Souza Freire, and Cortes de Vasconcellos (2011) studied the relationship between CSR and corporate financial performance in Brazil using financial and CSR data of 78 non-financial listed companies in the period 2001-2006 and concluded that in Brazil, CSR did not enhance the value of firms because managers failed to involve stakeholders. They posit that many external stakeholders were not aware of firms' CSR initiatives in Brazil to assist these firms translate such investments to better financial performance. On the whole, business entities in emerging economies do not appreciate the importance of communicating their CSR activities to their stakeholders (Min Foo, 2007;Wright, Filatotchev, Buck, & Bishop, 2003). Mellahi and Wood (2003) argued that because stakeholders are not normally aware of CSR activities in developing economies, such initiatives are likely to have a negative impact on financial performance. The next perspective, closely connected to stakeholder engagement analysis, advanced that CSR activities can have a positive effect on firm's performance by providing better access to valuable resources (Cochran & Wood, 1984), attracting and retaining higher quality employees (Greening & Turban, 2000;Turban & Greening, 1997), allowing for better marketing of products and services (Fomburn, 1996;Moskowitz, 1972), creating unforeseen opportunities (Fombrun, Gardberg, & Barnett, 2000), and contributing toward gaining social legitimacy (Hawn, Chatterji, & Mitchell, 2011). Furthermore, CSR may function in similar ways as advertising does, increasing demand for products and services and/or reducing consumer price sensitivity (Dorfman & Steiner, 1954;Milgrom & Roberts, 1986;Navarro, 1988;Sen & Bhattacharya, 2001) and even enabling firms to develop intangible assets such as goodwill (Gardberg & Fombrun, 2006;Hull & Rothenberg, 2008). The stakeholder theory (Freeman, 1984;Freeman et al., 2007;Freeman, Harrison, Wicks, Parmer & de Colle, 2010) expounds that CSR includes managing multiple stakeholder ties concurrently which can mitigate the likelihood of negative regulatory, legislative, or fiscal action (Berman, Wicks, Kotha, & Jones, 1999;Freeman, 1984;Hillman & Keim, 2001). Carroll and Shabana (2010) summed up the reasons why CSR could have a positive relationship with financial performance to include; reduction in cost and risk, enhanced competitive advantage; developing reputation and legitimacy and seeking win-win outcomes through synergistic value creation. Longo, Mura, and Bonoli (2005) and Torugsa, O'Donohue, and Hecker (2012) have demonstrated that SMEs can enhance their financial performance while proactively making progress toward CSR initiatives. Some empirical evidence have, however, assumed the absence of any kind of relationship between CSR and financial performance. The proponents (e.g. Ullman, 1985) of this position argue that there are so many intervening variables between CSR and financial performance to expect a relationship exist, except by chance. Based on the above discussions, it is hypothesized that: H1: There is a positive relationship between CSR and financial performance.

