What is the weakness and the strength of research method of determinants of corporate social responsibility disclosure: the case of Islamic banks Authors):Sayd Farook?

Electronic copy available at: http://ssrn.com/abstract=1828624 DETERMINANTS OF CORPORATE SOCIAL RESPONSIBILITY DISCLOSURE: THE CASE OF ISLAMIC BANKS S A Y D FA R O O K , M . KA B I R HA S S A N A N D RO M A N LA N I S Sayd Farook Global Head of Islamic Capital Markets 10th Floor, West Tower, Bahrain Financial Harbour PO Box 1030, Kingdom of Bahrain Phone: 973 1750 2033 Email: [email protected] M. Kabir Hassan* Department of Economics and Finance University of New Orleans New Orleans, LA 70148 Phone: 610-529-1247 Email: [email protected] ; [email protected] Roman Lanis School of Accounting University of Technology, Sydney PO Box 123, Broadway NSW 2007, Australia Fax: +61-2-9514-3669 Email: [email protected] * C O R R E S P O N D I N G A U T H O U R Electronic copy available at: http://ssrn.com/abstract=1828624 1 DETERMINANTS OF CORPORATE SOCIAL RESPONSIBILITY DISCLOSURE: THE CASE OF ISLAMIC BANKS A B S T R A C T Islamic banks offer distinct financial services and as such have grown significantly in Bahrain, Bangladesh, Jordan, Kuwait, Malaysia, Qatar, Turkey and U.A.E over the past two decades. They are unique in the sense that they are accountable t o fulfil a social and ethical role inherent in their character as an ‘Islamic’ institution. They also ha ve a duty to discharge their accountability through disclosing corporate social responsibility (CSR) information consistent with the principles of Islam. However, recent anecdotal evidence finds that Islamic banks may not be fulfilling their social role in accordance with the prescriptions of Islam because they disclose less CSR information than expected. It has been suggeste d that disclosure may also be driven by the extant economic incentives. Hence, the exact nature of the CSR disclosure process by Islamic banks remains unclear due to a paucity of a priori research and statistical analysis of extant data.

In light of that, this study develops an a priori model linking CSR disclosure to socio-political influences and corporate governance factors. Then r esultant hypotheses are tested using regression analyses on a sample of 47 Islamic banks ’ annual reports from 14 countries. Categorisation: Research Paper Keywords: Corporate Governance, Corporate Social Re sponsibility Disclosure, Islamic Banks, Political Economy, Legitimacy Theory. JEL Classification: G21, M14, Z12, Z13 2 DETERMINANTS OF CORPORATE SOCIAL RESPONSIBILITY DISCLOSURE: THE CASE OF ISLAMIC BANKS Introduction A combination of political, economic and demographi c factors, including and not limited to the impact of the Iranian revolution, a growing Muslim middle class, the rise of the Asian tigers, increased deregulation and the oil shocks of the 19 70’s have stimulated the development of Islamic banks (Akacem and Gilliam, 2002).[1] Some r eports suggest that Islamic banks are growing at an annual rate of more than 15% making t hem the fastest growing segment of the credit market in Muslim countries (Zaher and Hassan , 2001; Bahrain Monetary Agency, 2004).

The latest figures suggest that the assets of Islam ic banks worldwide stand at over US$260 billion, while the number of Islamic banks has incr eased from 176 in 1997 to 267 in 2004 operating in over 60 countries throughout the world (Bahrain Monetary Agency, 2004).

Islamic Banks should ideally operate in accordance with the principles laid down by Islamic law ( Shari’a). [2] The primary contributing factor that hastened t he need for Islamic Banks is the prohibition of usury ( riba).[3] Merged with this function is the social role of Islamic Banks that entails social justice and accountability, requirin g the banks to disclose corporate social responsibility (CSR) information. Usmani asserts th at the philosophy behind Islamic banking was “aimed at establishing distributive justice free fr om all sorts of exploitation” (2002, p. 113).

According to Islamic principles, business transacti ons can never be separated from the moral objectives of society (Usmani, 2002). As such, a nu mber of scholars have developed a normative standard for reporting (Gambling and Karim, 1986, 1 991; Baydoun and Willett, 2000; Lewis, 2001) and indeed social reporting for Islamic busin esses based on Islamic principles (Haniffa, 2001; Maali, Casson and Napier, 2003). Governments in Muslim populated countries such as Malaysia and international regulatory institutions such as the Accounting and Auditing Organisation for Islamic Financial Institutions (AA OIFI) have also voiced their support for the development and adoption of such CSR reporting stan dards encouraged and propagated by Islam (Sharani, 2004; Yunus, 2004). Recent ad hoc studies indicate that Islamic banks are not comple tely fulfilling their social role in accordance with the prescriptions of Islam (Metwall y, 1992; Aggarwal and Youssef, 2000; Maali et al., 2003). Usmani emphasises that Islamic banks “were supposed to adopt new financing 3 policies and to explore new channels of investment which may encourage development and support of small scale traders to lift up their eco nomic level” (2002, p. 116). Islamic banks should have advanced towards profit and loss sharing (musharakah) in gradual phases and should have increased the size of musharakah financing (Usmani, 2002). However, it remains that very few Islamic banks and financial institutions have paid attention to this social aspect. Usmani (2002) further highlights that in a number of Islamic bank s, other permitted forms of financing are not effected according to the procedures required by th e Shari’a. Additionally, Aggarwal and Youssef find that while Islamic banks are expected to “favour small entrepreneurs who do not have access to credit in t he conventional banking system” (2000, p. 99), they rarely offer finance to these segments of the market, contrary to Islamic injunctions to promote the development of the under-privileged ech elons of society. They infer that this is a rational response by Islamic banks in the face of s evere agency problems in their attempts to provide funds to entrepreneurs. This leads them to conclude that “economic incentives shape the structure of Islamic banking more so than religious norms” (Aggarwal and Youssef, 2000, p. 99).

Maali et al.’s (2003) rudimentary analysis also suggests that I slamic banks’ CSR reporting falls short of the benchmark for entities whose operation s are founded on Islamic principles. Based on an Islamic perspective, they develop a pragmatic be nchmark for social disclosures that they would expect Islamic Banks to provide. They find th at there is considerable variation in the voluntary social reporting of Islamic banks, from s ome banks reporting 35% of expected social disclosure to others disclosing almost no social in formation (Maali et al., 2003). In addition, they find that the annual reports of Islamic banks prese nt elements in the process of constructing an Islamic reality (Hines, 1988; Maali et al., 2003). As such, Maali et al. conclude that “with a few exceptions, Islamic banks have a long way to go to meeting expectations of the Islamic community” (2003, p. 31). However, they fail to pro vide an a priori basis to support their results.

In addition, their sampling and statistical analyse s are rudimentary leaving their conclusions theoretically inexplicable (Maali et al., 2003). Consistent with Maali et al. (2003), the first objective of this study is to me asure the annual report social disclosure levels of Islamic banks based on a benchmark derived from Islamic principles.

More importantly, the second objective of this pape r is to ascertain a priori the determinants of Islamic banks’ social disclosures which will subseq uently be tested utilising the disclosure measures obtained. . In particular, tests for the determinants of social disclosure from a legitimacy 4 and political economy perspectives are to be performed (Gray, Kouhy and Lavers, 1995; Williams, 1999). Departing from previous research, this study also tests for the relationship between corporate governance mechanisms and CSR dis closure. This study contributes to the limited literature in the area of international corporate governance research and to the growing literature on Islam and its impact on corporate disclosure. It does so by providing empirical evidence on the nature and d eterminants of social disclosures by Islamic Banks. Further, this paper contributes to the empir ical research exploring corporate governance mechanisms and their effect on social disclosures a nd corporate transparency. 2. Prior research Despite the growth of Islamic banks in size and com plexity, few researchers have addressed the issue of social responsibility in the context of Is lamic banks. While a number of papers have explored Islamic accounting and corporate reporting of Islamic institutions and banks, they have generally been either normative or analytical in na ture and lack the empirical analysis of disclosure practices of Islamic organisations. Gambling and Karim’s (1986) seminal work developed a theoretical foundation for the analysis of Islamic accounting elaborating on its peculiarit ies and analysing its social orientation. Hamid, Craig and Clarke (1993), Karim (1995) and Lewis (20 01) subsequently detailed the intricacies of the influence of Islam on all areas of accounting f rom practice through to disclosure paving the path for developing a conceptual framework for Isla mic accounting and disclosure requirements particularly for Islamic banks. Sadeghzadeh (1995) enhances the structure of an Islamic perspective of social responsibility and sustainabi lity accounting by giving it theoretical depth and contrasting it with conventional theories relat ing to the area. Baydoun and Willett (2000) supplement Sadeghzadeh (1995) by developing normati ve Islamic reporting standards comprising value added statements following the two complement ary principles of full disclosure and social disclosure based on Islamic ethical values. The stu dies above provide the overall theoretical framework and practical application of Islamic acco unting.

