Higher Ethical Objective (Maqasid al-Shari’ah) Augmented Framework for Islamic Banks: Assessing Ethical Performance and Exploring Its Determinants

In contrast to classical views on the performance assessment of financial institutions which are based mainly on financial and economic measures, a performance view in light of the maqasid al-Shari’ah is multidimensional and has a wider scope with the purpose of enhancing and sustaining human and societal well-being (Bedoui and Mansour 2014). However, Mohammed et al. (2008) argue that the failure of academic scholars to construct the multidimensional objectives for IBs has left no choice but to adopt the conventional unidimensional benchmarks of financial indicators to evaluate the performance of IBs. Therefore, performance assessment models and criteria are important to link IB activities with maqasid al-Shari’ah objectives in the context of IME.

To fill the gap in performance measurement-related theoretical frameworks, several studies have been conducted. For example, Hameed et al. (2004) developed the ‘Islamicity Disclosure Index’, which consists of Shari’ah compliance, CG and social/environmental indicators, and some financial ratios. In addition, Mohammed et al. (2008) constructed performance ratios for assessing the achievements of IBs in nine dimensions and ultimately covered three objectives including educating individuals, establishing justice and promoting welfare. Furthermore, Ngalim et al. (2015) has developed an ‘Islamic Vision for Development’ based indicators for the assessment of the maqasid al-Shari’ah of IBs in light of the five Ghazalian maqasid with supplementary corollaries. In addition to maqasid al-Shari’ah performance measurement frameworks, some studies analysed the ethical identity of IBs through CSR-based disclosure analysis based on annual reports, such as Haniffa and Hudaib (2007) and Belal et al. (2014). Moreover, Kasri and Ahmed (2015), in their attempt to develop operational indicators for a maqasid-based multidimensional poverty measure, included merely 21 items over five dimensions. In another attempt, Amir-Ud-Din (2014:25), using the fiqhi classification, “developed a Maqasid al-Shari’ah Index (MSI) corresponding to three levels of necessities, complementarities and embellishments”, while this study is mainly related to expressed morality in terms of IME rather than ‘rational-legal reasoning’ provided by fiqh. In developing a structural model for human development, Mili (2014) utilises Ghazalian maqasid by adding a further dimension that is ‘human well-being’ resulting into six dimensions and 15 elements. In addition, with the objective of developing policy oriented maqasid index, Hasan and Ali (2018) constructs a maqasid al-Shari’ah deprivation perception index (MSDPI) for the Muslim countries utilising five dimensional Ghazalian maqasid index with 26 deprivation-related consequences to evidence the impact of the lack of such maqasid on human well-being. Despite the resemblances and utilising similar constructs, the methodological frameworks and measurements used in these attempts do not cover all the aspects of maqasid and they remain within the methodology and framework of CSRs applied to conventional banks. Therefore, a comprehensive and integrated maqasid index should be developed as attempted by this study, which covers an important gap in the literature.

In terms of the evaluation of actual ethical performance, several empirical studies provided evidence that while IBs achieve significant financial performance, they perform poorly in ethical and social dimensions targeted by IME. Asutay (2012) concluded that, based on empirical research, it can be claimed that there is a ‘social failure’ of IBs as a result of their shift from IME’s equity-based objective to the neo-classical objective of efficiency.

In evidencing this, a number of benchmarks can be utilised. For example, in contrast to the ideal PLS or ‘equity’-based financial contracts, IBs currently heavily utilize debt financing instruments, which has been referred to as the ‘murabahah syndrome’ (Yousef 2004). Among others, Nagaoka (2007) and Asutay (2012) demonstrate the dominance of debt-based Islamic instruments in the operations of Islamic banks, accounting for over 90% of their financing. As a result, more stakeholders perceive that the financing methods of IBs are the same as those of conventional banks, which provide debt financing deploying different mechanics.

Second, despite their impact on economic growth, the performance of IBs has not promoted the economic and social development of the communities they serve. Aksak and Asutay (2015) pointed out that the GCC countries, which themselves have a dynamic economic growth and a high level of wealth per capita in general, have failed in their human development. They showed that IBs have not helped in contributing to such objectives as part of their maqasid efforts. Therefore, it is claimed that IBs have not fulfilled the aspirations of IME in socio-economic developmentalist areas or embedding their operations within the maqasid frame.