The link between CSR and stakeholder engagement
There is a natural link between the idea of CSR and firms' stakeholders. Carroll (1991) opined that the word "social" does not really mean much in terms of the parties to whom a firm is responsible. Thus, the concept of stakeholder personalizes social and social responsibilities by identifying the specific groups or persons a business should consider in its CSR orientation. Freeman's (1984) work initiated the re-conceptualization of business organizations as entities that are expected to be sensitive and alert to the constituents in their internal and external environment because these constituents, known as stakeholders, are potent forces to accelerate or decelerate a firm's performance. Relying on stakeholder theory, Prado-Lorenzo, Gallego-Alvarez, and Garcia-Sanchez (2009, p. 95) opined that "the maximization of profits and the creation of value for the shareholders can no longer be the sole objectives of management; rather, they must be obtained through or in co-existence with a grid of values of the other stakeholders." A firm's survival and success depends on the ability of its managers to create sufficient wealth and satisfaction for its stakeholders (Clarkson, 1995).
Stakeholder engagement is explained as "practices the organization undertakes to involve stakeholders in a positive manner in organizational activities" (Greenwood, 2007, p. 315). The impetus behind the use of the term "engagement" instead of other words such as "participation" or "interaction" in the stakeholder theory and CSR literatures is to emphasize the fact that it is no longer sufficient for firms to merely interact with their stakeholders because interaction with stakeholders is a logically necessary activity of business. However, it is possible to trade with another actor without ever engaging him or her as a fellow person; that is, transacting without inquiring as to his or her wants, needs, well-being, or capabilities. In light of this, "engagement," is used to recommend a type of interaction that involves, at minimum, recognition and respect of common humanity and the ways in which the actions of each may affect the other. Because of the financial, physical, and environmental effects that businesses can have on individuals and communities, it is important that businesses actually identify and communicate with those persons who have some legitimate stake in them. Furthermore, because the public may often see businesses as monolithic entities, it is important that they recognize that individuals make up these businesses, and that these individuals are also important stakeholders (Noland & Philips, 2010).
CSR is viewed as "the responsibility of the corporation to act in the interest of legitimate organizational stakeholders" (Greenwood, 2007, p. 315). The thrust of CSR is to bridge the relationship between business and society. It is therefore logical to assume that businesses that are more prone to undertaking CSR initiatives would be compelled to show commitment, through policy and practice, to stakeholder engagement since stakeholders' participation serves as the medium to unearth the real needs and interests of these interest groups. A firm's interaction with its stakeholders is considered as an important aspect of CSR. Pedersen (2006) opined that without an effective interaction between business organizations and their internal and external constituents, companies will find it difficult to grasp the fluctuating nature of the values, attitudes, and behavior of their stakeholders and respond accordingly. Beyond the attainment of social goals, the win-win perspective to CSR initiatives connects CSR efforts to stakeholder engagement in a manner that makes economic sense to firms. The proponents (e.g. Kurucz, Colbert, & Wheeler, 2008) argue that by engaging stakeholders and meeting their needs, firms in turn discover other business opportunities and solutions to expand their investment portfolios leading to enhanced financial performance in the long run. Although, stakeholder engagement is not necessarily the same as CSR, it gives an impression of a firm's commitment to be accountable and responsible to their stakeholders (Greenwood, 2007;Morsing, 2006). Hence, it is hypothesized that, firms that are prone to CSR initiatives are more likely to engage their stakeholders.
H2: There is a positive relationship between CSR and stakeholder engagement.

The link between stakeholder engagement and financial performance
There have been considerable studies on stakeholder concept since the publication of the landmark work by Freeman (1984). The positions held by these works have been diverse and sometimes contradictory (Donaldson & Preston, 1995). However, two key visible strands run through most of these works. The first strand (the normative stakeholder accountability model) argues that organizations are expected to be responsible toward all stakeholders because a firm's activities affect the wellbeing of several stakeholders in addition to shareholders, therefore, the firm is morally responsible and accountable to all these stakeholders (Gray, Owen, & Adams, 1996). The second variant (the stakeholder management model) advanced that the effective management of a wide range of stakeholders can result in improved financial performance (Berman et al., 1999).
The benefits of stakeholder engagement emanates from an increase in shareholder value through insurance-like protection, improved risk management, market appeal to customers, improved transparency, and easier access to financial markets. All or most of these factors could potentially increase the financial performance of firms (Jo & Na, 2012). Employing resource-based theory, effective stakeholder engagement can be viewed as a rare, unique, and inimitable resource that does not only lead to strong financial performance but also sustained financial performance because of the difficulty competitors are likely to encounter in cultivating such relations (Choi & Wang, 2009). Firms that establish relationship with their stakeholders beyond market transactions gain competitive advantage (Mishra & Suar, 2010). Effective management of key stakeholders acts as a value driver by leveraging performance and reducing stakeholder-inflicted costs. For instance, lower employee turnover reduces hiring and training costs, loyal suppliers reduce quality certification costs and supportive communities reduce legal and public relations overheads (Mishra & Suar, 2010).
Empirical studies on stakeholder engagements have documented that firms that relate well with stakeholders tend to improve their financial performance (Barney & Hansen, 1994;Choi & Wang, 2009;Orlitzky et al., 2003;Roman, Hayibor, & Agle, 1999). For instance, Scholtens and Zhou (2008) measured stakeholder engagement by a firm's community involvement, corporate governance, employee relations, environmental conduct, diversity of the workforce, human rights policies, and product attributes and found that these different components of stakeholder engagements appear to be associated in a complex manner with financial performance. Branco and Rodrigues (2006) have argued that firms that engage stakeholders including customers, suppliers, competitors, bankers, and investors turn to have improved financial performance. Donker, Poff, and Zahir (2008) also analyzed the code of ethics of Canadian companies and discovered that the inclusion of key words that reflect stakeholder engagement were positively correlated with financial performance. Jones (1995, p. 442, 430) elucidates the intervening role of stakeholder engagement on CSR-firm performance nexus by advancing that certain types of CSR initiatives are "manifestations of attempts to establish trusting, cooperative firm-stakeholder relationships and should be positively linked to a company's financial performance … firms that contract with their stakeholders on the basis of mutual trust and cooperation will have a competitive advantage over firms that do not … [This advantage stems from] reduced agency costs, transaction costs, and costs associated with team production. More specifically, monitoring costs, bonding costs, search costs, warranty costs, and residual losses will be reduced." Tang, Hull, and Rothenberg (2012) observed that generally stakeholders such as employees, customers, governments, and the media, tend to respond favorably to CSR activities on the part of firms. This favorable response leads to better reputation, improved human capital, and greater innovative capability for firms, which in turn lead to better financial performance. Although, the strength of such a CSR-stakeholder engagement-firm financial performance link can vary in different contexts (e.g. , it has been largely confirmed that CSR activities can establish or fortify the firm-stakeholder relationship (e.g. philanthropic donation; Brammer & Millington, 2008), revive or strengthen the product's differentiation factor in the market (e.g. environmental friendliness; Cheema & Javed, 2017;Klassen & Whybark, 1999), refine the costefficiency aspect of business processes (e.g. "best practices"; Christmann, 2000), attract the best human capital (e.g. Johnson & Greening, 1999), or improve other intangible assets (e.g. reputation; Margolis, Elfenbein & Walsh, 2009) for the firm, which will ultimately benefit the firm financially. The theoretical framework underlying the business case proposes that CSR improves key stakeholder relationships, which decreases costs and increases income and so increases firms' financial performance (Barnett, 2007). Hence, it is anticipated that stakeholder engagement could mediate the relationship between CSR and financial performance (Table 1).
H3a: Stakeholder engagement is positively associated with financial performance.
H3b: Stakeholder engagement mediates the relationship between CSR and financial performance.

