Exploring the Ethical Identity of Islamic Banks
The population of our study consists of IBs operat- ing in the Arabian Gulf region, which encompasses Bahrain, Saudi Arabia, Qatar, Kuwait, Oman and the United Arab Emirates. Our sample was initially based on the list of Islamic financial insti- tutions on the website IBF NET (http://www. Islamic-finance.net/bank.html). There are 24 such institutions in the region (including investment banks and insurance providers) and our final sample consists of seven Islamic banks in four of those countries whose annual reports for the years 2002–2004 were available in an English version on the Internet. These banks are Abu Dhabi Islamic Bank (ADIB), Al-Baraka Bank (ABB), Al-Rajhi Bank (ARB), Bahrain Islamic Bank (BIB), Dubai Islamic Bank (DIB), Kuwait Fi- nance House (KFH) and Shamil Islamic Bank (SIB). To explore the communication practices of IBs in our sample, we used content analysis, which is a method of codifying the text (or content) of a piece of writing into various groups (or categories) depending on selected criteria (Krippendorff, 1980; Weber, 1988). An essential element of content analysis is the selection and development of categories into which content units can be classified. The categories and items were drawn mainly from our understanding of Islamic literature as discussed in the last section and the Islamic perspective of a social responsibility reporting framework suggested by Haniffa (2002). We also considered items required to be disclosed by the Accounting and Auditing Organisation for Islamic Financial Institutions (AAOIFI, 2001) and previous research in the area of social responsibility reporting (Cowen et al., 1987; Ernst and Ernst, 1978; Gray et al., 1995; Guthrie and Parker, 1989, 1990). Since the interest is to assess the degree of con- gruence between the ideal and communicated ethical identities, we designed our research instrument (checklist) to cover aspects of the five traits (themes) constituting the ideal ethical identity as suggested in the last section. The five themes were extended into eight dimensions. The first theme, underlying phi- losophy and values’, consists of two dimensions: vision and mission of the organisation and details of its board of directors and top management, including aspects of good corporate governance practice. The second and third themes, viz. interest-free product’ and engagement in only Islamically acceptable deals 12
(halal)’, are both addressed under the dimension products. Aspects of the fourth theme, focus on developmental and social goals’, are covered under four dimensions: zakah, charity and benevolent loans; employees; debtors; and community. The fifth theme, additional reviews by the Shari’ah Supervi- sory Board’, is addressed under the dimension SSB. We further identified constructs under each dimension based on prior literature including those that capture pictures and graphics, as they are powerful and highly effective methods of commu- nication (Beattie and Jones, 1992, 1994; Preston et al., 1996) and excluding them may be considered a limitation. The final ideal ethical identity checklist instrument consists of 5 themes, 8 dimensions and 78 constructs (see Appendix).
Two methodological issues related to content analysis are reliability and validity (Weber, 1985). Reliability problems may arise due to ambiguity of meanings or category definitions while validity problems are related to the "extent to which (a variable) measures the construct the investigator in- tends it to measure" (Weber, 1985, p. 15). Hence, precautionary measures have been taken to improve both reliability and validity. One way of improving reliability is to use multiple coders (Holsti, 1969) and in this study, two coders scored the research instrument and any problems and discrepancies that arose were discussed and resolved accordingly via a set of basic coding rules. To enhance validity, categories were carefully developed from the extant Islamic and social responsibility literature.
The approach to scoring items is essentially dichotomous in that an item in the research instru- ment scores one if communicated and zero if it is not (see Haniffa and Cooke, 2002), although no penalty is imposed if the item is considered irrelevant.13 To ensure that judgement of relevance is not biased, the entire annual report is read before any decision is made (Cooke, 1996). The approach to scoring is additive and equally weighted.14 To assess the com- municated and ideal ethical identities, we then expressed the scores in the form of an index, which we termed Ethical Identity Index (EII), as follows:
E I I j ¼