Dealing specifically with Islamic banks, Archer, Ka rim and Al-Deehani (1998) detail the contractual basis of Islamic banks to outline the s pecial need for corporate governance and disclosure by Islamic banks due to the monitoring w eaknesses inherent in the Islamic banking system. They recommend a number of solutions rangin g from tighter ex ante contractual 5 conditions to improving transparency of financial reporting and monitoring. Their analysis is a significant attempt to extrapolate the various cont racting issues inherent in Islamic banks. The only noteworthy exceptions to these normative a nd analytical papers in the area of Islamic corporate reporting and social reporting are the wo rks of Askary (2001), Maali et al. (2003) and Haniffa and Hudaib (2007) . Askary (2001) draws on research examining the influence of culture on accounting to classify accounting practices in d ifferent Muslim countries according to cultural variables developed by Hofstede (1980), Gray (1988) and Perera (1989). He compares the actual disclosure practices of companies in Muslim countri es to the benchmark of Islamic accounting practices as measured by those cultural variables. However, his research focuses on all companies in Muslim countries rather than on Islamic banks sp ecifically and he measures overall disclosure rather than CSR disclosure. Lastly, Maali et al. (2003) make a significant attempt to substantiate the actual social disclosure practices of Islamic banks. They utilize a sample o f 29 Islamic banks from a number of different countries and compare CSR disclosure to a pragmatic benchmark based on Islamic values. They find that Islamic banks are disclosing CSR informat ion far below the expected level. However, their rudimentary analysis makes only vague inferen ces as to what may drive the social disclosure practices of these banks. They conjecture as to a n umber of possible explanations from the CSR literature (Maali et al., 2003). Alluding to economic incentives that may dr ive CSR disclosure, Maali et al., (2003) also suggest that Islamic banks may only di sclose CSR information to construct an Islamic reality while not subscribing to that reality and its resultant obligations. Haniffa and Hudaib (2007) examined the ethical iden tity of Islamic banks in the Gulf region.

Based on published annual report, Haniffa and Hudai b measured the level of ethical identity for seven Islamic banks in the Arabian Gulf Region base d on ideal versus communicated ethical identity framework. The gap between ideal and communicated ethical identity for Islamic banks was wide. In their study, they used eight dimensio ns of ethical identity namely (1) mission and vision statement, (2) board of directors and top ma nagement, (3) products and services, (4) zakah, charity and benevolent funds, (5) commitments towar ds employee, (6) commitment towards debtors, (7) commitment towards society, and (8) Shariah supervisory board. 6 3. Theoretical Framework Corporate social responsibility disclosure has been the subject of substantial academic accounting research. Gray, Owen and Maunders define CSR disclo sure: …as the process of providing information designed t o discharge social accountability.

Typically this act would…be undertaken by the accou ntable organisation and thus might include information in the annual report, special p ublications or even socially oriented advertising (1987, p. 4). To that effect, a number of theories have been prop osed to explain why corporations voluntarily disclose CSR information. A review of accounting research indicates that theo ry development related to CSR disclosure in general is fragmented and rudimentary, while almost no theory development has occurred in relation to CSR disclosure of Islamic banks (Maali et al., 2003; Sadeghzadeh, 1995). At most, the literature on Islamic banks suggests that a priori there are two major influences on the Islamic banks CSR disclosure:

1. socio-political context within which the banks o perate; and 2. economic opportunities available to Islamic bank s.

In the CSR literature, the former influence is iden tified with systems oriented theories such as political economy, legitimacy and stakeholder theor ies (Wilmhurst and Frost, 1999; Deegan 2002; Campbell, Craven and Shrives, 2002). As part of that program, the political and social contexts have been found to be important determinan ts of the decision to disclose CSR information (Roberts, 1992; Williams, 1999). The economic incentives viewpoint is consistent wit h research that explains CSR disclosure in the context of agency theory (Cowen, Ferreri and Pa rker, 1987; Adams 2002; Campbell, 2000).

Thus, the theoretical framework development here wi ll incorporate both influences. The proposed model is presented in Figure 1. FIGURE 1 HERE The diagrammatic portrayal is an endeavour to captu re the two major possible influences. The socio-political context variables, as derived from systems oriented theories, measure cross- national differences whereas the corporate governan ce variables measure organisation specific differences that may influence the CSR disclosure o f Islamic banks. 7 3.1. Systems theories, social responsibility and Islam Systems oriented theories such as political economy , stakeholder and legitimacy propose that individuals, institutions and organizations, seekin g to preserve their own self-interest, attempt to operate and interact within the system through vari ous relationships with others (Williams, 1999).

The theories also emphasise “that the actors, wheth er they are individuals or organisations, in this system have the right to pursue their own goals and self-interests” (Williams, 1999, p. 211).

However, these rights to self-interest are moderate d by the social and political environment in which they interact (Williams, 1999). This idea is consistent with Islam where the concep t of Unity (Tawhid) prevails. According to this concept, God is the Creator, Owner and Source of al l things (Maali et al., 2003).[4] In light of God’s ownership of everything, it is believed that God has entrusted mankind to the use of resources.[5] Thus, in return for the use of the ph ysical universe, mankind agrees to be accountable for how the universe is used (DeLorenzo , 2002). This position of trust is the source of accountability for individuals and consequently organisations.[6] The trusteeship requires a total commitment to the will of God and therefore “ involves both submission and a mission to follow the Shari’a in all aspects of life”, including economic aspects (Baydoun and Willett, 2000, p. 80). However, Islam does not deny individual rights to s elf-interest. Enjoyment of self-interest is only conditioned by the permanent needs of greater socie ty (Umma ) (Sadeghzadeh, 1995). As such, individual freedom is sacred only as long as it doe s not conflict with the larger societal interest or as long as the individual does not transgress the r ights of others (huqquq-al-ibad) . Some practical examples of this concept are demonstrated by the fo rbiddance of a number of activities such as drinking alcohol, adultery and gambling because of their contributory effects to families and societies. The conceptual basis of this implicit co ntract between the individual and greater society are emphasised in great lengths in the Qur’an and the teachings of Prophet Muhammad. Similarly defined by the proponents of political ec onomy, stakeholder and legitimacy theories, the relationship between individuals, organisations and society is consequently viewed as a ‘social contract’ (Ramanathan, 1976; Deegan, 2002; Williams, 1999). Organisations themselves play a large role in socie ty and have responsibilities assigned to them 8 based on their status in society. As such, they “…exist only to the extent that the particular society considers that they are legitimate” (Deegan , 2002, p 292). Accordingly, organisational legitimacy is defined a s: …a status, which exists when an entity’s value syst em is congruent with the value system of the larger social system of which an entity is a part. When a disparity, actual or potential exists between the two value systems ther e is a threat to the entity’s legitimacy (Lindblom, 1994 cited by Newson and Deegan, 2002, p . 184). To paraphrase, organisations continually seek to en sure that they operate within the bounds and norms of their respective societies (Deegan, 2002). In this context, CSR is defined broadly as “including the concern for the impact of all of the corporation's activities on the total welfare of society” (Bowman and Haire, 1976, p. 13). The concept of corporate social responsibility in I slam likewise emerges from this ‘social contract’ that neccesarily has to do with the congr uency of the value system of the organisation to the larger value system of an Islamic society.[7] I n Islam, organisations are similarly, if not more so, accountable to society as are individuals. As L ewis (2001) elaborates, the implications for business enterprises is that “both managers and pro viders of capital, are accountable for their actions both inside and outside their firms; accoun tability in this context means accountability to the community” to establish socio-economic justice within their own capacity (2001, p.113). A number of commandments in the Qur’an and the tradition of the Prophet Muhammad stipulat e what must be done in order to establish socio-econo mic justice and therefore be socially responsible. Some examples of these are the obligatory payment out of income and wealth (zakah) , philanthropic trusts (waqf), alms, charity (sadaqa), interest free loans ( qard-ul-hassan) (Sadeghzadeh, 1995). The forbiddance of riba also stems out of principles of socio-economic justice in Islam in that the objective is to disall ow any unjust distribution of wealth through forced or undeserved loss to one party or unearned gain to the other party (fixed interest). Against that background, it is reasonable to conclude that Islam ic business values and norms are consistent in general with the definition of corporate social res ponsibility.

