Islamic Banking - More Financial Inclusion for Arab States?

Transcription

1. Islamic Banking More Financial Inclusion for Arab States ? ARAB BAROMETER WORKI NG PAPER NO. 3 June 15, 2016 Clement M. Henry
2. Islamic banking continues to grow more rapidly than conventional banking in most Arab and many Muslim majority countries , but the sceptic may ask whether it really has a future, mobilizing broad classes of new economic actors, or is simply diversifying the portfolios of wealthy Gulf investors, where, apart from Iran, most Islamic banking assets are presently concentrated. This study will explore the potential of Islamic finance to be inclusive, to introduce financial intermediation to populations up till now confined to informal economies. Might it promote social as well as financial inclusion, reaching out to lower income groups? At what political costs and possible benefits to the state actors who regulate these banks in each country? This study will first examine whether any greater inclusion is yet associated with the penetration of Islamic finance into the banking systems of Muslim majority states. While time series data are not yet available, a cross section of 38 states including most of the Muslim world except India was devised.1 It displays varying degrees of penetration of Islamic banking and therefore offers an opportunity to examine its possible relationship to World Bank assessments of financial inclusion. Worldwide surveys launched in 2011 by the Bank’s Global Financial Inclusion (Findex) Program offer extensive data about banking behaviour, based on country surveys. Financial inclusion can be indicated by the percentages of people having bank accounts and who actually borrow money from banks rather than other informal sources. The surveys also document the percentages of people who cite religious reasons for not having a bank account. At the country level these results will be included in our cross sectional data set. How, if at all, do levels of financial inclusion correlate with those of Islamic financial penetration? The overview of financial development in much of the Muslim world will also point to strategic areas in the Arab region outside the GCC countries where Islamic finance is underdeveloped and may have important potential to bring wider populations of informal economies into their respective banking systems. The rest of this paper will focus on public opinion in the Arab region and particularly on the larger parts of it –Algeria, Egypt, Iraq, Saudi Arabia, and Tunisia -- for which Arab Barometer survey data are available. The Arab Barometer was devised by political scientists and is principally concerned with political attitudes, not financial behaviour. Islamic finance, however, is too closely associated with both local and international politics to be left exclusively in the hands of financial analysts. Specifying the attitudes as well as socioeconomic backgrounds of those most closely associated in a variety of Arab countries with systematic rejection of interest-based finance will offer a more informed appreciation of the potential and limitations of Islamic finance, including their potential political costs and benefits for state regulators. An overview of Islamic banking penetration Just four of the GCC countries, Kuwait, Qatar, Saudi Arabia, and the United Arab Emirates, carry the bulk of Islamic banking assets. Their domestic commercial banking systems held 1 The countries for which market shares of Islamic banks could be estimated were Afghanistan, Algeria, Azerbaijan, Bahrain, Bangladesh, Brunei, Egypt, Indonesia, Iran, Iraq, Jordan, Kazakhstan, Kuwait, Kyrgyzstan, Lebanon, Malaysia, Maldives, Mauritania, Mauritius, Morocco, Nigeria, Oman, Pakistan, Qatar, Saudi Arabia, Senegal, South Africa, Sri Lanka, Sudan, Syria, Tajikistan, Tunisia, Turkey, Turkmenistan, UAE, Uzbekistan, West Bank and Gaza, and Yemen. Ben Naceur et al (2015) present an alternative way of assessing the financial inclusionary potential of Islamic banking. They compare financial performances of the states in the Organization of the Islamic Conference which host Islamic banks with those that do not. They also use World Bank data (p.13), some of which were incomplete or out of date.
3. over half of the sharia-compliant deposits registered in the 38 principal Muslim majority or adjacent countries identified in the present study . GCC banks also fielded many banking subsidiaries elsewhere. Table 1 presents the evolving market shares of Islamic banks in the commercial banking systems of Arab countries and of other Muslim majority countries having either large populations or large market shares. The four Gulf States were also among the best banked in our sample and registered among the highest rates of financial inclusion, in the sense of people holding bank accounts. Even among the poorest of their respective populations, over 65 per cent had bank accounts except in Qatar, and the only other states in our sample enjoying such high rates of participation were Bahrain and Oman, also members of the GCC, and Iran and Malaysia. The latter were also major players in Islamic finance: Iran had totally Islamized its banking system, and Malaysia actively encouraged Islamic capital markets as well as banking, having issued over half the sukuk outstanding in 2015, when Islamic banks held at least 23 per cent of the country’s commercial bank deposits.2 It did not follow, however, that Islamic finance was attracting depositors who would otherwise stay away from banks for religious reasons. Even in Saudi Arabia, the system most penetrated with retail Islamic banking, where 69.4 per cent of those surveyed had a bank account, 23 per cent of those without one cited religious reasons for staying away from banks. On other indicators of financial inclusion, such as proportions of people having savings accounts or those receiving financing from banks rather than friends, families, suppliers, employers or the like, the only country with consistently top ratings was Bahrain, the pioneer of conventional offshore banking.3 As the offshore banking business diminished in the 1980s and 1990s, the Bahrain repositioned itself as a major center and hub for Islamic finance, hosting the Accounting and Auditing Organization for Islamic Financial Institutions (AAOIFI) and launching the world’s first sovereign sukuk in 2001. The Central Bank of Bahrain subsequently took the initiative in developing liquidity instruments for Islamic banks. Its high rates of financial inclusion, however, were perhaps more the fruit of the country’s history as a banking center and its educational edge over the rest of the GCC, reflecting earlier oil wealth, than more recent surges in Islamic banking. The available cross sectional data can only correlate Islamic banking penetration with recent rates of financial inclusion surveyed by the World Bank, including the propensity to 2 Sukuk originating in Malaysia and the GCC countries constituted 54.7 and 31.3 per cent, respectively, of the total of outstanding sovereign and corporate sukuk in 2015, according to the Sukuk Monitor http://www.zawya.com/sukuk/ (accessed April 15, 2015). 3 With 82 per cent of its adult sample having bank accounts, Bahrain was second only to Iran, with 92 per cent; and in bank lending Bahrain was number 1, with Iran falling back to eleventh place, just ahead of Lebanon and Saudi Arabia.