Third, the IME paradigm provides a rationale for CSR as a social responsibility towards society and extends it through maqasid expectations. In contrast, according to several empirical studies, IBs have failed in their CSR practices. For example, among others, Sairally (2007), Haniffa and Hudaib (2007), Platonova (2013) and Belal et al. (2014) found that most IBs conduct only zakah distribution and other charitable activities without a systemic approach towards CSR. Thus, the social performance is a weak area in IB performance indicating dis-embeddedness in relation to the maqasid al-Shari’ah defined objectives.

Maqasid frame and IME identifies that IBs are expected to have the best practices and a comprehensive CG framework. According to Choudhury and Hoque (2006), tawhidi or shuratic models of CG have wider groups of stakeholders in comparison with Anglo-Saxon and even European models. Nevertheless, currently IBs have not developed distinct CG structures which continue to utilize shareholder-oriented models; therefore, Chapra and Ahmed (2002) identified that IBs do not protect the interests of all stakeholders. In relation to Shari’ah governance (SG), Hassan (2012) identified that there are no ‘best practices’ in SG, which could lead to the deterioration of Shari’ah dimensions in IB operations and therefore increase Shari’ah non-compliance risks.

Based on the stated empirical evidence, it can be argued IBs have not met the expectations of Islamic economists and Muslims in general, which is further substantiated by El-Gamal (2006: xii) who stated that “Islamic finance has arguably failed to serve the objectives of Islamic law (maqasid al-Shari’ah)”. Therefore, it is important to develop an empirical framework based on maqasid to test the observed shortcomings in the ethical and social performance of IBs in the context of maqasid al-Shari’ah by adopting an extended maqasid frame beyond the Ghazalian framework, as discussed above.

Despite the dynamic growth of IBs in asset size and the complexity of their operations, few studies have addressed questions regarding the determinants of the financial, ethical, social and environmental performance of IBs, which is essentialised by the maqasid al-Shari’ah frame within the scope of Islamic finance, including both the institutional or socio-political context as well as organizational features. However, most of the empirical studies have explored only determinants of financial performance and the development of IBs. For example, Gazdar and Grassa (2015) identified the positive influence of macroeconomic factors, such as income per capita, economic openness and population, for Islamic finance in GCC countries while institutional factors were not considered relevant to the development of IBs. In another study, Mollah and Zaman (2015) investigated the positive influence of Shari’ah supervisory boards (SSB) on financial performance, and they found that corporate governance attributes, such as the board of directors’ size, independence and its CEO, overall negatively impact IB financial performance. In focusing on ownership, Zouari and Taktak (2012) trace family and government ownership of IBs with higher financial performance while institutional and foreign shareholders do not have such influence on IBs.

In relation to the ethical, social and environmental dimensions, most studies have explored the determinants of the social and ethical reporting of IBs. Farook et al. (2011) report that the relative size of Muslim populations, the level of social and political freedom, the SSB, and the share of investment account holders (IAHs) to total assets positively influence CSR disclosure.

In addition, Yousef’s (2004) cross-country analysis of ‘murabahah syndrome’ showed that legal institutions and political factors significantly influence the prevalence of debt-based instruments over equity-based ones in Islamic finance. As a result, Yousef (2004, p. 76) expects “that overall institutional environment will be a more important determinant of the evolution of financial structures in Muslim societies than … rigid religious interpretations”.

In conclusion, the IBs, as institutions of IME, are expected to emphasize their socio-economic role in society by enhancing their economic, social, ethical and environmental performance. However, as mentioned in the above-presented studies, there is a gap in the current practices of IBs with regard to their ideal behaviour. Moreover, there is a lack of research in relation to the determinants of the economic, social, ethical and environmental performance of IBs. As a result, further empirical studies are required to determine the factors affecting the performance of IBs in light of maqasid al-Shari’ah, a task this study aims to fulfil in this paper.