3.2. CSR disclosure as a means of legitimation To discharge their corporate social responsibility, political economy, legitimacy and stakeholder theories’ proponents argue that corporations (manag ement) provide CSR information as part of 9 the dialogue between the corporation and greater society (Gray et al., 1995). Even if the organisation is complying with society’s expectatio ns, organisational legitimacy can be threatened if it has failed to make disclosures tha t show it is complying with societal expectations (Newson and Deegan, 2002). Hence, managers need to demonstrate that they are complying with the ‘social contract’ by disclosing information in line with society’s expectations.[8] Although different obligations and responsibilities are due of different types of organisations in different contexts, the overall general framework f or social responsibility and accountability in Islam is derived from Islamic teachings embodied in the detailed jurisprudence surrounding the Qur’an and the teachings of Prophet Muhammad. Hence, the expectations of an Islamic populace towards any organisation that claims to be Islamic are unambiguous. Islamic banks are expected to disclose relevant CSR information to discharge t heir responsibility and to earn legitimacy for their continued existence (Sadeghzadeh, 1995; Baydo un and Willett, 2000; Haniffa, 2001; Lewis, 2001; Maali et al., 2003). However, the expectation to disclose is o nly a necessary condition for disclosure. The disclosure of CSR information by Islamic banks will depend on a number of other factors necessarily “focussing on the role of information a nd disclosure in the relationships” between the organisations, the State, individuals, groups and p articularly the Islamic society (Umma) (Gray, Owen and Adams, 1996, p.45). In light of this, two factors from systems oriented theories are identified that will directly influence the level o f CSR disclosure by Islamic banks. Extracted from political economy theory, the first factor foc uses on the broader social and political environments in which organisations interact and is captured by the ‘political rights and civil liberties’ variable (Williams, 1999). The second fa ctor, proxied by the ‘relative size of Muslim population’ variable, is derived from legitimacy th eory and attempts to capture the concept of the ‘relevant publics’ from which the organisation requ ires legitimation to exist (Newson and Deegan, 2002). Figure 2 indicates the proposed effe ct of both the socio-political factors on the organisation. The figure demonstrates that multiple groups of ‘relevant publics’ or ‘stakeholders’ influence the organisation within the context of th e social, political and economic frameworks represented by the ‘political economy’ (Gray et al., 1995). In addition, Figure 2 indicates that each theory addresses a ‘different level of resolut ion of perception’ (Gray et al., 1995). [9] FIGURE 2 HERE 4. Hypothesis Development 10 4.1. Hypothesis: Political Economy As Islamic banks operate in a broader social and po litical environment, the interplay of power affecting the rights and responsibilities of all ac tors within this environment will determine the flow of information and dialogue. This will consequ ently affect the level of CSR disclosure by Islamic banks. 4.1.1. Political rights and civil liberties One factor that may influence the interplay of powe r is the political rights and civil liberties in a country. Williams suggests that “violations of poli tical rights and civil liberties associated with various forms of political structure may restrict p olitical and civil freedom and therefore act as a hindrance to full and fair disclosure” (1999, p. 21 3). Gastil (1981) points out that as political and civil repression increases, the influence and effec tiveness of social interest bodies decrease. For example, repressive governments will restrain the a ctivities of institutions such as the press which ensure a level of accountability through the effect ive flow of information and dialogue. Within countries exhibiting limited political right s and civil liberties, societal interest groups, whether Islamic or otherwise, lack the capacity to voice their concerns regarding organisational conduct. Under such repressive regimes, organisatio ns may face lower social expectations and pressure (Williams, 1999). On the other hand, organ isations and particularly Islamic banks operating in relatively open communities with great er freedom may need to provide further justification to legitimize their existence and hen ce disclose CSR information. As with Williams (1999), a significant negative relationship is expe cted between the levels of CSR disclosure presented by Islamic banks and the extent of politi cal and civil repression. H 1: There is a negative association between the level of political and civil repression and the level of CSR disclosure presented in annual reports by Islamic Banks. 4.2. Hypothesis: Relevant Public 4.2.1. Relative size of Muslim population While the broader social and political environment may affect the flow of information and the effectiveness of social interest groups in a partic ular country, the relative size of the Muslim population as a proxy for the Islamic society will contemporaneously determine the level of CSR disclosure presented by Islamic banks. 11 Newson and Deegan (2002) point to the crucial facto r that is directly relevant to management: the ‘relevant publics’ to which the organisation is acc ountable. As articulated in the theory development section, it could reasonably be justifi ed that the main ‘relevant public’ in the case of Islamic banks is the Islamic society. Applying News on and Deegan’s (2002) concept, while there may be limited expectations about social responsibi lities within a particular country, if an Islamic bank relies on support from the Islamic public, the n it must demonstrate adherence to the expectations thereof. Newson and Deegan (2002) elaborate that while there may be national differences across countries with regards to disclosure perhaps attrib uted to cultural and other factors, their notion of a ‘global’ culture and Islamic culture in this stud y, should work against the differences in CSR disclosure policies. Reiterating this point, Maali et al. (2003) indicate that Islamic banks claiming to follow Islamic principles are required to make c ertain universal voluntary disclosures regardless of local standards, because the need to report such items is based on accountability to the Islamic public or society. However, the extent to which Islamic banks comply w ith these universal expectations of CSR disclosures depends on the relative power of the ‘r elevant public’ to influence the activities of Islamic banks. Given that the relevant public in th e case of Islamic banks is the Islamic society, the proportion of population in a country that adhe re to the principles of Islam or in other words, the relative size of the Islamic society could argu ably represent the influence of the ‘relevant public’ for Islamic banks (Karim, 1990). If the ‘re levant public’ comprise a larger proportion of the overall population, there will be increased pre ssure on the Islamic banks to legitimise their actions to this constituency who enjoy a relatively stronger position in the social and political environment within which Islamic banks operate (Rob erts, 1992). Hence, a significant positive relationship is expected between the proportion of adherent Muslims and the level of CSR disclosure presented by Islamic banks. H 2: There is a positive association between the propo rtion of adherent Muslims (in a Muslim country) and the level of CSR disclosure presented in annual reports by Islamic banks. 4.2.2. Corporate governance 12 It has been suggested that Islamic banks are also d riven by economic realities within which they operate. According to Archer et al. (1998), Islamic banks are an efficient means of co ntracting for investors and fund users wishing to comply with the laws and principles of Islam. The banks usually comprise a group of shareholders that attra ct funds for investment using a version of the mudaraba contract. The mudaraba contract is a profit sharing financial instrument w ith “detailed juristic rules derived from the Shari’a which regulates the relationship between investment account holders (IAH) as providers of funds and the bank in its capacity as entrepreneur ( mudarib) ” (Archer et al., 1998, p. 152). The aggregate investment portfolio of an Islamic ba nk is financed by IAHs’ funds[10], shareholder’s equity and other sources of funds ava ilable to the bank (Archer et al., 1998). The bank’s management acts as an agent not only for the shareholders, but also for investment account holders as the mudarib. The Islamic bank then invests these funds in acti vities acceptable in the Shari’a. [11] Since Islamic banks cannot earn or pay any interest , Islamic banks invest their funds through profit and loss sharing arrangements such as partne rship contracting (mudarabah or musharakah) or other acceptable contracts such as mark-up (murabahah), deferred payment (bay muajjal), leasing (ijarah), forward sale contract (salam) and manufacturing contract (istisna) (Usmani, 2002). 4.2.3. Agency Problems In light of the above, a number of agency problems are encountered in the contractual structure of Islamic banks which may also affect the CSR disclos ure presented by Islamic banks in their annual reports. The principal problem is that the two types of prin cipals (the IAH and shareholders) have inferior information to that possessed by management, partic ularly about the application of Islamic laws in relation to the banks’ operations. Islamic banks are under an implicit contractual obligation to both their shareholders and IAH to function accordi ng to the laws and principles of Islam. Bakar states that “ Shari’a compliance is the very essence of an Islamic bank and its banking business” (2002, p. 76). 13 Another agency problem is the fact that investment accounts are not a liability with a fixed claim on the company’s assets and hence are only given a residual claim to the bank’s earnings or assets pari passu with that of shareholders (Archer et al., 1998). The IAH has no formal right to show their disapproval of management actions except to ‘ vote with their feet’ (Archer et al., 1998).

Archer et al. (1998) explain that in that case, due to the inhere nt benefits of higher returns from funds invested by IAH, shareholders may ‘vicariousl y’ monitor for their other counterparts: the IAH. They state that the relationship between IAH a nd shareholders “exhibits some features of bilateral dependency, in that the IAH depend on shareholders for monitori ng while the shareholders depend on IAH as a source of profits v ia the mudarib share” (Archer et al., 1998, p.

164). The IAH, if it is comprised of Islamic invest ors would also be interested in the level of compliance of the bank with Islamic laws and princi ples. Consequently, the extent of Shari’a compliance by an Islamic bank will depend on the l evel of monitoring in place to limit the divergence of inte rest between the principals who are particularly interested in Shari’a compliance of the bank and the agent which is the b anks’ management.

Karim (1990) broadly classifies the three main type s of shareholders of Islamic banks:

Management Islamic investors [12] Economic investors [13] (Karim, 1990). The same categorisation could be used for classifyi ng IAH perhaps with the exclusion of management shares. The segment most interested in t he banks’ compliance with Islamic laws and principles within this categorisation would be the Islamic investors. The greater the level of monitoring by Islamic investors, the greater the co mpliance of the Islamic bank with Islamic laws and principles. Hence, the extent of CSR disclosure could arguably depend on the level of monitoring by the Islamic investor group. Two major determinants of the level of monitoring are identified in the literature: monitoring mechanisms and ownership structure. 4.3. Hypothesis: Monitoring Mechanism 4.3.1. Islamic governance: The Shari’a supervisory board A number of Islamic banks employ a special form of monitoring to limit the divergence of interest between Islamic investors and the manageme nt of the Islamic bank. Shari’a supervisory boards (SSB) assure investors of the compliance of Islamic banks with Islamic laws and principles. The demand for the services of an SSB a rises out of a “perceived need to constantly 14 check innovations in banking practice [as well as i n accounting] against the principles of Islamic orthodoxy” (Karim, 1995, p. 287). It is not mandato ry for an Islamic bank to have its own SSB.