4. Table 1 : Sharia-compliant shares of commercial bank deposits4 1st year Bahrain Kuwait Oman Qatar Saudi Arabia UAE Algeria Egypt Iran Iraq Jordan Lebanon Morocco Sudan Syria* Tunisia Turkey Yemen Bangladesh Brunei** Indonesia Kazakhstan Malaysia Nigeria Pakistan 1979 1977 2013 1982 1988 1975 19961986 1998 2007 2011 6.7% 18.0% 9.8% 16.2% 11.5% 24.4% 46% 31% 10.4% 18.1% 11.5% 7.9% 13.2% 13.7% 16.1% 35% 17.0% 3.2% 2012 2013 32.4% 47.3% 2.0% 26.0% 48.9** 24.7% 1991 0.8% 1.6% 1977 9.7% 8.1% 9.7% 10.3% 10.3% 1983 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 1992 25.0% 25.0% 1978 7.0% 8.9% 11.3% 11.3% 11.4% 1991 0.1% 0.4% 0.4% 2015 1977 17.0% 27.9% 100.0% 100.0% 100.0% 100.0% 2005 7.3% 1983 0.2% 0.6% 2.7% 3.3% 1985 0.8% 3.6% 4.3% 5% 6.6% 1996 4% 41.8% 31.9% 32.6% 42.1% 1983 1991 1992 2013 1983 2011 1997 0.1% 1.9% 65% 37.1% 4.1% 1.6% 14.9% 21.1% 39.1% 4.6% 22.3% 9% 20.0% 41.0% 5.0% 0.1% 22.7% 0.4% 10.3% * Syrian sharia-compliant percentage of private resident deposits in local currency ** Percentage of total assets, not deposits cite religion as a reason for not having a bank account.5 Within the sample of 38 countries for which data were available, there are weak positive but statistically insignificant correlations between market shares and/or the longevity of Islamic banking in the country, on 4 Sources: Harvard Islamic Finance Program, various years; McKinsey Co, 2005; Kuwait Bankers training data; EY Islamic Finance Competitive Reports; Indonesian Financial Regulatory Authority (OJK); Brunei: http://www.ambd.gov.bn/Pages/Banks-Financial-Highlights; OIC Islamic Finance in the OIC Member Countries, 2011; APBT, Tunis, 2012, 2013; various Islamic bank annual reports; author’s data set. 5 The national averages were extracted from the Findex survey available online. Since the percentages referred only to the people not having bank accounts, they were adjusted downward to include the entire national sample in the denominator.
5. the one hand , and, on the other hand, the two principal indicators of financial inclusion derived from the Findex data, the percentages of bank account holders and of bank borrowers.6 There is little evidence of any significant causal relation between Islamic banking and financial inclusion. The penetration of Islamic finance might be expected in general –with the possible exception of Saudi Arabia noted above-- to reduce religious opposition to banks, but the statistical relationship was very weak, even if pointed in the expected direction. Market shares of Islamic banking were negatively associated with religious reservations about banking (r=-.064), and the years of Islamic banking activity in a country only slightly more so (r=-.152). In cross country research financial inclusion is obviously associated with other variables such as banking traditions, per capita income, the broad money supply (M2) as a proportion of GDP, bank branches per hundreds of thousands of people, credit to the private sector as a percentage of GDP, countrywide education levels and literacy rates, degrees of urbanization, access to the Internet and the like. Since many of these possible determinants are inter-correlated, just a few suffice as controls for examining any possible relationships of Islamic financial penetration or religious objections to financial inclusion. A series of regressions, reported in Table 2, were run on both indicators of financial inclusion. Table 2: Regressions of Bank Accounts and Borrowing on possible correlates Dependent variable Bank Accounts Islamic share -4.605 years of Islamic banking -0.334 religious reason -176.9 ** GNI per capita Money (M2/ GDP) 0.001 ** Borrowing from Banks -15.603 -108.97 ** -175.499 ** 0.001 ** 0.001 ** -142.85 ** 0.001 ** 8.474 Branches per 100,000 adjusted R-squared 0.49 N= 30 * p<.05; ** p<.01 ***p<.001 -2.43 -4.63 0.001 11.457 private sec credit/ GDP -10.69 -39.219 * 0 0.032 -30.57 * 0 0 0 0.032 0.342 *** 0.07 * 0.475 30 -29.47 * 3.086 0.744 0.66 32 -37.2 * 0.496 31 0.28 30 0.479 32 0.072 * 0.202 31 0.478 32 Whatever the mix of control variables, neither the market shares of Islamic banking nor years doing business in a country have a significant impact on either of the two indicators of financial inclusion. Religious reasons for shunning banks retain a strong statistically significant negative relationship to both indicators of financial inclusion. In other words, Islamic banking does not seem to date to have lessened opposition to banking or to have attracted more people into the formal sector of the economy. If anything, market share and years of stagnant Islamic finance in a country are associated with fewer people having bank accounts or borrowing money from banks, but the results are not statistically significant. Selfexclusion on religious grounds seems to limit financial inclusion in this sample of Muslim countries, and even the existence of Islamic finance is associated with people borrowing more 6 Among the 38 countries in the sample these indicators of financial inclusion were highly correlated (r=.752) but deserve to be analyzed separately. Accounts and Borrowing were respectively associated with market shares of Islamic banks (r=.218; r=.303) and with longevity, the number of years since the first Islamic bank was established (r=.117; r=.228).