Research methodology
The thrust of this study is to establish the mediating effect of stakeholder engagement on the relationship between CSR and financial performance of SMEs employing partial least squares (PLS) estimation technique. The target population comprised all SMEs within the Accra Metropolis of Ghana. The accessible population consists of owner-managers of SMEs which had registered with the National Board for Small-Scale Industries (NBSSI) and the Association of Ghana Industries (AGI) in the Accra Metropolis as at September, 2013. SME owners/managers were the sampling unit because of the key role the play in the administration of these firms. Brancato et al. (2006) advanced that the personal knowledge of respondents can be useful both to answer directly survey questions and to know where to search for information sources across the firm. The total number of SMEs recorded in the NBSSI's and AGI's registers by location in the Metropolis was 2,083. Following Krejcie and Morgan (1970), to ensure a 5% margin error, 254 medium-sized and 302 small-sized firms were randomly selected from 750 medium-sized and 1,400 small-sized firms, respectively. Simple random sampling technique was employed because in Ghana and other parts of Africa data collection is difficult as most people are not used to responding to questionnaires and returning them (Gyensare, Anku-Tsede, Sanda, & Okpoti, 2016).
The instrument used for this study was a questionnaire administered to owners/managers of SMEs by the researcher and five trained research assistants of the University of Cape Coast. According to Sweeney (2009), questionnaires do not emerge fully fledged; they have to be created or modified, shaped and developed to maturity after several test flights. Every aspect of a survey has to be tried out beforehand to make sure that it works as intended. Based on the recommendation of Netemeyer, Boles, and McMurrian (1996), once the questionnaire was developed in reflection of current literature, it was peer reviewed by academic colleagues who have undergone the process of survey development and analysis previously.
While a pilot study is unlikely to reveal all challenges of the main survey, it should result in important improvements to the questionnaire and may influence the scope and perhaps necessity of the main survey (Sweeney, 2009). Hence, a pilot study of the questionnaire was undertaken in February 2014. The questionnaire was administered to a sample of 50 SMEs in the Sekondi-Takoradi Metropolis. Each respondent was informed that this was a pilot study and was encouraged to provide feedback on any problems that they experienced while completing the survey (see De Vaus, 1993). In total, the pilot study recovered a response rate of 62%. The response rate was high (65%) in spite of some complaints from respondents about the length of the questionnaire. Questions relating to ascertaining actual data on financial performance proved to be the most problematic and as such recorded low response rate. Feedback from the respondents indicates that the low response rate is due to the confidential Table 1