However, the AAOIFI require both the SSB and the fi nancial auditors of Islamic banks to report on compliance with Shari’a doctrines (AAOIFI, 2004). The AAOIFI standards exp licitly state that the Shari’a supervision is “intended to investigate to what ex tent the financial institution has adhered to Shari’a rules and principles in all its activities” (Bakar 2002, p. 81). The investigation would include examination of the “bank’s memorandum and articles of association, its contracts…its financial reports and various other r eports…” (Bakar, 2002, p. 81). Karim stresses that “in most cases, SSB’s authority is equal to th ose of external auditors” (1995, p. 35). Ideally, one would expect the SSB to represent orth odox Islamic laws and principles more so than management. The Islamic credentials of the members of the SSB are considered to be impeccable (Karim, 1990). If an SSB is employed to ensure comp liance of the Islamic bank to Islamic laws and principles, it can be deduced that it may play a role in mandating CSR activities and also CSR disclosure. However, the extent to which the SS B influences CSR disclosure depends on the function of the SSB in monitoring on behalf of inve stors. The extant literature suggests two competing viewpo ints: the impact of internal governance mechanisms on corporate disclosures may be compleme ntary or substitutive (Ho and Wong, 2001). A greater magnitude of disclosures would be expected if it is complementary as “more governance mechanisms will strengthen the internal control of companies and provide an intensive monitoring package for a firm to reduce o pportunistic behavior and information asymmetry” (Ho and Wong, 2001, p143). Islamic banks would thereby be under greater pressure from the SSB to comply with Islamic laws and princi ples and hence disclose more CSR information. This view is consonant with the role a nd function of the SSB described above. Alternatively, corporate governance mechanisms may be substitutive and may result in lower disclosures (Ho and Wong, 2001). Where there is an additional governance mechanism installed that leads to greater monitoring, the need for disc losure as a form of monitoring then decreases (Ho and Wong, 2001). The SSB may see no need to em phasize additional CSR disclosures if the bank’s activities are complying with Islamic laws a nd principles and investors are assured of that through the Shari’a compliance report (Maali et al., 2003). If information asymmetry can be reduced because of existing monitoring packages suc h as the SSB, the need for installing 15 additional monitoring through greater CSR disclosur es to assure Islamic investors of the banks commitment to Islamic laws and principles is lower (Ho and Wong, 2001). While these two conflicting viewpoints have not bee n totally resolved, Hill argues that no single mechanism is a governance panacea and suggests that it “is desirable to have a system of overlapping checks and balances” (Ho and Wong, 2001 , p. 143). Nothwithstanding, it is expected that the former ex planation holds true. This is because the nature of compliance with Islamic laws and principles from an Islamic point of view entails not only assurance of compliance through issuing the Shari’a report, but also greater involvment in CSR activities and CSR disclosures. The SSB’s function as stated by the AAOIFI also concurs with this rationale. Hence, it is generally expected tha t the existence of a SSB in an Islamic bank would lead to greater levels of CSR disclosures. However, while the existence of a SSB may lead to g reater monitoring and thereby greater disclosures of CSR information, the degree to which the SSB would influence CSR disclosures may also depend on the characteristics of this corp orate governance mechanism (Haniffa and Cooke, 2002; Ho and Wong, 2001). Hence, a multitude of factors that relate to the SSB’s characteristics may determine how effectively the S SB conducts its function and subsequently the level of CSR information disclosed by Islamic banks . A number of characteristics are elaborated upon, after which a hypothesis is formulated. 4.3.2. Number of board members An increase in the number of SSB members may lead t o higher levels of CSR disclosure as the capacity for monitoring increases. With regard to t he minimum number of members of any SSB, the AAOIFI standards have required at least three m embers. This is a common requirement in many Islamic banks such as Faysal Islamic Bank of S udan. The greater the number of members in a SSB, the greater the amount of monitoring, imp lying a greater level of compliance with Islamic laws and principles. The SSB would be able to allocate its functions across a larger group of members, allowing the SSB to review more aspects of the banks’ activities and hence ensure greater compliance. One aspect of this compliance i s more CSR disclosure. Further, synergies could also be present in boards with a large number of members pooling their ideas and perspectives to derive better applications of Islam ic law, particularly with regards to disclosure.

The AAOIFI recommends a number of different people from different professions to sit on the 16 SSB, including bankers, economists and full-time me mber(s) (AAOIFI, 2004). This allows for the implementation of diverse perspectives on the appli cation of the Shari’a. To enable this to happen, a large SSB would be required to represent these sectional professions. The above analysis suggests that the SSB size should have a p ositive relationship with CSR disclosure.

4.3.3. Cross memberships Cross memberships of SSB members may also lead to h igher disclosure of CSR information (Dahya, Lonie and Power , 1996). The literature suggests that cross-director ships increase transparency for two reasons: firstly, members with cross-directorships can make comparisons from knowledge gained in other companies; and secon dly “decisions at one board become part of the raw material for decisions at other boards” (Ha niffa and Cooke, 2002, p. 321). SSB members with cross-memberships will be exposed to more disc ussions about the application of Islamic law in banking. This increased experience should enhanc e their knowledge about the application of Islamic principles to corporate reporting and in pa rticular to CSR disclosure. 4.3.4. Secular educational qualifications In the extant literature, the director’s education has been proposed to influence the level of disclosure. Hambrick and Mason (1984) indicated tha t the more educated the director, the more likely he/she is to adopt innovative activities and accept ambiguity. The level of education of the SSB members may influence the level of CSR disclosu re. Bakar states that “ideally a Shari’a adviser (board member) must be able to understand n ot only Shari’a issues but also issues pertaining to law and economics, because such issue s in many cases are overlapping” (2002, p.

79). SSB members usually comprise scholars of Islam ic law who may not be highly educated in secular studies. SSB members who are not highly edu cated may be undermined in their abilities as Islamic law scholars to fully apply theoretical Islamic laws and principles because of their lack of practical commercial knowledge (Bakar, 2002; Bok hari, 2002). Hence, scholars with a doctorate degree are arguably better informed of th e current implications of Islam for financial institutions, particularly with regards to CSR disc losure.

4.3.5. Reputable scholars Some Shari’a scholars have a significant amount of tacit knowled ge about the application of Islamic law. However, they may not have formally re cognised qualifications from secular educational institutions. Hussain and Mallin (2003) report that the factors influencing the appointment of directors in Bahraini companies are relevant skills, business experience and 17 reputation. Following that reasoning, it is expecte d that reputation is a proxy for industry knowledge and hence reputable scholars are more lik ely to understand the current implications of Islamic banking, particularly with regards to discl osure. Hence, reputable scholars are more likely to emphasize CSR activities and the subsequent disc losure of CSR information. 4.3.6. IG-Score A number of previous studies have combined corporat e governance factors into an index which attempts to capture their aggregate effect. Gompers , Ishii and Metrick (2001) and Hanlon, Rajgopal and Shevlin (2003) combine a number of variables proxying for g overnance factors to produce a g-score. Applying the same reasoning, this study develops an Islamic governance score (IG-SCORE). In particular, a score is constructed b ased on the existence and characteristics of the SSB as detailed above. The score sums the value of the dichotomous characteristics of the board, namely the existence of the SSB, the number of boar d members, the existence of members with cross-memberships, the existence of members with do ctoral qualifications and the existence of reputable scholars presiding on the board. The meth od is elaborated in the research design. A positive relationship is expected between the Islam ic governance score (IG-SCORE) and the level of CSR disclosure presented by an Islamic bank. H 3: There is a positive association between the Islam ic governance scores obtained and the level of CSR disclosure presented by Islamic banks in the ir annual reports. 4.4. Hypothesis: Ownership Structure 4.4.1. Investment Account Holders’ (IAH) rights The structure of ownership also determines the leve l of monitoring and thereby the level of disclosure (Jensen and Meckling, 1976). A number of prior studies look at the effect of ownership structure on voluntary disclosure (Ruland, Tung and George, 1990; Eng and Mak, 2003; El- Gazzar, 1998; Mitchell, Chia and Loh, 1995; McKinno n and Dalimunthe, 1993; Schadewitz and Blevins, 1998). As elaborated above, Islamic investors determine th e extent of compliance with Islamic principles and consequently the level of CSR disclosures. Isl amic investors are more likely to invest their funds as IAH rather than as shareholders since Isla mic investors are primarily interested in the services that Islamic banks offer, rather than shar e ownership of the Islamic banks per se. Further, investment accounts with Islamic banks are generall y more accessible than shares of Islamic 18 banks. In the context of SSB reports, Karim states that: …If a negative report were to be published, it woul d be expected that this group (IAH) would begin to doubt management’s commitment to Isl am. Such a belief would have a detrimental impact on the bank since these clients are likely to refrain from dealing with the bank (1990, p. 41). While the IAH do not have any formal voting rights, they nevertheless influence the level of monitoring of management ‘vicariously’ through shar eholders (Archer et al., 1998). This is due to the fact that the profits of shareholders are deter mined by the profits earned through the utilisation of IAH funds. If the IAH are more interested than the shareholder s in the bank’s compliance with Islamic laws and principles, then the relative influence of the IAH will determine the extent to which the bank complies with Islamic laws and principles and conse quently the level of disclosure presented by the bank. This suggests that CSR disclosure is posi tively related to the relative size of IAH funds as a proportion of shareholder funds. H 4: There is a positive association between the propo rtion of IAH funds to shareholder funds and the level of CSR disclosure presented by Islamic ba nks in their annual reports.