6. from family , employer, or shopkeepers, not from banks.7 The stubborn negative relationship between religion and financial inclusion in Muslim countries still remains a challenge to the infant industry of Islamic finance. Arab Barometer survey data, to be discussed below, further confirm a widespread rejection of conventional banking on religious grounds. Possibly Islamic banking will enjoy greater success in countering the exclusionary impact of these beliefs as it progresses beyond the traditional confines of the GCC countries. Some of the Muslim countries to the west and as well as east of the GCC have the greatest untapped markets, as indicated by the size of their respective economies. As financial deepening (M2/GDP) proceeds, much of the increased money supply will flow into commercial banking, and Islamic market shares may also continue to increase. One way of assessing the untapped potential is to compare the respective sizes of these economies with their respective Islamic bank deposits. The larger the GDP and weaker the penetration of Islamic finance, measured in deposits, the greater the potential of the Muslim majority country for developing Islamic banking. Table 3 estimates the present Islamic deposits bases of the Arab countries along with other principal Muslim majority countries included in the data set.8 From the data in Table 1 about market shares, the total of sharia-compliant deposits is estimated. In the final column of Table 3 the ratio of current GDP to sharia-compliant deposits is computed to offer a rough idea of each country’s potential as the Islamic finance movement progresses. Thus Indonesia, for instance, having the largest economy and also the greatest number of Muslim inhabitants, may have a market share of only about 5.1 per cent to date, but its ratio of 56 suggests tremendous potential. Nigeria’s high ratio portends an African powerhouse of Islamic finance if financial deepening occurs and its banks gain market share. Within the GCC, Oman is the late comer to Islamic finance and may have greater potential to expand than the other more mature banking markets. But the Arab country standouts, in terms of potential growth, are Algeria, Morocco, and to a lesser extent Tunisia, Egypt, and Iraq, where Islamic banks have already gained 25 per cent of the deposits (from a very low deposit base). Lebanon also shows high potential, albeit from a tiny Islamic market share. Its Findex banking profile closely matches Algeria’s in numbers of account holders and of those not holding account for religious reasons.9 The Arab Barometer, however, offers data that reveal the large differences in banking attitudes between the two countries and point to Algeria’s far greater potential. 7 Neither the market shares nor the longevity of Islamic finance seem to bear much relationship to borrowing from banks, but market share is significantly related to informal borrowing (r=.482, p<.003, n=35), whether from family (r=.358), employer (r=.415), or shopkeepers (r=.427 and .630). 8 World Development Indicators provide GDP and M2 in current USD. IMF International Financial Statistics provide the amount of M2 outside the depository system to be subtracted from M2. 9 51 per cent of the Algerians and only 46 per cent of the Lebanese had bank accounts, with about 8 per cent in each country citing religious objections. On the financing or borrowing side, however, 16 per cent of the Lebanese used banks, compared to only 2 per cent of the Algerians.
7. Table 3 : Financial Surfaces of Arab and other Muslim States Billions USD 2013 Bahrain Kuwait Oman Qatar Saudi Arabia UAE GDP $32.9 $175.8 $79.7 $203.2 $748.4 $402.3 M2 $24.3 $114.3 $31.1 $126.0 $419.1 $245.4 Algeria Egypt Iraq Jordan Lebanon Morocco Sudan Syria (2010) Tunisia WBG Yemen $210.2 $272.0 $229.3 $33.7 $44.4 $103.8 $66.6 $132.4 $214.9 $75.7 $41.8 $110.9 $116.3 $14.0 $47.0 $11.3 $36.0 $31.5 $1.8 $14.0 Bangladesh Brunei Indonesia Iran Kazakhstan Malaysia Nigeria Pakistan Turkey $150.0 $16.1 $868.3 $368.9 $231.9 $313.2 $521.8 $232.3 $822.1 $91.5 $11.3 $356.0 $48.0 $76.5 $450.9 $114.8 $95.2 $501.5 Bank Islamic deposits share $23.1 32.4% $109.1 47.3% $27.3 2.0% $122.3 26.0% $379.6 48.9% $234.3 24.7% $96.0 $174.2 $40.3 $36.4 $105.4 $88.2 $9.7 $19.0 $26.9 $9.0 1.6% 10.3% 25.0% 11.4% 0.4% 0.0% 100.0% 7.3% 3.6% 8.4% 42.1% $80.5 $10.4 $308.8 $45.3 $60.8 $427.6 $99.9 $72.2 $458.1 20.0% 41.0% 5.1% 100.0% 0.0% 22.7% 0.4% 10.3% 6.6% Islamic deposits $7.5 $51.6 $0.5 $31.8 $185.6 $57.9 GDP/Islamic deposits 4.4 3.4 145.9 6.4 4.0 7.0 $1.6 $18.0 $10.1 $4.1 $0.4 $0.0 $9.7 $1.4 $1.0 133.2 15.1 22.8 8.1 105.2 $3.8 9.5 $16.1 $4.2 $15.5 $45.3 $0.0 $97.1 $0.4 $7.4 $30.2 9.3 3.8 56.1 8.1 6.9 48.6 . 3.2 1305.8 31.2 27.2 Preliminary Findings of the Arab Barometer Arab Barometer surveys conducted in 2007-08, 2011, and 2013 indicate that large majorities of representative national samples across the Arab world believe that conventional bank interest is contrary to the teachings of Islam.10 Table 4 summarizes the information, including only those who agreed or disagreed with the proposition that banks charging interest contradict the teachings of Islam. In the second wave of surveys, conducted in 2011, 10 The Arab Barometer (http://www.arabbarometer.org/), originally founded at the University of Michigan in 2005 in consultation with the Global Barometer (http://www.globalbarometer.net ), partners with the Arab Reform Initiative (http://www.arab-reform.net/) in association with other universities in the Arab world to carry out periodic surveys of public opinion. The three waves surveyed representative national samples, each including over 1000 respondents). The countries covered are reported in Table 4.