Hp1
There is a positive relationship between CSR and financial performance of SMEs + Hp2 There is a positive relationship between CSR and stakeholder engagement +

Hp3a
Stakeholder engagement is positively associated with financial performance +

Hp3b
Stakeholder engagement mediates the relationship between CSR and financial performance nature of the information sought. Therefore, it was evident that it would be extremely difficult to obtain actual profit figures to measure financial performance, rather, subjective measures of financial performance was used to ensure high response rate in the final questionnaire.
The questionnaire used for the study was subdivided into five sections. The first section covered the background information of respondents' age, gender, educational level, work experience, and startup experience. The second section captured information about the organization's size, core business, age, location, ownership structure, and number of branches. The third and fourth sections sought for data on CSR and stakeholder engagement. Some of the CSR information involved energy conservation, supply of clear and accurate information, and labeling of products and services, resolving of customer complaints in timely manner, quality assurance criteria adhered to in production, being committed to the health and safety of employees, recruitment policies that favor the local communities and donation to charity. The stakeholder engagement questions solicited for data on the extent the firm engages with stakeholders who are directly affected by the organization's operations; those who have an interest in, or influence over the organization's operations; stakeholders who have knowledge about the impact of the operations of the firm, authorities or regulators who exercise control over an industry and several other types of stakeholders. The final section collected data on financial performance with respect to sales growth, profit growth, and leverage.

Measurement of variables
The following section presents a narration of how the variables of interest employed in this study were measured. A questionnaire was developed in reflection of the extant literature in the area. It was also peer reviewed by academic colleagues who have undergone the process of survey development and analysis previously. This was carried out to ensure clarity and that no irrelevant questions were included in the survey. All the variables were measured on a continuous scale of 1-5. Following Sweeney (2009), Man (2011, Burton and Goldsby (2009) the study adopted the subjective approach of measuring financial performance consisting of profit growth, sales growth, and leverage. Several studies (Dawes, 1999;Man, 2011;Sweeney, 2009) have adopted this approach to measuring the financial performance of SMEs. According to Man (2011), the use of scales is a better alternative to measure SMEs' performance than to use actual figures due to the unwillingness of SME owners/managers to disclose such sensitive data.

Independent and mediating variables
CSR and stakeholder engagement were measured based on the items employed by Sweeney (2009). Following the position of Sweeney (2009), this research did not seek to reinvent the wheel with regard to the measurement of CSR and stakeholder engagement, instead each question within the section relating to these concepts has been taken from previous research works on the area. Carroll (1991) cites five major groups that are recognized as priorities by most firms, across industry lines and in spite of size or location: owners (shareholders), employees, customers, local communities, and the society at large. However, Sweeney (2009) argue that owners (shareholders) are not major stakeholders with reference to CSR initiatives among SMEs. This is probably due to the rareness of "outside" shareholders besides the founders of these enterprises in funding and managing their operations. This is the case in most developing countries including Ghana (Agyemang & Ansong, 2017). Thus, the dimensions of CSR and stakeholder engagement in this paper centered on employees, customers, community, and environment.
Some of the CSR information involved energy conservation, supply of clear and accurate information and labeling of products and services, resolving of customer complaints in timely manner, quality assurance criteria adhered to in production, being committed to the health and safety of employees, recruitment policies that favor the local communities and donation to charity. Respondents rated the extent to which their respective firms undertake or pursue the aforementioned activities using a five-point continuous scale.
Based on Sweeney (2009), stakeholder engagement was measured based on an eight-item scale consisting of responses on stakeholders directly affected by the firm's operations; stakeholders who have an interest in or influence over the organization's operation; stakeholders who have knowledge about the impact of the operations of the firm; stakeholders who are part of the broader community who have an interest in, concern with, or influence over the operations of the firm; authorities or regulators at the national or local level; authorities who control or issue licenses or permit to operate; authorities or regulators who exercise control over the firm's sector; and authorities responsible for social and economic development, infrastructure and service provision (Table 2). 4. Stakeholders who are part of the broader community who have an interest in, concern with, or influence over the operation of your firm 5. Authorities or regulators at the national or local level 6. Authorities who control or issue licenses or permits to operate 7. Authorities or regulators who exercise control over your sector or industry 8. Authorities responsible for social and economic development, infrastructure and service provision, town and regional planning Financial performance 1. Sales growth Man (2011) 2. Profit growth 3. Leverage (total equity/total debts) Burton and Goldsby (2009)