5. Data and Research Design 5.1. Sample and Data Description In order to test the hypotheses, annual reports [14 ] of fully fledged Islamic banks from a number of countries were collected for the years ending 2002 or 2003. The collection o f two years of annual reports improved the chances of achieving a larger sample size as a number of the banks had not published their 2003 annual report at the t ime of data collection. Islamic banks are classified as any financial inter mediary that claims to operate according to the laws and principles of Islam (IAIB, 2001). Therefor e, the sample of Islamic banks contains service banks, investment banks, mortgage companies and leasing companies operating as either publicly listed companies, private companies or as government owned institutions. The structure of the banks and the types of services they offer a re not significant in determining CSR disclosure, as the expectation to disclose CSR info rmation is based on the fact that the banks 19 claim to be Islamic. Hence, mission statements and/ or information on their websites are used to verify the banks’ claim to operate according to Isl amic laws and principles.

The initial sample population for the study compris ed all Islamic banks in the world derived from a comprehensive list compiled by Archer and Karim ( 2002). These were cross-checked with the International Directory of Islamic Banks and Instit utions (2000) issued by the Institute of Islamic Banking and Insurance in London and the relevant st ock exchanges. Approximately 187 fully fledged Islamic banks in 29 countries were identifi ed. The annual reports for Islamic banks were primarily obtained from their websites and the rele vant stock exchange web-sites. Approximately 33 banks’ annual reports were collected in this man ner. To enlarge the sample size, another 48 banks were identified in the sample for which maili ng addresses were available. Each bank was sent a letter requesting their 2002 and 2003 annual reports. In total, 14 Islamic banks’ annual reports were received giving an overall response ra te of 29.8%. This is somewhat lower than the response rates achieved for studies in which Islami c banks annual reports were used. Maali et al.

(2003) achieved a response rate of 37.5% with an in itial sample of 88 banks. The final sample consists of 47 banks from 14 countries with the dis tribution as follows: Bahrain (6), Bangladesh (5), Egypt (1), Iran (4), Jordan (2), Kuwait (5), M alaysia (2), Pakistan (8), Qatar (2), Saudi Arabia (5), Sudan (1), Turkey (2), United Arab Emirates (3 ), and Yemen (1). 5.2. Research Design Ordinary least squares (OLS) regression is used to examine the relationship between CSR disclosure and the explanatory variables.[15] The r egression equation is as follows:

CSRDIS = + 1PRCL + 2MUSPOP + 3 IG-SCORE + 4IAH + 5SIZE + CSRDIS is the CSR disclosure index score of the ban k. Table 1 provides a summary of the operationalisation of the independent variables. TABLE I HERE 5.2.1. Dependent Variable The dependent variable was constructed using an ind ex of expected CSR disclosure of Islamic banks operating in Muslim countries. The index was derived from Maali et al. (2003) and contains 32 items.[16] Following Inchausti (1997) a nd Maali et al. (2003), each item in the disclosure index was given the same weight (Barrett 1997). Items disclosed were given a weight 20 of 1 while undisclosed items were weighted 0. The a nnual report of each bank was reviewed and a judgment made by the author as to which items wer e relevant to each bank. Irrelevant items were not considered as part of the overall score fo r the respective bank. For instance, some banks are required by law to pay Zakah, while others are not. Hence, the CSR disclosure s core was constructed as a ratio of the actual score achieved by the bank to the maximum possible value for each bank from the 32 disclosure items (Maali et al., 2003).

5.2.2. Independent Variables The independent variables are categorized into thre e groups: socio-political context, corporate governance and control. Since a number of the varia bles were never considered in previously published research on disclosure and particularly C SR disclosure, the appropriateness of these variables was discussed with and confirmed by a man aging director and a Shari’a supervisory board member of an Islamic bank in Australia. 5.2.3. Control Variable Firm size (SIZE) – Log of total assets. This measur e has been consistently found to influence CSR disclosure (Belkaoui and Karpik, 1989; Roberts, 1992; Williams, 1999; Patten, 1991). The arguments for larger firms disclosing higher levels of CSR flow from a number of different theoretical perspectives. Agency and positive accou nting theories predict that managers use CSR disclosure as part of their overall strategy to red uce agency costs and in particular political costs (Watts & Zimmerman, 1978). Large firms are more pol itically visible and therefore disclose more information. The legitimacy theory argument is that the more social exposure a firm receives by being larger, the greater the need to legitimize it s existence to its relevant publics (Patten, 1991). Hence, a significant positive relation is expected between firm size and voluntary disclosure. Alternative definitions of explanatory variables There are a number of alternative definitions for t he explanatory variables stated above.

Therefore, sensitivity testing is carried out, the results of which are reported subsequent to the main regression model. 6. Analysis of Results 6.1. Descriptive statistics The mean CSR disclosure for the 47 banks in the sam ple is approximately 16.8% of expected disclosure, representing an increase from Maali et al. (2003) who reported a mean disclosure of 21 13.3%. This is perhaps due to the fact that there a re more banks in the sample used in this study.

The maximum reported disclosure was 48.3% of expect ed disclosure, still falling significantly below expectations of the bank’s Islamic stakeholde rs or any Islamic society. A number of the banks had no CSR disclosures. The Pearson parametric and Spearman non-parametric correlation tests indicate that CSR disclosure is significantly related to size (SIZE), the Islamic governance score (IG-SCORE) and investment account holders’ rights (IAH). Size is a lso significantly related to the Islamic governance score (IG-SCORE) and investment account holders’ rights (IAH). This is intuitively appropriate since larger banks would have better mo nitoring mechanisms in place and would be funded to a greater extent by investment account ho lders rather than shareholders. The proportion of Muslim population (MUSPOP) is significantly rela ted to the level of political rights and civil liberties (PRCL) in a country. 6.2. Main regression model Table 2 shows the results of the multivariate regre ssion analysis. The model is statistically significant at the 1% level and has an adjusted R-square of approximately 0.393, indicating that the model explains a significant amount of the vari ation in disclosure. The control variable SIZE (log total assets) is not statistically significant . This is inconsistent with prior studies which fin d size to be significantly related to CSR disclosure (Williams, 1999). This may be due partly to the fact that the study is cross-national and log total assets is not a relative measure which captures the differences in relative size between countries. TABLE II HERE The political rights and civil liberties (PRCL) var iable is in the predicted direction. However, it is not statistically significant ( p-value=0.917), which is inconsistent with prior stu dies on disclosure and CSR disclosure (Archambault and Archambault, 20 03; Williams, 1999). The proportion of Muslim population (MUSPOP) variable is in the predi cted direction and statistically significant at the 5% level ( p-value=0.027). This indicates that the size of the relativ e publics’ proxying for their ability to influence and apply pressure on Is lamic banks is a significant factor in determining the level of CSR disclosure presented by these bank s. One of the corporate governance factors is signific ant at or below the 5% level. The combination 22 of the SSB characteristics (IG-SCORE) is highly sig nificant at the 1% level and in the predicted direction ( p-value=0.001) while the investment account holders’ rights (IAH) variable is in the predicted direction but only marginally significant at the 10% level (p-value=0.061). The highly significant IG- SCORE indicates that the SSB and it s characteristics are important in influencing CSR disclosures. The insignificance of the IAH vari able is contrary to H 4. Overall, the preliminary results suggest that the r elevant publics and monitoring mechanisms such as the SSB increase the level of CSR disclosure by Islamic banks.

6.3. Sensitivity testing Since there is a degree of arbitrariness in selecti ng the proxies for the model variables, sensitivity tests are conducted to ensure the robustness of the results and to further enhance the empirical model. Replacing the measure of political rights an d civil liberties with a relatively simpler index ranking countries from ‘free’ (1) to ‘partially fre e’ (2) to ‘not free’ (3), increases the adjusted R- square from 0.393 in the original model to 0.452 an d the variable PRCL becomes significant at the 5% level ( p-value=0.042). Since size was found to be insignific ant contrary to expectations, a relative measure of size was used by dividing the t otal assets of the bank by the gross domestic product of countries and then logging the value. Wh ile this was done to adjust for the relative differences in size between countries, no significa nt results were found. Two alternative measures of IAH were utilised. The first measure utilised was the ratio of total investment account holders’ fund to total assets. T he IAH variable increases in significance to the 5% level ( p-value=0.029) and the adjusted R-square increases from 0.393 in the original model t o 0.411. The second measure utilised is the ratio of the total investment account holders’ funds to total shareholders’ equity. The IAH variable loses its significance using this measure and the R- square decreases from 0.393 in the original model t o 0.348. There are no significant changes in the other variables. Further, to ensure that the si gnificant correlation between IAH rights and size is not introducing noise into the regression, tests were conducted excluding either IAH rights or size from the regression model. While each of the v ariables increase in significance when one of them are excluded from the model, excluding IAH fro m the model reduces the total R-square more than excluding size, indicating that the IAH c ontributes more significantly in explaining disclosure than SIZE. Hence, it may be inferred tha t the significance of the SIZE variable in the test conducted without IAH was only capturing the e ffect of an omitted correlated variable.

23 Since the IG-SCORE attempts to combine the effect o f five difference sub-variables, it is not informative as to the effect of each sub-variable. Tests are run excluding each one of the sub- variables one at a time from the IG-SCORE to review the consequent effect on the significance of the IG-SCORE and the adjusted R-square of the total model. The IG-SCORE variable dr ops in significance and the R-square is vastly reduced at every instance a sub-va riable is dropped except for when the sub-variable NUM is dropped indicating that the number of scholars is not an important factor in contributing to higher CSR disc losure.