8. respondents were also asked whether , “in order to meet the demands of the modern economy, banks should be allowed to charge interest.” While the responses were significantly correlated with the question about the teachings of Islam, the relationship was weak (r= -.162, p<.000): a fair number of respondents who had agreed that interest was forbidden in Islam also thought that banks should be allowed to charge interest regardless of the religious teachings.11 The final two columns of Table 4 therefore report the percentages of those really interest-averse people who both agreed that interest was haram (forbidden) and also that banks should not charge interest even for the sake of economic development. Table 4: Arab Attitudes toward Bank Interest Interest is: contrary to teachings of Islam 2006-2008 (n=) 2011 88.5% 1123 86.3% 72.3% 80.0% 85.6% 1038 87.2% 75.8% 650 64.8% 938 68.9% Algeria Egypt Iraq Jordan Kuwait Lebanon Libya Morocco Palestine Saudi Arabia Sudan Tunisia Yemen Average 86.0% 85.1% 1193 1221 81.3% 966 85.5% 76.7% 64.2% 85.1% 76.8% 81.9% 7129 77.8% (n=) 980 1126 1059 1130 1202 1152 1077 1299 957 1037 11019 2013 89.9% 70.8% 64.7% 88.9% 77.4% 71.7% 82.6% 90.5% 84.1% (n=) 1124 937 994 1737 983 1038 1149 1000 1154 60.2% 79.5% 80.7% 79.1% never acceptable 2011 (n=) 51.6% 854 33.5% 1086 60.1% 939 47.1% 1109 27.4% 1115 1126 1096 1087 61.3% 63.6% 41.6% 42.4% 54.1% 1115 992 1224 817 938 13425 47.8% 10189 These results point to the sharp differences in opinion between the Lebanese and the Algerians.12 The countries combining substantial financial surface with intense objection to conventional finance are Algeria, Iraq, and possibly Morocco, which is unfortunately missing from the second wave of data. Despite its smaller surface Tunisia also displays strong potential, with 42 per cent unconditionally opposing bank interest despite its heritage of modernization under Habib Bourguiba. 11 Demirguc-Kunt et al (2013: 5) found in a 2012 survey of over 5000 adults focused on banking in Algeria, Egypt, Morocco, Tunisia, and Yemen that 45 per cent preferred a sharia-compliant product over a less expensive conventional bank product, but that 37 percent preferred the conventional product or did not have a preference. 12 These data cannot really be compared to the Findex findings about those citing religion as a reason for not having a bank account. It is still interesting to note the relatively good fit between the two data sets. The raw Findex data would rank the eight countries (excluding Lebanon and Tunisia, the latter for lack of data) as follows, comparing the ranks with the percentages recorded in Table 4: 1) Iraq 60.1% 2)Palestine 61.3% 3)Saudi 63.6% 4)Jordan 47.1% 5)Yemen 54.1% 6)Algeria 51.6% 7)Sudan 41.6% 8)Egypt 33.5%
9. The present analysis must focus on the findings of this second wave of public opinion surveys , since the other waves had no independent check on perceptions of banking to complement the very widespread belief that bank interest is religiously unacceptable. The other surveys suggest that Morocco may share Algeria’s potential for Islamic finance, and under its present government Morocco is taking important first steps in promoting Islamic banks in the kingdom. But further analysis of the survey data is only possible for the ten countries included in the final two columns of Table 4. The data set offers a wealth of information about social and political attitudes as well as personal data that the potential clients of Islamic banks may share. In 2011, when the survey was conducted, the Arab world was exceptionally open to political research. If indeed the likeliest clientele for Islamic banks is people who are unconditionally opposed to interestbased conventional banking, it becomes interesting to examine the relationships to their respective polities. To what extent are these potential participants in Islamic banking interested in politics? How much trust do they have in state institutions? How do they relate to other ideas and behaviors relating to religious identity? In particular, variables connoting religiosity, conservative family values, secularism, and Islamism can be constructed out of the data set, along with trust in authorities, interest in politics, and socio-economic status. The inclusion of Saudi Arabia also permits us to test the hypothesis of portfolio diversification in the leading dynamo of Islamic finance. Which personal characteristics and social and political attitudes are most closely associated with the Saudis who unconditionally reject conventional interest-based banking? Wealth, age, and religiosity may point to a clientele in search of sharia-compliant substitutes for conventional investment portfolios. Do the other considerably less wealthy Arab countries show similar relationships between these personal characteristics and the rejection of conventional banking? How, moreover, do other political and social attitudes interact with these personal characteristics? Another hypothesis to be tested is whether being part of the potential Islamic finance clientele is associated with a strong interest in politics and/or with political Islam. Or is this potential