Analyses
The data were analyzed using PLS analytical approach. The application of this approach does not necessarily require the data collected to have multivariate normal distributions, and it is appropriate for testing theories in developmental stages (Fornell, 1982). In addition, this approach is apt when the goal is a causal predictive test instead of a test of an entire theory (Chin, 1998). Henseler, Ringle, and Sarstedt (2015) posit that an indicator is considered reliable when its outer loading is higher than 0.7. Majority of the indicators used in the model loaded well above 0.7. Indicators that loaded below the minimum threshold value of 0.7 have not been removed because the average variance extracted (AVE) for all the constructs were above the minimum threshold of 0.5 at a significant level of p < 0.05 (see Henseler et al., 2015). Table 3 presents the list of latent variables, indicators retained, and their respective outer loadings.

Checking for reliability and validity of the model
The reliability of each construct was assessed by observing the composite reliabilities of the constructs used in the model. The composite reliability of each construct is well above the minimum value of 0.7, ranging from 0.809 to 0.940 (see Table 4), thereby confirming the reliability of the constructs (Hair, Sarstedt, Hopkins, & Kuppelwieser, 2014).  Convergent validity was assessed by observing the AVE for each construct. As suggested by Hair et al. (2014), the minimum value of the AVE of all the latent variables used in a study should not be less than 0.5. The results (see Table 2) show that the AVE of each latent variable used in this study is above the cut-off value of 0.5, indicating that the requirement of convergent validity has been met. Also, the AVE values of each construct (latent variable) are higher than the squared value of the correlation between the constructs (see Table 5), thereby suggesting the requirement of discriminant validity is achieved. Table 6 shows that each indicator loaded is higher on the construct it is measuring than on all other constructs in the model. This also goes to confirm the discriminant validity of the model (Hair et al., 2014).
Heterotrait-Monotrait Ratio (HTMT) is also another means of determining the discriminant validity of a PLS-SEM model (see Table 7). According to Henseler et al. (2015), a latent construct has discriminant validity when its HTMT ratio is below 0.825. As presented in Table 5, the HTMT ratios of all the constructs used in the model were well below the threshold value of 0.825 indicating that the constructs used in the model have discriminant validity.
The multicollinearity tests results for the independent variables employed in the study are reported in Table 8. The calculated Variance inflation factor (VIF) in all cases is less than 2 indicating that there is no multicollinearity problem for the analysis (Henseler et al., 2015).   Table 9 presents the results of the R 2 values in the structural model. It shows that financial performance and stakeholder engagement have R 2 values of 0.097 and 0.416, respectively. Additionally, Table 9 indicates Q 2 values, which are used to assess the predictive relevance of a reflective structural model. The results show Q 2 values of 0.051 for financial performance and 0.252 for stakeholder engagement, respectively, which met the general requirement that Q 2 should be greater than 0 (Hair et al., 2014). Table 9, all the hypothesized direct relationships (with the exception of CSR to financial performance) were supported by the structural model data. CSR accounted for 64.5% of the variance in stakeholder engagement whereas stakeholder engagement explained 24.4% of the variance in financial performance. In line with existing CSR literature, it was found that CSR had a significant positive effect on stakeholder engagement (β = 0.645, p = 0.000) and stakeholder engagement also had a significant positive effect on financial performance (β = 0.244, p = 0.000). Thus, the results lend support to hypothesis 2 and hypothesis 3a. However, hypothesis 1 which predicted the direct path from CSR to financial performance was not supported (β = 0.092, p > 0.05).

Mediation analysis
From Table 9 and Figure 1, the results of the structural model indicate that CSR contributes nonsignificantly in predicting financial performance (β = 0.092, p > 0.05). Additionally, the results also show that CSR plays a crucial role in determining (β = 0.645, p < 0.000) stakeholder engagement. Stakeholder engagement also plays a significant role in explaining financial performance (β = 0.244, p < 0.005). As indicated earlier, while the direct effect between CSR and financial performance is not significant (β = 0.092, p > 0.05), the indirect effect (β = 0.157, p < 0.005) between CSR and financial performance is significant (see Table 10). Since the indirect effect of CSR on financial performance is significant while the direct effect is not, it can be concluded that stakeholder engagement fully mediates the positive relationship between CSR and financial performance. Hence, hypothesis 3b is supported by the study.