TAKE IN TABLE III 6.4. Optimal regression model Based on the sensitivity analysis conducted above, an optimal regression model is formed utilising two alternative measures derived from the sensitivity testing. The original political rights and civil liberties variable (PRCL) is replaced by the measure elaborated earlier. The IAH variable is replaced by the alternative IAH rights’ measure elaborated earlie: the ratio of total IAH funds to total assets. The results are reported in table 3. The model provides an adjusted R- square of approximately 0.471 recording a significa nt increase from the original model which had an adjusted R-square of approximately 0.393. All variables except SIZE are significant with PRCL ( p-value=0.034) and IAH ( p-value=0.039) being significant at the 5% level and MUSPOP ( p- value=0.005) and IG-SCORE ( p-value=0.007) being significant at the 1% level. 7. Conclusion Islamic banks have grown in size and significance i n the past three decades. In line with Islamic principles, Islamic banks should fulfil an ethical role inherent in their character as an ‘Islamic’ bank. The objective of this study was to measure th e CSR disclosure levels of Islamic banks and subsequently ascertain the likely determinants of t hat disclosure. Contrary to expectations of full disclosure and accountability, it was found that th e majority of Islamic banks disclose significantly less than expected, with apparent dif ferences in the levels of disclosure. To explain these differences a number of hypotheses were deriv ed and subsequently tested. The significance of the variable PRCL suggests that the extent of political and civil repression influences the level of CSR disclosure by Islamic b anks. This gives weight to the view consistent with Williams (1999) that organisations operating i n relatively repressed societies face lower 24 social expectations while organisations operating i n relatively open communities need to provide further justification to legitimize their existence . The significance of the variable MUSPOP lends weight to the hypothesis that the level of CSR acti vities and consequently disclosure will depend on the extent of influence that the ‘relevant publi cs’ have on the organisation. The combined results of these two hypotheses demonstrate the the oretical significance of applying ‘levels of resolution of perception’ to understand the complex interactions between organisations and society in the broader socio-political environment. The significance of the corporate governance variab les lends support to the view that Islamic banks are also driven by economic realities. In par ticular, the existence of Shari’a board members with cross memberships, doctorate qualifications an d/or international repute results in greater monitoring and hence greater compliance with Islami c laws and principles, an output of which is higher levels of CSR disclosure. This implies that skilled scholars are required to decipher Islamic law to apply it to modern Islamic banks, pa rticularly with regards to CSR disclosure.

Further, the significance of IAH rights in influenc ing CSR disclosures imply that Islamic banks disclose CSR in order to bond their activities to t heir ‘Islamic’ investors. Overall, the results suggest that there are a numbe r of factors which contemporaneously influence CSR disclosure of Islamic banks including socio-pol itical pressures and economic incentives. The results here have a number of potential policy impl ications for Islamic banks and regulators.

While socio-political factors may restrain the leve l of CSR disclosure presented by Islamic banks, increasing the level of monitoring within the banks , such as the installation of a SSB can counter that and lead to greater CSR disclosure. As highlig hted in the literature, these results further emphasize the need for Islamic banks to invest more in monitoring mechanisms, such as greater training of SSB members to increase the confidence of the Islamic investors and society (Bakar, 2002). Furthermore, these results also give weight to the contention that uniform accounting standards and Shari’a rulings across the globe need to be implemented in order to ensure a uniform level of disclosure by Islamic banks (Dudle y, 2004; Karim, 2001).

A number of limitations, however, exist about the r esults of this paper. The sample size of 47 observations is one such limitation. While a number of tests were conducted to ensure that the sample represented a normal distribution, there is the likelihood that the results represent a random occurrence or are biased towards the variabl es. Further, this study only uses Islamic banks’ annual reports to measure CSR disclosure. It is likely that Islamic banks also use other 25 forms of media to communicate with investors and gr eater society such as company websites, press releases, annual general meetings, special bo oklets and pamphlets detailing their contribution to society. Using a larger sample and a larger set of information or variables about the operating environment and individual characteri stics of Islamic banks, future research can attempt to further generalise these results and enh ance knowledge about the affect of other factors not theorised in this study, particularly regarding transparency and social responsibility and the ensuing CSR disclosures. Lastly, future research co uld also attempt to generalise the results of this study about the potential impact of the socio- political environment in determining the level of CSR disclosure using conventional organisations ope rating in diverse cross-national contexts rather than solely Islamic banks. 26 References AAOIFI (2003), Accounting, Auditing & Governance St andards for Islamic Financial Institutions, 4th Edition, AAOIFI, Bahrain. Akacem, M and Gilliam, L. (2002), “Principles of Is lamic banking: Debt versus equity financing”, Middle East Policy, Washington, Vol. 9, No. 1, pp. 124-139.

Adams, C.A. (2002), “Internal organisational factor s influencing corporate social and ethical reporting: Beyond current theorising”, Accounting, auditing and accountability, Vol. 15, No. 2, pp. 223-250. Aggarwal, R. K. and Youssef, T. (2000), “Islamic ba nks and investment financing”, Journal of Money, Credit and Banking, Vol. 32, pp. 93-120. Ali, A.Y. (1989), The Holy Qur’an: Text, Translatio n and Commentary, Amana Corporation, Maryland, Washington. Archambault, J.J. and Archambault, M.E. (2003), “A multinational test of determinants of corporate disclosure”, The International Journal of Accounting Vol. 38, pp. 173-194.

Archer, S. and Karim, R.A.A. (2002), Islamic Financ e: Innovation and Growth, Euromoney Books and AAOIFI, London. Archer, S.; Karim, R.A.A. and Al-Deehani, T. (1998) , “Financial contracting, governance structures and the accounting regulation of Islamic banks: An analysis of agency theory and transaction cost economics”, Journal of Management and Governance, Vol.2 pp. 149-170.

Askary, S. (2001), “The Influence of Islamic Cultur e on the Accounting Values and Practices of Muslim Countries” Doctoral Dissertation, University of Newcastle.

Bahrain Monetary Agency (2004), Islamic finance hig hlighted, Islamic Finance Review April 2004, Issue 5 Bakar, M.D. (2002), “The Shari’a supervisory board and issues of Shari’a rulings and their harmonisation in Islamic banking and finance” in Is lamic Finance: Innovation and Growth, eds S.

Archer and R. A. A. Karim, Euromoney Books and AAOI FI, London, pp. 74-89.

Barrett, M. (1997), “Financial Reporting Practices: Disclosure and comprehensiveness in an international setting”, Journal of Accounting Resea rch, Vol. 14, pp. 10-26.

Baydoun, N. and Willett, R. (2000), “Islamic Corpor ate Reports”, Abacus, Vol. 36, No. 1, pp. 71- 90. Belkaoui, A. and Karpik, P.G. (1989), “Determinants of the Corporate Decision to Disclose Social Information”, Accounting, Auditing and Accou ntability Journal, Vol. 2, No.1, pp. 36-51.

Bokhari, F. (2002), “Lack of experts threatens Isla mic investing: Banks and funds wooing Muslims need scholars versed in religion and financ e”, Financial Times, London, May 29, p.24.

Bowman, E.H. and Haire, M. (1976), “Social Impact D isclosure and Corporate Annual Reports”, Accounting, Organizations and Society, Vol. 1, No. 1, pp. 11-21. 27 Campbell, D.J (2000) “Legitimacy theory or manageri al reality construction? Corporate social disclosure in Marks and Spencer plc corporate repor ts 1969-1997”, Accounting Forum Vol 24, No.1, pp. 80-101. Campbell, D., Craven, B. and Shrives, P. (2002), “V oluntary social reporting in three FTSE sectors: a comment on perception and legitimacy” Ac counting, Auditing and Accountability Journal, Vol. 16, No. 4, pp. 558-581. Central Intelligency Agency (2004), CIA World Factb ook, Central Intelligence Agency, viewed 31 July, 2004 Cowen, S.S., Ferreri, L.B. and Parker, L.D. (1987), “The impact of corporate characteristics on social responsibility disclosure: A typology and fr equency-based analysis”, Accounting, Organizations and Society, Vol. 12, No. 2, pp. 111- 122.

Dahya, J., Lonie, A.A. and Power, D.M. (1996), “The case for separating the roles of chairman and CEO: An analysis of stock market and accounting data”, Corporate Governance: An International Review, Vol. 4, No. 2, pp. 52-68. Deegan, C. (2002), “The legitimising effect of soci al and environmental disclosures- a theoretical foundation”, Accounting, Auditing and Accountabilit y Journal, Vol. 15, No. 3, pp. 282-312.

DeLorenzo, Y.T. (2002), “The religious foundations of Islamic finance” in Islamic Finance:

Innovation and Growth, eds S. Archer and R. A. A. K arim, Euromoney Books and AAOIFI, London, pp. 9-28. Dowling, J. and Pfeffer, J. (1975), “Organizational legitimacy: social values and organisational behaviour”, Pacific Sociological Review, Vol. 18, N o. 1, pp. 122-136.