clientele basically apolitical and conservative, or is it anti-secular or Islamist? Also to be investigated is whether it generally distrusts the state authority, perhaps associated in public opinion with conventional interest-based banking ever since foreign colonizers established both the state and its banks. If so, official efforts to promote Islamic finance might face an uphill challenge, seeking support from a hostile constituency— or perhaps gaining its trust. Correlates of Interest Rejection As a first step toward locating these potential Islamic bank customers, what sorts of personal characteristics are associated with the unconditional rejection of bank interest? In addition to location (urban/rural), age, sex, wealth, and education, “religiosity” is also reported as a variable summarizing personal religious activities,13 as is exposure through travel to a 13 Religiosity was derived from a set of interrelated questions. This analysis assumes that “religiosity” may be understood as a continuum, ranging in intensity, rather than a special state of mind reflected by the respondents’ answers to one or two questions (depending on the analyst’s definition of what “real” religiosity might be). Respondents were asked on a four point scale ranging from always to rarely (Tunisians were given “never” as a fifth point) whether they prayed, fasted during Ramadan or Lent, watched or listened to religious programs on TV or radio, attended religious classes in mosque or church, attended Friday or Sunday prayer services, listened to or read the Quran or Bible, and read religious books. The scores were then averaged to form a new continuous variable, “religiosity,” with
10. Western society .14 Does the personal profile of the potential clientele vary across the Arab states, which differ not only in percentages rejecting bank interest but also in their economic structures, educational systems, degrees of urbanization, and experiences with Islamic finance reported in Table 1? Table 5: Objective Personal Correlates of Interest Rejection All Algeria Egypt Iraq Jordan Lebanon Palestine Saudi Sudan Tunisia Yemen urban .031** 0.065 .134** -0.015 0.01 -0.006 0.058 -0.049 0.025 -0.049 0.022 10189 854 1086 939 1109 1115 1115 992 1224 817 938 age 0.011 .074* -0.04 0.026 0.053 -0.058 0.054 .101** -0.006 0.054 0.042 N= 10150 854 1086 939 1098 1115 1100 991 1216 817 934 female -.027** -.077* 0.026 0.001 -.133** 0.043 -0.015 0.001 -.112** -0.021 -.064* N= 10189 854 1086 939 1109 1115 1115 992 1224 817 938 education.054** 0.011 0.047 -0.062 0.059 -.086** 0.021 0.044 .193** -0.063 .101** N= 9327 854 1083 939 1106 1115 1111 992 1210 814 917 wealth .071** -0.042 .068* 0.058 -0.009 -.118** -0.031 .127** -0.045 -0.052 -0.021 N= 7951 352 974 834 1096 894 1052 821 581 669 678 unemployed -.023* -.072* 0.009 0.008 -.097** 0.037 -0.049 -0.025 -.076** -0.065 -.129** N= 10148 854 1086 939 1103 1115 1115 992 1218 817 909 married .032** -0.006 0.008 0.035 -0.002 0.041 0.011 0.018 -0.013 0.034 .085* N= 10035 826 1086 939 1065 1115 1073 992 1215 805 919 Christian -.170** -.099** -0.013 -0.039 -.316** -.096** 0.025 -0.066 N= 8229 1082 936 1105 1114 1107 1217 806 Religiosity.078** .272** .062* -.076* .080** .135** .161** 0.012 0.032 .136** 0.056 N= 9310 825 1086 929 1106 1108 1111 990 1221 808 934 Trips West-.061** -.164** -.118** -.096** -0.015 -.083** -0.056 0.015 -.100** -0.032 0.005 N= 10021 837 1079 931 1094 1103 1103 949 1194 811 920 **. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed). N= The aggregate correlation coefficients reported in the first column clearly mask wide variations within the national samples. Despite its apparently general statistical significance, for instance, it is only living in an Egyptian urban environment that seems significantly associated with the unconditional rejection of interest-based banking. Age, too, has mixed but significant effects, despite its overall apparent irrelevance. It is most strongly associated with interest rejection in Saudi Arabia. Indeed, looking further down the list of possible Saudi correlates, the only other statistically significant one is wealth, pointing to a potential constituency for Islamic investments. These survey results apparently reflect the appetite of wealthy Saudi investors for sharia-compliant banking assets that indeed stimulated the development of Islamic windows in conventional banks and the conversion of some of them to “Islamic” banking since the late 1990s. Curiously, however, personal religiosity did not correlate in a statistically significant way with the rejection of interest in the Saudi sample. Elsewhere, except Sudan and Iraq, the correlation coefficients were positive and statistically significant. Apparently males tend in general to reject bank interest more than females, but again, country by country the results are mixed. As for education, its positive and statistically the highest scores awarded those who engaged the most in these activities. Alpha (coefficient of reliability) = 0.78. 14 Respondents were asked whether they had traveled to a Western country in the past five years and, if so, for how many months, up to six or more.