Implications for theory
The findings of this study represent a step beyond previous CSR research, which has reported a direct linear relationship between CSR and firm performance. In particular, the study found an indirect effect of CSR on SMEs' financial performance through stakeholder engagement. The findings of this study provide a theoretical support to the dual opinions held about stakeholder theory by researchers (Donaldson & Preston, 1995;Freeman, 1983). First, the findings advance the normative stakeholder accountability model view that organizations are expected to be responsible toward all stakeholders because a firm's activities affect the well-being of several stakeholders in addition to shareholders (Gray et al., 1996). In addition, it provides support to the stakeholder management model which advocates that the effective management of a wide range of stakeholders can result in improved financial performance (Berman et al., 1999). The findings extend earlier findings on the important role intervening variables plays in explaining the relationship between CSR and firm performance (Agyemang & Ansong, 2017). Specifically, the findings provide empirical evidence that the effect of CSR on financial performance is influenced by stakeholder engagement (Ansong & Agyemang, 2016;Fombrun, 2005;Fombrun & Shanley, 1990;Freeman et al., 2007). This is in line with Waddock and Graves (1997) assertion that CSR initiatives that do not involve stakeholders are less likely to have any reward for firms.
Furthermore, an understanding of the study setting is imperative. Individualist cultures assume individuals look primarily after their own interests and the interests of their immediate family (husband, wife, and children). Collectivist cultures assume that individuals -through birth and possibly later events -belong to one or more close "in-groups," from which they cannot detach themselves. The in-group (whether extended family, clan, or organization) protects the interest of its members, but in turn expects their permanent loyalty. A collectivist society is tightly integrated while an individualist society is loosely integrated (Hofstede, 1980). Ghana is a collectivistic culture and therefore the Ghanaian context provides support for the extrapolation of the CSR-stakeholder engagementfirm performance research findings in a collectivist culture as against other similar research   (Gyensare et al., 2016). Finally, although the SME sector is seen as the engine of Ghana's economy, there is a paucity of research in this context. Hence, this study is the first of its kind to empirically examine the indirect effect of CSR on financial performance through stakeholder engagement among SMEs in Ghana.

Implications for practice
The findings of this study provide some implications for managers of SMEs. First, since stakeholder engagement improves the financial performance of firms, the study recommends that managers should pursue robust engagement of stakeholders by incorporating engagement tactics in their routine operations in a way that creates a safe, trusting, and collaborative environment in which stakeholders feel engaged, valued and motivated to contribute to the wealth creation efforts of firms. Most essentially, the study provides understanding of the mediating mechanism that relates CSR to financial performance. The study also highlights the potential importance of CSR in organizations at enhancing their relationship with their stakeholders. It was established that SMEs with improved CSR practices are better positioned to engage more with their stakeholders, which translates into improved financial performance. Thus, it is recommended that for SMEs to improve upon their CSR practices, which will eventually result in enhanced financial performance, stakeholder engagement should be a major part of their operations.

Limitations and future research
The findings of this study should be interpreted with some considerations in the light of the following shortcomings. The first limitation deals with the sample employed for the study. Although, the study has undoubtedly contributed to understanding the mediating role of stakeholder engagement to SMEs' financial performance, they should be treated as preliminary until future studies replicate that with samples from a broad range of organizations consisting of both small and large firms. Future longitudinal and experimental research would help confirm the causal paths investigated in the present study. Again, the study relied on self-reported measures. Despite the fact that some researchers have shown that common method bias is trivial (e.g. Crampton & Wagner, 1994;Spector, 1987) and rarely strong enough to invalidate research findings (e.g. Spector, 2006), it is possible that the findings of the present study may be inflated by same source bias. It is recommended that future studies incorporate objective financial performance measures, and measures that tap directly into the notion of stakeholder engagement and CSR initiatives.

Conclusions
The purpose of the present study was to examine the relationship between CSR and financial performance of SMEs in Ghana, by focusing on the mediating role of stakeholder engagement. It was argued that the inconsistent relationship between CSR and financial Performance of firms could stem from the mediating role of stakeholder engagement. The findings indicate that firms that embark on CSR initiatives tend to engage their stakeholders more than those who do not. Moreover, firms that engage their stakeholders tend to perform better financially. Overall, the analyses did not support the view that CSR has a direct significant positive relationship with financial performance; however, the results confirmed that stakeholder engagement fully mediates the nexus between CSR and financial performance.