Dudley, N. (2004), “Islamic finance needs solid fou ndations” Euromoney, January 2004, Vol. 35, No. 7, pp. 38-43. Eng, L.L. and Mak, Y.T. (2003), “Corporate Governan ce and Voluntary Disclosure”, Journal of Accounting and Public Policy, Vol. 22, No. 4, pp. 3 25-345.

El-Gazzar, S.M. (1998) “Predisclosure information a nd institutional ownership: A cross-sectional examination of market revaluations during earnings announcement periods”, The Accounting Review, Vol. 7, No. 1, pp. 119-129. Fama, E.F. and Jensen, M.C. (1983) “Agency Problems and Residual Claims” Journal of Law and Economics, Vol. 26, No. 3, pp. 327-351. Freedom House (2004), ‘Freedom in the world’ index, Freedom House, viewed 31 July 2004, Gambling, T.E. and Karim, R.A.A. (1986), “Islam and 'Social Accounting'”, Journal of Business Finance and Accounting, Vol. 13, No.1, pp. 39-49. Gambling, T.E. and Karim, R.A.A. (1991), Business a nd Accounting Ethics in Islam, 1 st Edition, Mansell Publishing Limited, London. 28 Gastil, R.D. (1981), Freedom in the World: Politica l Rights and Civil Liberties Freedom House, New York. Gompers, P., Ishii, J. and Metrick, A. (2001), Corp orate governance and stock prices, NBER working paper 8449. Gray, R., Owen, D. and Maunders, K. (1987) Corporat e Social Reporting: Accounting and Accountability, Prentice-Hall, London. Gray, R., Kouhy, R. and Lavers, S. (1995), “Corpora te social and environmental reporting: A review of the literature and a longitudinal study o f UK disclosure”, Accounting, Auditing and Accountability, Vol. 8, No. 2, pp. 47-77. Gray, R., Owen, D. and Adams, C. (1996), Accounting and Accountability: Changes and Challenges in Corporate and Social Reporting, Prent ice Hall, London.

Gray, S.J. (1988), “Towards a theory of cultural in fluence on the development of accounting systems internationally”, Abacus, Vol. 24, No.1, pp .1-15. Hambrick, D.C. and Mason, P.A. (1984), “Upper Echel ons: The Organization as a Reflection of Its Top Managers”, Academy of Management Review, Vo l. 9, No. 2, pp. 193-206.

Hamid, S., Craig, R. and Clarke, F. (1993), “Religi on: A Confounding Cultural Element in the International Harmonization of Accounting”, Abacus, Vol. 29, No.2, p. 131-149.

Hanlon, M., Rajgopal, S. and Shevlin, T. (2003), “A re executive stock options associated with future earnings?” Journal of Accounting and Economi cs Vol. 36, No.1-3, pp.3-43.

Haniffa, R. (2001), Social Responsibility Disclosur e: An Islamic Perspective, Discussion Paper, University of Exeter, UK. Haniffa, R.M and Cooke, T.E. (2002), “Culture, Corp orate Governance and Disclosure in Malaysian Corporations”, Abacus, Vol. 38, No. 3, pp . 317-350.

Haniffa, R., and Hudaib, M. (2007). Exploring the e thical identity of Islamic banks via communication in annual reports. Journal of Business Ethics, 76, 97-116 Hill, J.G. (1999), “Deconstructing Sunbeam-Contempo rary issues in corporate governance”, University of Cincinnati Law Review, Vol. 67, pp. 1 099-1127.

Hines, R.D. (1988), “Financial Accounting: In Commu nicating Reality, We Construct Reality”, Accounting, Organizations and Society, Vol. 13, No. 3, pp. 251-261.

Hofstede, G. (1980), Culture's Consequences: Intern ational differences in work related values, Sage Publications, Beverley Hills, California. Ho, S.S.M. and Wong, K.S. (2001), “A study of the r elationship between corporate governance structures and the extent of voluntary disclosure”, Journal of International Accounting, Auditing & Taxation, Vol. 10, pp.139-156. 29 Hussain, S.H. and Mallin, C. (2003), “The Dynamics of Corporate Governance in Bahrain:

structure, responsibilities and operation of corpor ate boards”, Corporate Governance: An International Review, Vol. 11, No. 3, pp. 249-262. Inchausti, B. (1997), “The influence of company cha racteristics and accounting regulation on information disclosed by Spanish firms”, European A ccounting Review, Vol. 6, No.1, pp.45-68.

Institute of Islamic Banking and Insurance (IIBI) ( 2000), International Directory of Islamic banks and Institutions, IIBI, London. International Association of Islamic Banks (IAIB) ( 2001), Directory of Islamic Banks and Financial Institutions, IAIB, Jeddah. Jensen, M.C. and Meckling, W.H. (1976), “Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure”, Journal of Financia l Economics, Vol. 3, pp. 305-360.

Karim, R.A.A. (1990), “The independence of religiou s and external auditors: The case of Islamic banks”, Accounting, Auditing, and Accountability Jo urnal Vol. 3, No.3, pp. 34-44. Karim, R.A.A. (1995), “The Nature and Rationale of a Conceptual Framework for Financial Reporting by Islamic Banks”, Accounting and Busines s Research, Vol. 25, No. 100, pp. 285-300.

Karim, R.A.A. (2001), “International accounting har monization, banking regulation and Islamic banks”, International Journal of Accounting, Vol. 3 6, No. 2, pp. 169-196.

Lewis, M.K. (2001), “Islam and Accounting” Accounti ng Forum, Vol. 25, No.2, pp 103-127.

Lindblom, C.K. (1994), “The implications of organis ational legitimacy for corporate social performance and disclosure”, cited in Deegan 2002. Maali, B., Casson, P. and Napier, C. (2003), Social Reporting by Islamic Banks, Discussion Papers in Accounting and Finance, University of Sou thampton.

McKinnon, J.L. and Dalimunthe, L. (1993), “Voluntar y disclosure of segment information by Australian diversified companies”, Accounting and F inance, Vol. 33, No. 1, pp. 33-50.

Metwally, S. (1992), “The Aggregate Balance Sheet a nd the Results of Transactions and Financial Indicators for Islamic Banks and Financia l Institutions”, Journal of Islamic Banking and Finance, pp. 7-61. Mitchell, J.D., Chia, C.W.L and Loh, A.S. (1995), “ Voluntary disclosure of segment information:

Further Australian evidence”, Accounting and Financ e, Vol. 35, No. 2, pp. 1-16.

Newson, M. and Deegan, C. (2002), “Global expectati ons and their association with corporate social disclosure practices in Australia, Singapore and South Korea”, The International Journal of Accounting, Vol. 37, pp. 183-213. Patten, D.M. (1991), “Exposure, Legitimacy and Soci al Disclosure” Journal of Accounting and Public Policy, Vol. 10, pp. 297-308. Perera, M.H.B. (1989), “Towards a framework to anal yze the impact of culture on accounting”, 30 International Journal of Accounting, Vol. 24, No. 1 , pp. 42-57.

Ramanathan, K.V. (1976), “Toward a theory for corpo rate social accounting”, Accounting Review, Vol. 51, No.3, pp. 516-529. Roberts, R.W. (1992), “Determinants of corporate so cial responsibility disclosure: An application of stakeholder theory”, Accounting, Organizations a nd Society, Vol. 17, No.6, pp. 595-612.

Ruland, W., Tung, S. and George, N.E. (1990), “Fact ors associated with the disclosure of managers’ forecasts”, The Accounting Review, Vol. 6 5, No. 3, pp.710-721.

Sadeghzadeh, A. (1995), Social responsibility accou nting, sustainability accounting and Islam, Ph.D. Dissertation, University of Wollongong. Schadewitz, H.J. and Blevins, D.R. (1998), “Major d eterminants of interim disclosure in an emerging market”, American Business Review, Vol. 16 , No. 1, pp. 41-55.

Sharani, U.M. (2004), “Corporate social responsibil ity underlines values propagated by Islam” Bernama, Kuala Lumpur, June 21, p. 1. United Nations Development Program (2004), 2004 Hum an Development Report, United Nations Development Program, viewed 31 July, 2004, Usmani, M.T. (2002), An introduction to Islamic fin ance Arab and Islamic Law Series, Kluwer Law International, Amsterdam. Watts. R. and Zimmerman, J. (1978), “Towards a posi tive theory of the determination of accounting standards”, The Accounting Review, Janua ry, pp. 112-134. Williams, S.M. (1999), “Voluntary environmental and social accounting disclosure practices in the Asia-Pacific region: An international empirical test of political economy theory”, The International Journal of Accounting Vol. 34, No.2, pp. 209-238.

Wilmhurst, T.D. and Frost, G. R. (2000), “Corporate environmental reporting: A test of legitimacy theory”, Accounting, Auditing and Accoun tability, Vol. 13, No. 1, pp. 10-25.

Yunus, K. (2004), “Investment in Islamic funds soar s” Business Times, Kuala Lumpur, June 23, p. 2. Zaher, T.S and Hassan, M.K. (2001), “A Comparative Literature Survey of Islamic Finance and Banking” Financial Markets, Institutions and Instru ments, Vol. 10, No.4, pp. 155-199. 31 Footnotes 1. In this paper, Islamic Banks are synonymous with Islamic Financial Institutions namely commercial banks, investment banks, and Modaraba co mpanies. 2. The Shari’a is the law of Islam derived from the Muslim holy bo ok (Qur’an) , the sayings and deeds of the Prophet Muhammad (Sunnah), consensus ( ijma), reasoning by analogy (qiyas), and public interest (maslaha). 3. Riba is translated strictly as usury; however Islamic sc holars equate it as being equivalent to interest. 4. Ali 1989, The Holy Qur’an 2:116; 2:107; 3:189; 5:17; 18:40.