11. significant relationship for the sample as a whole is reversed in Lebanon and to a lesser extent in Tunisia . Clearly the content of education matters, and there is significant cross-country variation. Wealth, too, shows significant variations, again pointing to a “deviant” Lebanon, perhaps because much of the wealth is Christian. Indeed none of the personal characteristics works in the same way across all the countries in the sample. Iraq is the most puzzling anomaly because religiosity is negatively correlated with interest aversion. Everywhere else religiosity is positively associated with rejecting conventional bank interest. Even in the relatively traditional Islamic societies of Saudi Arabia, Sudan, and Yemen there is still a weak positive relationship. In Iraq, by contrast, the relationship is negative and statistically significant. As noted earlier, the country seems to offer promising prospects for Islamic finance, yet religiosity does not appear to be opposed to conventional interest-based banking. In this country study the survey data identified respondents by sect as well as religion.15 A closer look at the data indicates that clearly identified Sunnis scored lower on our religiosity scale but tended also to be more opposed to interest-based banking than their Shi’ite compatriots. Clearly multivariate analysis was needed, to disentangle the possible effects of religiosity from sectarian identity. A new independent variable of Shi’ite religiosity was therefore constructed. And since the dependent variable, Interest Rejection, is binary, a logistic regression could be performed on all of the independent variables to separate out the independent effects of each, controlling for the others. Controlling for all the other variables, education seems significantly associated with greater toleration of bank interest. In Iraq an added level on a seven point education scale is associated with 12 per cent less likelihood (1-.88) of rejecting interest. Household income, too, becomes statistically significant, but works in the opposite direction. Every incremental $100 of monthly income is associated with an increasing likelihood of 3 per cent for rejecting interest. Religiosity still carries a negative sign but its statistical significance vanishes when the interactive effects of religiosity and Shi’ite identity are taken into account. One unit of Religiosity on a four point scale among Shi’ites makes them almost 30 per cent less likely (1.707) to reject interest.16 Exposure to the West still carries significant impact: a couple of months visiting the West are associated with being 35 per cent less likely to reject interest! Before pursuing a similar analysis for other Arab states, the ideational correlates of rejecting interest need to be introduced so as to move the analysis beyond issues of wealth, 15 The questionnaire tactfully asked respondents “If I asked about your religion, would you prefer the answer be…(listing the following possibilities: 1) Orthodox. 2) Catholic. 3) Protestant. 4) Christian. 5) Sunni Muslim. 6) Shi’ite Muslim. 7) Muslim. 10) Other (specific). 37.4% responded Sunni; 43.6% Shi’ite Muslim; 18.2% Muslim. The self-identified Shi’ites were identified, and a new interactive variable of “religiosity” and “Shi’ite” was constructed. 16 In an alternative model, substituting Shi’ite for Shi’ite religiosity, being a Shi’ite is associated with a 40% decline in the odds or likelihood of rejecting interest. Religiosity then retains its statistical significance (p<.035) and is associated with a decline of 27% in the likelihood of rejecting interest.
12. Table 6 : Logistic Regression of Interest Rejection on Personal Correlates (Iraq) B Urban 0.027 Age -0.005 Female -0.081 Education -0.128 Income$100s 0.031 Unemployed -0.134 Married 0.159 Christian -1.491 Religiosity -0.136 Relig_Shi'ite -0.347 Trips West -0.44 Constant 3.697 S.E. Wald 0.172 0.025 0.007 0.508 0.177 0.208 0.055 5.425 0.016 3.908 0.192 0.485 0.201 0.624 1.234 1.46 0.153 0.791 0.054 40.636 0.175 6.296 1.419 6.784 df 1 1 1 1 1 1 1 1 1 1 1 1 Sig. Exp(B) 0.874 1.028 0.476 0.995 0.648 0.922 0.02 0.88 0.048 1.031 0.486 0.875 0.43 1.172 0.227 0.225 0.374 0.873 0 0.707 0.012 0.644 0.009 40.309 portfolio diversification, and personal religious practice to better understandings of the possible politics that may be associated with Islamic finance. If the analysis of public opinion can identify a potential constituency for Islamic finance, how may policy makers either in government or business attract it, with what sorts of political consequences? Assuming that the unconditional rejection of interest signifies a potential sympathy for Islamic finance, how interested is this potential constituency in politics?17 Is it Islamist or predominantly secular in outlook, preferring a separation between politics and religion?18 If tending toward soft Islamism, the other end of the secularism scale devised from the survey questions, was it a really “fundamentalist,” version of Islamism, to be interpreted from responses to additional questions such as affirming hudud punishments, death to Muslim apostates, and a theocratic form of government.19 Or are these potential sympathizers of Islamic finance simply more conservative in their family values.20 What relationships, if any, 17 Interest in politics was indicated by averaging the scores of each respondent’s self-evaluation and by how much he or she followed the news in their country. Alpha (coefficient of reliability) = 0.86 18 Secularism was the mean of the responses on a four point scale of strong agreement to strong disagreement to the following: 1) Religious leaders (imams, preachers, priests) should not interfere in voters’ decisions in elections; 2) Religious practices are private and should be separated from social and political life; 3) Religious associations and institutions (excluding political parties) should not influence voters’ decisions in elections; 4) Mosques and churches should not be used for election campaigning. The scores were averaged to form a new variable, Secularism, or conversely soft Islamism, which had borderline reliability: Alpha =.687. The scores were reversed to give the secularist responses the higher scores. 19 “Hardline” Islam was the mean of responses on a four point scale to the following: 1) the appropriateness of a parliamentary system in which only Islamist parties compete in parliamentary elections; 2) of a system governed by Islamic law without elections or political parties; 3) agreeing that the government and Shura Council should enact penal laws in accordance with Islamic law; 4) that your country is better off if religious people hold public positions in the state; 5) that religious leaders (imams, preachers, priests) should have influence over government decisions; 6) that in a Muslim country, non-Muslims should enjoy less political rights than Muslims; 7) that when a person changes his/her religion he/she should be penalized with death. The scores were reversed to give the higher scores to the more hardline responses. Alpha (coefficient of reliability)=.773 20 Family values were derived from the responses on a four point scale of strong agreement to strong disagreement to the following: 1) A woman can become the prime minister or president of a Muslim