For the purposes of designing the evaluative framework for the maqasid or ethical performance of the sampled IBs through disclosure analysis, Najjar’s extended maqasid framework is implemented with four primary objectives and eight corollaries, as discussed in Bedoui and Mansour (2014). However, to link practice-related indicators of the economic, social, ethical and environmental performance of IBs with theoretically developed maqasid al-Shari’ah objectives and corollaries, Sekaran’s (2000) method is utilized. Accordingly, ‘concepts and notions’ can be broken down into observable characteristics or ‘dimensions’, and, subsequently, they can be divided into measurable behaviours or ‘elements.’ In line with Mohammed et al. (2008), ‘elements’ are separated into a list of indicators of various performance ratios. Therefore, a comprehensive evaluative framework in relation to the maqasid al-Shari’ah framework is presented in Fig. 1, which is utilised in the empirical analysis.

Based on the existing literature, such as ethical performance by Belal et al. (2014), along with current ‘best practices’ in both Islamic and conventional finance (for example AAOIFI, IFSB and others), as well as macro and micro-level maqasid indicators developed by Mohamed (2018), a comprehensive list of benchmark performance indicators is constructed which are then traced to each ‘element’, to each ‘dimension’ and then ultimately to each ‘concept’ or maqasid objective. The index in its fullness is presented in “Appendix”. It should be noted that not all performance indicators from the above-mentioned sources are utilised as only relevant indicators are included in the list. In quantifying, most indicators are scored based on a ‘dichotomous approach’: a score of ‘1’ is given if the indicator is found in the annual report and ‘0’ is given if not (Belal et al. 2014). In addition, some indicators are calculated based on ratios or other financial calculations.

To measure the overall maqasid performance of IBs and their separate performance in each primary objective, a maqasid al-Shariah index with sub-indices in each corollary is constructed. As mentioned above, the calculation of the index is based on an un-weighted approach and all indicators were weighted equally in each element’s groups to control possible bias in the measurement of maqasid performance of the sampled IBs in line with previous studies such as Haniffa and Hudaib (2007). It should be noted that Antonio et al. (2012) used a weighted approach, which is not considered in this study due to its heavily subjective nature, as he constructed the weights through the interviews with the Shari’ah scholar. In addition, Moqbel (2014) in his research referred to the frequency of each of the MSI dimensions mentioned in the Qur’an is determining the weightage. However, tawhid axiom of the IME being complementarity within unitarity suggests that each dimension of MSI should be considered equally, as assigning different weights will essentialise certain dimensions on other by creating dominance. Tawhid by definition prevents any domination of any of the dimensions of MSI over the other, as this would be contrary to ‘justice (adl)’ and ‘beneficence (ihsan)’ axiom of IME too. Therefore, this paper refers to the unweighted method to fulfil the tawhidi expectation.

The calculation of maqasid al-Shari’ah sub-indices according to each corollary is completed through the following formula (Haniffa and Hudaib 2007):where I_i,t,c: Maqasid al-Shari’ah sub-indices of ‘c’ corollaries of Najjar’s maqasid of ‘i’ IB in year t; X_i,j: score of indicators of ‘i’ IBs in year t; n_j: the overall number of indicators in ‘c’ corollaries.

Consequently, the calculation of the overall maqasid al-Shariah index is based on the following formula:where I_i,t is the maqasid al-Shari’ah index of the performance of IBs i in year t.

The research hypotheses in this study were developed based on the literature review, both on the determinants of performance and the disclosure practices of IBs as well as of conventional banks and corporations.

There are several factors which may have an impact on the economic, social, ethical and environmental performance of organizations. For example, Aguinis and Glavas (2012) highlight that companies conduct CSR activities due to institutional, organisational or individual factors. As for social reporting of IBs, Farook et al. (2011) analyse the influence of both country and IB-specific factors, in which two main groups of determinants are considered, including country-specific factors such as socio-economic development and political context and IB-level specific factors.

As part of the business environment, political and socio-economic factors shape the environment in which every corporation operates, including IBs. Jones (1999) argues that a country’s institutional arrangements in politics, law and economics along with the level of national development can enhance the social responsibility of organizations, neutrally sustain it or actively resist it. In line with this, Yousef (2004) traced the prevalence of ‘murabahah syndrome’ in civil law countries with weak institutions and non-democratic politics. Farook et al. (2011) outlined that within countries where political rights and civil liberties are limited, IBs could face lower social expectations and, thus, have incentives towards lower social disclosure. In contrast, in countries with a higher degree of freedom, IBs provide more disclosure to legitimize their existence. Therefore, it is expected that the higher a country’s economic and human development, the higher the maqasid performance of IBs. Meanwhile, the higher the repression of political rights and civil liberties along with civil law system, the lower the maqasid performance of IBs. Hence, hypotheses 1, 2 and 3 are defined as follows:

H₁: Human development positively affects the maqasid performance of IBs;

H₂: Political and civil repression negatively influences the maqasid performance of IBs;

H₃: Civil law based legal systems negatively influence the maqasid performance of IBs.