5. Ali 1989, The Holy Qur’an 33:72. Vicegerency ( istikhlaf) or trusteeship is similar to the concept of Stewardship in Christianity.

6. Ali 1989 The Holy Qur’an 102:8.

7. Ali 1989, The Holy Qur’an 22:40.

8. Dowling and Pfeffer (1975) and Lindblom (1994) e laborate on the number of strategies that organisations undertake to maintain or create congr uence between social values implied by the organisation’s operations, and the values embraced by society, all of which require disclosures and all of which may not be genuine attempts at soc ial responsibility.

9. Level of resolution of perception refers to the varying levels of perception through which the social system can be analysed (Gray et al., 1995). For brevity’s sake, we have concentrated on two levels of resolution, the broader political eco nomy within which the organisations interact and the specific relevant ‘publics’ with whom the organisations interact. 10. IAH funds usually comprise a mix of unrestricted and restricted mudarabah contracts. The IAH who use unrestricted mudarabah authorise their mudarib to invest their funds at its discretion including co-mingling the IAH’s funds wi th those of the shareholders. The restricted mudarabah IAH, on the other hand specify to the bank the type of investment in which their funds should be invested e.g. real estate, currenci es or leasing. 11. An Islamic bank is not allowed to invest in act ivities that are associated with gambling, alcohol, pork and generally encouraged to invest in social development activities. 12. Islamic investors invest in Islamic banks prima rily for religious reasons but also for economic reasons. 13. Economic investors invest in Islamic banks pure ly for economic reasons. 14. Annual reports may not be the only means by whi ch Islamic banks communicate their CSR information. They may utilize other means such as a dvertising, public relations and internet sites to convey social information. Notwithstanding, Gray et al. (1995) state that the annual report is the only document produced regularly to comply with regulatory requirements and more importantly is central to the organisations constru ction of its own external image. Hence, this study only considers the information disclosed in a nnual reports.

15. A number of tests were conducted to confirm the normality of the distribution. Normal probability plots and histograms of the distributio n of the dependent variable (CSRDIS) and scatter plots of standardized residuals against sta ndardized estimates of CSRDIS were reviewed on all models to verify that the normality and homo scedasticity assumptions were valid.

Multicollinearity tests are also conducted on all m odels. The variance inflation factors (VIF) range from 1-2 and eigenvalues and condition indice s were checked to ensure there was no excessive multicollinearity. However, caution must nonetheless be drawn in generalizing from these results as sample size may limit the signific ant findings in this study. Further, An expert statistician from the University of Technology, Syd ney was consulted to ensure that the data did not violate any assumptions that are usually sacrif iced with small sample sizes.

16. Refer to Maali et al. (2003) for details on the construction of the disc losure index.

32 FIGURE 1. Diagrammatic portrayal of the influences on CSR disclosure by Islamic banks \ FIGURE 2. Diagrammatic portrayal of the relationshi p between the organisation and greater society. SOCIO -POLITICAL CONTEXT Political Economy Relevant Publics Islamic Banks CSR DISCLOSURE CORPORATE GOVERNANCE Monitoring Mechanisms Ownership Structure ORGANISATION POLITICAL ECONOMY RELEVANT PUBLICS or STAKEHOLDERS RELEVANT PUBLICS or STAKEHOLDERS RELEVANT PUBLICS or STAKEHOLDERS 33 TABLE I: Summary of the source and operationalisation of independent variables Hypothesis Independent Variables Operationalisation Source of information Socio-political context- Political economy H1 PRCL Political rights and civil liberties Overall combined index scores of political rights and civil liberties based on work of Gastil (1978) for the given nation:1 (freedom) to 14 (repression) Freedom House ‘freedom in the world’ Index http://www.freedomhouse.org/ratings/allscore04.xls Socio-political context- Relevant publics H2 MUSPOP Proportion of Muslim population Ratio of Muslim population to total population CIA World Factbook, 2004 http://www.odci.gov/cia/publications/factbook/geos/ Corporate Governance- Shari’a Supervisory Board H3 IG-SCORE Islamic governance score Existence of SSB + Number of SSB members + Cross memberships + Doctorate qualification of SSB member + Reputable scholars on SSB Derived from sub-variables below SSB Existence of SSB Dichotomous; yes/no Bank Annual Report NUM Number of SSB members Dichotomous; 1 for banks with 7 or more members and 0 otherwise Bank Annual Report CROSS Cross Memberships Dichotomous; If any SSB member with cross- directorship: 1, otherwise: 0. Bank Annual Report PHD Doctoral qualification of SSB member Dichotomous: If any SSB member with doctorate qualification: 1, otherwise: 0 Bank Annual Report and website REP Reputable Scholars on SSB Dichotomous: If any reputable SSB member: 1; otherwise: 0. Reputable scholar is one that has a position in the SSB of the AAOIFI and at least two Shari’a board memberships. AAOIFI Standards and Bank Annual Report Corporate Governance- Ownership Structure H4 IAH Investment Account Holders Ratio of total amount of investment account funds to paid up capital in shareholders equity. Bank Annual Report Control SIZE Size Natural Log of Book value of assets translated into USD Bank Annual Report TABLE II: Regression analysis main model Table 2 displays the regression results for the determinants of CSR disclosure. The coefficients are based on the following equatio n: CSRDIS = + 1PRCL + 2MUSPOP + 3 IG-SCORE + 4IAH + 5SIZE + Independent Variables Predicted Sign Coefficients t-statistics p-value VIF Intercept -0.517 -2.457 0.018 PRCL - -0.001 -0.105 0.917 1.148 MUSPOP + 0.373 2.298 0.027 * 1.168 IG-SCORE + 0.046 3.625 0.001 ** 1.376 IAH + 0.003 1.927 0.061 1.362 SIZE + 0.023 1.258 0.215 1.591 Std Error 0.100 F value 6.950 Sig. F (p-value) 0.000 R 0.677 R2 0.459 Adjusted R2 0.393 ** Coefficient is significant at the 0.01 level (2-t ailed).

* Coefficient is significant at the 0.05 level (2-tailed). Variable Definitions:

PRCL Political rights and civil liberties index sco res from Freedom House ranging from 1 (highest amount of political rights and civil liber ties) to 14 (lowest amount of political rights and civil liberties). MUS POP Proportion of Muslim population to total population of the given nation.

IG-SCORE (SSB+NUM+CROSS+PHD+REP).

IAH Ratio of investment account holders' funds to paid- up capital.

SIZE Natural log of the bank’s total assets.

CSR DIS CSR disclosure index score of bank. IG-SCORE consists of:

SSB 1 for Islamic banks with Shari’a Supervisory boards and 0 otherwise. NUM 2 for Islamic banks with 7 or more SSB members and 1 otherwise.

CROSS 1 for Islamic banks with one or more cross-members and 0 otherwise.

PHD 1 for Islamic banks with one or more members w ith doctorate qualifications and 0 otherwise.

REP 1 for Islamic banks with one or more reputable scholars as members of the SSB and 0 otherwise.

35 TABLE III: Optimal regression model Table 3 displays the optimal regression model derived from the sensitivity tests conducted utilising alternative measures of political rights and civil liberties (PRCL) and IAH rights (IAH).

The coefficients are based on the following equatio n: CSRDIS = + 1PRCL + 2MUSPOP + 3 IG-SCORE + 4IAH + 5SIZE + Independent Variables Predicted Sign Coefficients t-statistics p-value VIF Intercept -0.412 -2.092 0.043 PRCL - -0.072 -2.187 0.034 * 1.448 MUSPOP + 0.481 3.091 0.004 ** 1.231 IG-SCORE + 0.035 2.823 0.007 ** 1.547 IAH + 0.096 2.136 0.039 * 1.333 SIZE + 0.020 1.166 0.250 1.629 Std Error 0.093 F value 9.187 Sig. F (p-value) 0.000 R 0.727 R2 0.528 Adjusted R2 0.471 ** Coefficient is significant at the 0.01 level (2-t ailed).

* Coefficient is significant at the 0.05 level (2-tailed). Variable Definitions:

PRCL Political rights and civil liberties index scores f rom Freedom House. 1: for countries classified free, 2: for countries classif ied partially free, 3: for countries classified not free. MUSPOP Proportion of Muslim population to total population of the given nation.

IG-SCORE (SSB+NUM+CROSS+PHD+REP).

IAH Ratio of investment account holders' funds to total assets.

SIZE Natural log of the bank’s total assets.

CSR DIS CSR disclosure index score of bank. IG-SCORE consists of:

SSB 1 for Islamic banks with Shari’a Supervisory boards and 0 otherwise. NUM 1 for Islamic banks with 7 or more SSB members and 0 otherwise.

CROSS 1 for Islamic banks with one or more cross-members and 0 otherwise.

PHD 1 for Islamic banks with one or more members w ith doctorate qualifications and 0 otherwise.

REP 1 for Islamic banks with one or more reputable scholars as members of the SSB and 0 otherwise.