13. might there be between political Islam and Islamic finance ? A further question concerns possible relationships between those opposing interest and their confidence or trust in state authorities. Attitudes toward the government, the judiciary, the police, and the army seemed from the survey responses to be sufficiently inter-correlated to allow for an index of trust in state institutions.21 Table 7: Ideational Correlates of Interest Rejection All Algeria Egypt Iraq Interest in politics -0.009 -0.002 .074* -.070* N= 10149 853 1084 938 Secularism -.079** -0.029 -0.027 -0.052 N= 10024 813 1082 934 Hardline Islamism .102** .161** 0 0.025 N= 10186 854 1086 939 family values .136** .201** -0.006 .086** N= 10167 850 1086 939 Distrust of State .035** -.130** .151** 0.003 N= 10119 851 1080 936 **. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed). Jordan Lebanon Palestine Saudi Sudan Tunisia Yemen 0.019 1107 -.073* 1090 0.017 1109 .210** 1107 .069* 1104 -0.039 1110 -0.024 1108 .132** 1115 .177** 1114 0.035 1110 0.041 1115 -.073* 1110 .169** 1115 .172** 1115 0.003 1100 -.210** 975 0.056 953 -.173** 989 -.155** 982 .065* 977 .059* 0.013 .154** 1215 817 935 0.012 -0.015 -.083* 1196 803 935 0.052 0.007 .171** 1224 817 938 .118** .091** .197** 1221 816 937 .067* 0.043 .162** 1215 811 935 Across the entire sample many of these variables apparently bear statistically significant relationships to the potential Islamic finance constituency. Within each country, however, these relationships vary widely, and multivariate analysis is needed to discover which ones are really significant, controlling for the effects of the others. To return to Iraq, it would appear from Table 7 that any constituency for Islamic finance is positively disinterested in politics (r= -.07; p<.05) and basks in family values (r= .087; p<.01). Controlling for religiosity and other correlates of interest rejection, are conservative family values still so highly correlated with potential Islamic finance sympathizers? Is there no significant relationship between these sympathizers and political Islamists, when all other sources of variation are taken into account? To examine the possible independent impact of political Islam upon attitudes toward conventional interest-based banking, a logistic regression on Interest Rejection was run including all of the possible correlates so as to offer a fuller analysis.22 Education is no longer a significant negative correlate, but it turns out that the relationships shown in Table 7 state; 2) A married woman can work outside the home; 3) In general, men are better at political leadership than women; 4) University education for males is more important than university education for females; 5) Men and women should have equal work opportunities; 6) It is permissible for a woman to travel abroad by herself; 7) A woman should obtain her inheritance (she should not be denied her inheritance); 8) Women’s share of inheritance should be equal to that of men; 9) Women can assume judicial positions; 10) Women can become ministers. Reversing the scoring for items 3 and 4, the scores were then averaged to form a new variable, “family values,” since the conservative responses received the highest scores. Alpha (coefficient of reliability) = 0.803. 21 On a four point scale respondents how much they trusted or distrusted the government, the judiciary, the police, and the army (among others where the responses were more mixed, less related to one another). The mean was constructed with a coefficient of reliability, Alpha = .837. 22 The model in Table 8 produced a pseudo R-square of 13.6%, more than double the model in Table 6.
14. Table 8 : Logistic Regression of Interest Rejection – Iraq Urban Age Female Education Income$100s Unemployed Married Christian Religiosity Religiosity-Shi'ite Trips West Interest in Politics Secularism Hardline Islam Hardline Sunni Family values Distrust of state Sunni Distrust Constant B -0.003 0.001 -0.044 -0.05 0.023 -0.224 0.092 -1.156 -0.324 -0.191 -0.216 -0.544 0.225 -0.379 0.695 0.685 0.27 -0.337 2.177 S.E. 0.182 0.007 0.184 0.058 0.015 0.197 0.207 1.442 0.157 0.085 0.18 0.117 0.144 0.219 0.219 0.182 0.138 0.178 1.796 Wald 0 0.02 0.057 0.739 2.282 1.293 0.2 0.642 4.272 5.042 1.438 21.645 2.454 3 10.103 14.192 3.824 3.593 1.471 df 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Sig. Exp(B) 0.986 0.997 0.888 1.001 0.811 0.957 0.39 0.951 0.131 1.023 0.256 0.799 0.655 1.097 0.423 0.315 0.039 0.723 0.025 0.826 0.231 0.806 0 0.581 0.117 1.253 0.083 0.684 0.001 2.004 0 1.984 0.051 1.31 0.058 0.714 0.225 8.823 still hold: (dis)interest in politics and family values remain strong correlates. From Table 8 it can be seen that a unit of interest in politics on a four point scale is associated with 42% less likelihood in unconditionally rejecting conventional bank interest. One unit on a four point scale of family values, by contrast, is associated with twice the likelihood of rejecting interest. The negative rather than positive relationship of religiosity to Interest Rejection still held up, and indeed it turned out, controlling for all the other variables, that its impact is just as statistically significant for Shi’ites as for Sunnis. Regressions were also run excluding one of the two Islamism variables but the remaining one just gained marginal statistical significance.23 However, fundamentalist Islam seemed to carry different weights in the respective Muslim sects. An interactive Hardline-Sunni variable added to the model is shown in Table 8. It is highly significant: among Sunnis a unit on the Hardline Islam four-point scale is associated with twice the likelihood of rejecting conventional bank interest, whereas for the rest of the Iraqi population the Hardliners seem less likely to reject interest. So also, since the degree of distrust of state institutions varied between the sects, an interactive Sunni Distrust variable was included to complement Distrust of state in Table 8. Here, too, there were interesting differences: Sunnis who were less trusting of state institutions tended to be less likely to reject interest, unlike the rest of the population. Put differently, more trusting Sunnis were also more likely to be potential clients of Islamic banks. 23 For Hardline Islam, p<.055; for Secularism, p<.06.