Secondly, the size of the Muslim population in countries where IBs operate may influence their maqasid performance. Based on the concept of a ‘relevant public’ to whom the corporations are accountable, Newson and Deegan (2002) found evidence of a positive impact of the size of Muslim populations on the social reporting of IBs. Therefore, it is expected that the higher the Muslim population, the higher the maqasid performance of IBs, as Muslim communities could put pressure on IBs to perform better in the ethical, social and environmental dimensions of their activities. Hence hypothesis 4 is formulated as follows:

H₄: The share of the Muslims population in a country’s total population positively affects the maqasid performance of IBs.

As for the third factor, the ownership structure of IBs, Zouari and Taktak (2012) indicated that, according to the existing literature, the ownership structure and concentration are crucial factors that impact a company’s financial performance and stability. However, they summarise that previous empirical studies reported mixed results on the relationship between ownership concentration and structure and performance. Nevertheless, their empirical study showed that family and government ownership of IBs positively influences their financial performance, whereas the same impact was not found in institutional and foreign shareholders. Therefore, it is expected that ownership structure either positively or negatively affects the maqasid performance of IBs, which is stated in hypothesis 5 as follows:

H₅: There is a positive relationship between ownership structure and the maqasid performance of IBs.

The fourth factor is corporate governance related organizational structures and the features of the board of directors in IBs. Mollah and Zaman (2015) summarised that, according to the existing available literature related to all types of companies, there is mixed empirical evidence of the influence of governance mechanisms on a company’s performance. Nevertheless, their study showed that a board of directors’ size and independence are negatively affected by the financial performance of IBs. In contrast, Zeitoun (2013) identified that, within several corporate governance features of companies, only ‘board independence’ has a significantly positive effect on social performance. Thus, based on the above rationalisation, it is expected that the corporate governance structure may affect the maqasid performance of IBs, as identified in hypothesis 6:

H₆: There is a positive relationship between corporate governance structures and the maqasid performance of IBs.

Finally, Shari’ah governance is one of the main features of IBs, ensuring that their products and operations are Shari’ah compliant, which are mainly presented by SSB features. Farook et al. (2011) proposed that SSBs could be a factor for CSR activities and social disclosures, and this was deduced from the SSBs’ role as promoters of Shari’ah compliance and presenters of Islamic laws and principles. Therefore, it is expected that the better the Shari’ah governance, the higher the maqasid performance of IBs, as stated in hypothesis 7:

H₇: Shari’ah governance structures positively affect the maqasid performance of IBs.

After identifying the hypotheses, the following section explains the panel data model used to examine the factors impacting the disclosure performance in relation to maqasid al-Shari’ah.

In order to analyse and test the identified hypotheses in line with the stated aims, the following linear regression model was utilized:where I_i,t is the Maqasid al-Shari’ah Index and its eight corollaries in bank i in year t; PSEC_i,t is the vector of time-varying political and socio-economic context variables; O_i,t is the vector of time-varying ownership structure variables; CG_i,t is the vector of time-varying board of directors’ variables; SSB_i,t is the vector of time-varying SSB variables; C_i,t is the vector of control variables; Ɛ_i,t is the white-noise error term.

In this paper, the dependent variable is the maqasid al-Shari’ah index (MSI), which is constructed by this study and calculated as explained above.

The independent variables in this paper are grouped into four main categories: (i) political and socio-economic context variables, (ii) corporate governance, (iii) SSB features related variables, and (iv) ownership structures of IBs along with economic and financial variables, as consistent with previous studies.

The political and socio-economic context variables include the ‘Human Development Index’¹ (HDI, as obtained from the UNDP website), ‘political and civil repressions’ (PCR, as presented using data from the Freedom in the World Index), and the ‘Muslim population ratio’ (MUSLIMRATIO, as a percentage of the Muslim population of a country using data from the CIA World Factbook). Three dummy variables relating to legal systems are ‘common law’ (COMLAW is marked ‘1’ if the country uses a common law system and ‘0’ if not), ‘civil law’ (‘CIVLAW’ is marked ‘1’ if the country uses a civil law system and ‘0’ if not) and ‘Shari’ah law’ (‘SHARLAW’, is marked ‘1’ if the country uses an Islamic law system and ‘0’ if not) by using data from the JuriGlobe database of Ottawa University. These variables are utilised in the existing literature, including by Jones (1999) and Farook et al. (2011).

The ‘corporate governance’ variables consists of ‘board independence’ (‘INED’ as a percentage of non-executive directors in the board of directors), ‘board pro-stakeholders directors’ (‘BPSD’ as a percentage of board members, for example politicians, academic scholars and others who have not held any other positions in the other organizations in the same industry), ‘board size’ (‘BS’, as the number of members in the board of directors), ‘CEO duality’ (‘CEOD’ is marked as ‘1’ if the CEO and chairman are the same person and ‘0’ if otherwise) and ‘investment account holders ratio’ (‘IAH’ as a percentage of IAHs account deposits to equity). It should be noted that these variables have all been utilized in previous studies, such as Farook et al. (2011) and Mollah and Zaman (2015).

The SSB-related variables are presented by a few indicators as ‘SSB composition’ (‘SSBC’ is marked ‘1’ if the SSB consists of more than 3 members and ‘0’ if otherwise), ‘SSB disclosure level’ (‘SSBD’ represents the level of compliance of the SSB report with AAOIFI standards) and ‘SSB control and monitoring abilities’ (‘SSBCM’ represents the average marks of the three components including the number of meetings, cross-membership in other IBFs, and the SSBs’ audit activities).

As regards to ownership structure, four dummy variables are included: ‘government owned’ (GOV), ‘family owned’ (FAM) ‘institutional’ (INST) and ‘foreign ownership’ (FORG). The respective variables are marked ‘1’ if the IBs are owned by such owners. The same variables have been used in a previous study by Zouari and Taktak (2012), among others.

All the control variables are grouped into bank-specific control variables since several IB characteristics may positively or negatively influence maqasid performance in the form of social, economic and environmental performance according to previous studies. The control variables included in this paper are the ‘IBs’ size of assets’ (SIZE), which is presented by a log of the total assets; ‘the leverage level’ (LEV), which is measured as the ratio of total debt to equity, and the ‘date of incorporation’ (AGE). All these variables have been utilized in several studies such as Beck et al. (2013), Zouari and Taktak (2012), and Rizkiningsih and Dewi (2015).

Table 2 presents the independent and dependent variables and the data sources.

In order to test the hypotheses developed, the annual reports of 33 IBs from 12 countries over nine-year period covering 2008–2016 were utilised for disclosure analysis, with a total number of 297 annual reports. The sample selection was determined by the availability of the annual reports on the IBs’ websites with a preference for large and well-established IBs in terms of assets in each country. The choice of selecting 12 countries was motivated by the necessity to cover a wider range of countries with IB presence along with the aim of capturing the impact of different socio-political, institutional and human development level contexts. The sample of IBs, with such characteristics as being listed or non-listed companies, their country of origin, as well as their total assets by the end of 2016, are presented in Table 3. Since the data generation is based on disclosure index and analysis, expanding the sample was not possible due to a very large maqasid index with 139 items.

As for the bank-specific, economic and political economy variables, the Zawya, BankScope, IMF, the UNDP and World Bank databases were utilised.

The first set of analyses in this section examines the maqasid performance of each of the sampled IBs, which is summarised in Table 4.² As can be seen, Safwa Islamic Bank (4.8488), Jordan Islamic Bank (2.473), Faisal Islamic Bank (Sudan) (1.8896) and PT Bank Syariah Mandiri (Indonesia) (1.8087), PT Bank BRI Syariah (Indonesia) (1.4398) are the five best banks with the highest scores for their maqasid performance, whereas BankIslami Pakistan Limited (0.6438), Al-Shamal Bank (Bahrain) (0.5661), KFH Bahrain (0.5243), Bahrain Islamic Bank (0.0329) and Al Rayan Bank UK (− 0.9155) demonstrated the lowest scores. Overall, except for a high dispersion in safeguarding the ‘wealth’ maqasid corollary, no best practices were identified in the performance of IBs within other maqasid corollaries, which supports the findings of previous studies including, among others, Hassan (2012), Belal et al. (2014), Haniffa and Hudaib (2007) and others. Only a few IBs, which are in the top five maqasid performers, scored higher than half of the score in the maqasid indices.

As regards to country level performance, as can be seen in Table 5, Jordan scored the highest average maqasid performance index score of 3.34, followed by Indonesia (1.44) and Sudan (1.15). The three countries with a low IB maqasid performance index are Kuwait (0.75), Bahrain (0.60) and the United Kingdom (− 0.91). Such a low maqasid performance from IBs in these two rich GCC countries supports criticism of the IBs, namely, that they are ‘delivering a new capitalism to the Muslim world’ which is oriented towards financialization rather than focusing on their social role (Asutay 2012, p. 108).

In addition, the performance of IBs per each maqasid corollary except I₇ (wealth), as IBs are inherently oriented towards financial performance, are presented in Fig. 2. As can be seen, mostly IBs in Indonesia, Bangladesh, Jordan and Pakistan are oriented towards safeguarding the ‘self’, ‘posterity’ and ‘social entity’ maqasid corollaries, which indicates their more active social and developmentalist role in these countries as opposed to the IBs in the GCC countries (Bahrain, Qatar, and Kuwait) and the UK. These results are consistent with Aksak and Asutay’s (2015) findings regarding the poor role of IBs in the economic and social development of the GCC countries.

Table 6 presents the summary of the descriptive statistics. As can be seen, the mean of the overall I_maqasid index is approximately 1.0284 while the range of this index is from a minimum of − 1.6493 to a maximum of 6.7370. Such wide dispersion in the I_maqasid index is due to the wide spread of the I₇ (wealth) index component (from − 14.0905 to 49.6030), which indicates that some IBs overperformed in financial terms, while others faced significant financial difficulties during the 2008–2016 period. Although the mean of the I₇ (wealth) index is approximately 5.2846, the mean of the I_maqasid index is significantly lower as a result of the low scores in the other maqasid performance indices as components of the I_maqasid index. Such indices as I₈ (ecology), I₄ (intellect), and I₆ (social entity) scored 0.1065, 0.1711 and 0.4162, respectively, as these indices have minimum zero (0) scores, which indicates that some IBs did not undertake any actions in these maqasid corollaries during 2008–2016. Furthermore, the most surprising finding is the low I₁ (faith) index mean score (0.3791), which is in contrast to the expectation that IBs are niche financial institutions that are expected to serve the financial needs of people according to Islamic expectations. Thus, low I_maqasid index is an indication of the low scores prevailing throughout the sample banks and periods as well as dimensions.

As can be seen from Fig. 3, the Jarque–Bera, skewness and kurtosis coefficients are approximately 37,764.748, 3.113,241 and 20.08,921, respectively, for the I_maqasid distribution, which indicates that the distribution is not normal.³ Nevertheless, a judgment can be made in favour of the normal distribution for this particular dependent variable. In addition, according to Brooks (2008, pp. 44, 164), an indication of non-normality for the distribution is not a significant concern for the purpose of determining the Best Linear Unbiased Estimator (BLUE) estimators using OLS if other assumptions are valid. Therefore, non-normality in dependent variables is not considered a restriction to our regression model.

The efficient conduct of a regression analysis requires the absence of multicollinearity; therefore, in this study, Pearson’s Correlation was utilised to explore the presence of multicollinearity in the relationship between dependent, independent and control variables. According to Gujarati (2003, p. 359), the pair-wise correlation between explanatory variables in excess of ± 0.8 creates a serious multicollinearity problem. As presented in Table 7, a significant pair-wise correlation can only be found between I_maqasid and I₇ indices (0.9529), which is not the main aim of this study. All other pair-wise correlations between variables are found to be less than 0.8, which helps to conclude that multicollinearity is not an issue for the pooled regression, FEM, REM models presented in this paper.

As this study is based on unbalanced panel data, several statistical tests were performed, including the simple OLS ‘pooled regression’, redundant fixed effect test, the random effect model (REM), the fixed effect models (FEM) according to the existing literature on econometrics (see: Brooks 2008; Gujarati 2003) (Table 7).

Tables 8 and 9 present the redundant fixed effect and Hausman test, which had the null hypothesis that no cross-sectional and period specific fixed effects exist, along with no significant difference between FEM and REM estimators. Since the results reject the null hypothesis, the FEM should be considered as an appropriate model for the purpose of this study. In addition, following the recommendation by Gujarati (2003, p. 418), White’s robust standard errors test in FEM was run in order to prevent the potential negative influence of heteroscedasticity,⁴ which is also supported by the higher adjusted R² coefficient (0.7134), as shown in Table 10 as an indicator that a significant part of the variation of the I_maqasid index is explained in comparison with other models. Nevertheless, the empirical results section presents all four models (pooled regression, REM, FEM, and FEM with robust standard errors test) since Gujarati (2003, p. 651) argues that “panel data do not provide a cure-all for all of an econometrician’s problem”.

Due to the specifics of the FEM model, which inherently includes differential intercept dummies, Eviews, the econometrics package, removed four dummy variables (CIVLAW, COMLAW, SHARLAW and FORG) in order to prevent a ‘dummy variable trap’ or perfect collinearity (as stated by Gujarati 2003, p. 642).

Table 10 presents the results of the panel data analysis using all models for overall maqasid performance index, I_maqasid.⁵ The results are presented in the form of ‘Pooled Regression Model’, ‘random effect model (REM)’, ‘fixed effect model (FEM)’ and ‘FEM with Robust Standard Errors’ for four models. While Model 1 tests all the variables included in the model with Pooled Regression and REM, Model 2 presents the results with FEM and FEM with Robust Standard Errors in which CIVLAW, COMLAW, SHARLAW and MUSLIMRATIO as well as FORG and AGE are excluded. In an attempt to locate the effect of corporate governance (GOV, BPSD, BS and CEOD) on maqasid performance by political and socio-economic environment (COMLAW, SHARLAW and MUSLIMRATIO), interaction variables between the corporate governance and political and socio-economic variables are introduced. The results for these presented in Table 10 as Model 3 and Model 4 in a similar way with Model 1 and Model 2. According to the results of FEM with the robust standard errors model, several political and socio-economic context related variables, ownership structure, corporate governance, and SSB variables all influence the maqasid performance of IBs. Concerning interaction variables in models 3 and 4:

As can be seen, the impact of HDI (− 4.9606) is negative and statistically significant with p = 0.3529, which confirms the finding by Jones (1999); hence hypothesis H₁ is not rejected. A possible explanation for this might be the active role of the government and regulators in enhancing ethical, social and environmental activities by IBs as additional resource providers for solving country-specific issues as well as the customers’ proactive positions. As a result, it is expected that IBs in low HDI index countries should expand their socio-ethical-economic activities in line with maqasid to fulfil the requirements of their regulators, minimize pressure from other stakeholders, and legitimize their role in society as active players in solving social and environmental problems.

For solving the non-linear relationship between dependent and independent variables, the macroeconomic variables of GDP and financial development (FD) were logarithmically transformed in order to preserve the linear model. The results show that FD has a negative significant impact on maqāsid (− 0.8180 with p = 0.0974) which is not surprising as the financial development variable in market economies has been determined by fundamental shifts towards a market-based financial system and its stability, but not by social expectations. Similarly, re-estimating the panel including the interaction variable between corporate governance (INED) and political variables (PCR) (PCR*INED), GDP also shows a significant negative impact (− 0.8041 with p = 0.0368) on maqāsid performance.

As the results depicted in Table 10 demonstrates, MUSLIMRATIO (4.8834, p = 0.2563) is one of the main contributors to the higher maqasid performance of IBs, which is consistent with previous studies (such as Rizkiningsih and Dewi 2015), which implies that hypothesis H₄ should not be rejected. This can be interpreted as IBs being niche-oriented financial institutions, and hence their activities are dependent on the ‘relevant public’ to whom they are accountable.

The last environmental factor which has a negative sign and is statistically significant at the 5% level (p = 0.0442) is PCR (− 0.0693). This finding agrees with Farook et al.’s (2011) findings, which showed that PCR has a negative impact on the level of social disclosure as a proxy of the social activity of IBs. Therefore, hypothesis H₂ is not rejected by the regression model.

Turning to the ownership and corporate governance variables, the results in this study indicate that both the INST (− 0.8718 with p = 0.0265) and INED (− 1.2982 with p = 0.0117) variables have a negative sign and are significant at the 1% level, indicating their impact on maqasid performance, while CEOD (0.2327 with p = 0.0869) has a positive impact and is significant only at the 10% level. This is in contrast to Kunapatarawong and Martinez-Ros (2013), who found that institutional pressure has a positive but statistically insignificant impact on social performance.

The findings in this study show the negative effect of institutional investors on overall maqasid-based disclosure performance, which might be due to the incentives of institutional investors to increase benefits by reducing costs on social activities, which could result in IBs facing reputational issues from depositors following the negative impact on financial performance. Furthermore, the negative impact of institutional investors on financial performance was also identified by Zouari and Taktak (2012). The same reasons could be suggested for the negative influence of INED variables as the main objective of INEDs in any board of directors is mostly to protect the interest of shareholders rather than of stakeholders, which may result in INEDs pressuring IBs to provide more benefits to shareholders, thus leading to a decrease in social activity and indirectly causing the deterioration of the image and financial performance of the IBs. However, the findings in this study contrast with Zeitoun (2013), who argued that most of the studies found positive relationships between board independence and the social performance of corporations. Nevertheless, as the findings in the present study indicate, hypotheses H₅ and H₆ are not rejected.

Furthermore, positive but statistically insignificant results are found for the GOV and FAM variables, which differ with those in Zouari and Taktak’s (2012) study, as the latter found a positive and significant relationship between these variables and financial performance.

As can be seen in Table 10, two of the three SSB-related variables such as SSBC (0.2103) and SSBCM (0.3374) have a positive and statistically significant impact with p = 0.0087 and p =0.0137 while SSBD (− 0.4447 with p = 0.4288) is not statistically significant. These outcomes support Farook et al.’s findings (2011) who argued that a ‘number of members’ increase monitoring capacity of SSB and positively impact the level of Shari’ah compliance in IBs’ activities. Therefore, hypothesis H₇ should not be rejected.

From the control variables, only LEV (0.0339 with p = 0.0622) is found to be statistically significant at the 10% level with a positive sign, while SIZE and AGE are found to be statistically insignificant, which is consistent with the results of Oikonomou et al. (2012). It should be noted that for Islamic banking, this implies that the high level of leverage is due to higher liabilities as the main source of leverage comes from customer deposits (no interbank loans are allowed for IBs), which makes IBs concentrate on social and ethical aspects in addition to financial performance since any deterioration of the public image of IBs may significantly impair their financial stability.

The interaction variables between corporate governance and COMLAW, CIVLAW, and SHARLAW through dummy variables were created for Model 3 and Model 4; however, none of them were significant. In addition, an interaction variable was created between corporate governance (INED) and political variable (PCR) (PCR*INED) across four regressions in Model 3 and Model 4, which is only significant with FEM with Robust Standard Errors in Model 4.

In order to test the robustness of the model, a robustness check was performed on the presence of dynamic features in the model. We tested the effect of a one-year time lag (− 1) on maqasid disclosure performance, as Mirzaei (2011) found that the financial performance of IBs is influenced by period lag variables. Considering this dynamic behaviour of the financial performance of IBs, a separate regression model including the time lag variable was run. In this dynamic panel data analysis, the findings remained unchanged for the HDI, MUSLIMRATIO, and INST independent variables. However, CEOD and other SSB-related variables were found to be statistically insignificant, whereas the results were changed for BPSD and BS, both of which were found to have a negative impact. Such change may be explained by a significant dynamism in changes to the board of directors while SSB-related variables are mostly stable in IBs. Therefore, our main findings remain as stated in the main regression model.

To test for potential endogeneity between corporate governance performance and socio-economic performance, the Durbin–Wu and Hausman tests were applied. In a statistical model, endogeneity may occur as a result of joint determination between independent (corporate governance) and dependent (maqāsid) variable, or omitted variables, or if there is a correlation between explanatory variables and the error term (Greene 2003). The results in Table 10 indicate that the F-test is not significant and thus the null hypothesis of the Durbin–Wu test cannot be rejected, confirming that endogeneity does not represent a problem (Gujarati 2003) in this research.