15. These potential clients seemed uninterested in politics . Indeed, those more interested in politics seemed to associate with secular values, not Muslim fundamentalism.24 Table 8 summarizes the potential profile of an Islamic finance constituency in Iraq: relatively nonpracticing Sunni Muslims, turned off by day-to-day politics and imbued with “fundamentalist” principles as well as conservative family values. And although the Sunnis in the sample on average had less confidence in state institutions than other Iraqis, the Sunnis who were relatively more trusting of state institutions also tended to be more averse to bank interest. Promoting Islamic finance in Iraq might therefore have been a way in 2011 for the government to reach out to some of its politically as well as financially excluded citizens. Might it still? Or might some other party steal the initiative? The other more promising prospects for Islamic finance among the larger underdeveloped financial surfaces of the Arab states may lie in Algeria and to a lesser extent in Tunisia and possibly Egypt. Prospects for Algerian Islamic Finance As Table 1 indicated, Islamic banking has stagnated since 1991, when it was introduced to Algeria. The Al Baraka Group created a joint venture at that time with the Banque d’Agriculture et Développement Rurale, one of Algeria’s four dominant public sector banks, and seems in effect to be managed like a public sector bank. It opened its doors for business shortly before Algeria’s decade of civil war (1992-1998). Efforts to privatize banking in Algeria then went astray in 2003 with a series of spectacular scandals and bankruptcies. Consequently the government has not encouraged private sector banking, other than foreign banks, to penetrate what is still predominantly a state-run commercial banking system. The looming international financial crisis, notably trouble with housing loans at Citibank, deterred the Algeria government from privatizing one of its large public sector banks in 2007. Algeria exemplifies the seriously underdeveloped financing that has retarded investment and growth in much of the MENA. One indicator of the impoverished nature of banking systems, indicated in Table 3, is the amount of loose currency outside them. People prefer to hold their cash, at least partly out of widespread distrust for conventional banking as indicated in the Arab Barometer surveys. The converse, reported in Table 3, is the proportion of the broad (M2) money supply serviced by the commercial banking systems. Algeria’s ratio was higher than Iraq’s but lower than those of Egypt or Tunisia. Despite efforts since 1986 to reform the banking system, finance remains the Achilles heel of Algerian development. The survey data point, however, to a potential constituency that Al Baraka Algeria is not reaching but that other private sector banks might tap, once an adequate regulatory system is in place. A logistic regression including all the ideational as well as personal 24 Political interest with Islam, r= -.039, p<.01; with Secularism, r= .063, p<.01.
16. Table 9 : Logistic regression of Interest Rejection – Algeria B S.E. Wald df Sig. Exp(B) correlations Urban 0.187 0.342 0.298 1 0.585 1.205 0.065 Age 0.006 0.015 0.162 1 0.688 1.006 .074* Female -0.277 0.36 0.59 1 0.442 0.758 -.077* Education 0.192 0.096 3.996 1 0.046 1.212 0.011 Income$100s -0.068 0.046 2.206 1 0.137 0.934 -0.042 Unemployed 0.437 0.313 1.946 1 0.163 1.547 -.072* Married 0.292 0.385 0.575 1 0.448 1.339 -0.006 Religiosity 1.073 0.285 14.161 1 0 2.923 .272** Trips West -0.453 0.175 6.699 1 0.01 0.636 -.164** Interest politics 0.426 0.202 4.451 1 0.035 1.531 -0.002 Secularism 0.032 0.271 0.014 1 0.906 1.033 -0.029 Hardline Islam -0.233 0.352 0.439 1 0.508 0.792 .161** Family values 0.39 0.293 1.769 1 0.184 1.477 .201** Distrust of state -0.91 0.217 17.61 1 0 0.402 -.130** Constant -1.975 1.542 1.64 1 0.2 0.139 **. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed). variables pointed to religiosity, Western exposure, interest in politics, and trust in rather than distrust of state institutions as the significant correlates, along with a positive if only marginally significant relationship to education. A final column recalls the marginal correlates for Algeria recorded in Tables 5 and 7. Age, gender, and employment status lose their statistical significance, whereas Religiosity and Western exposure remain robust correlates, respectively positive and negative, of Interest Rejection. But the most interesting casualty of this multivariate analysis is Hardline Islam. Even if secularism is removed from the model, Hardline Islam is no longer associated with statistical significance to Algeria’s potential clientele for Islamic finance.25 Interest in politics, moreover, becomes a significant correlate in the logistic regression. And Distrust of state retains its strong negative correlation. In other words, the relatively educated, politicized Algerian who is devout and trusting in state institutions will be most likely to reject conventional bank interest and be a potential customer for Islam banks. The findings suggest that Algeria is not only brimming with potential demand for Islamic finance but also that the demand presents a political opportunity. This demand is not associated with Islam, but rather with religiosity as indicated by various behaviours. The potential customers also express an interest in politics as well as strong support for state institutions. Encouraging Islamic banking would make good sense for governments seeking to consolidate the support of a sympathetic constituency. In Malaysia successive governments took positive steps to encourage Islamic finance so as to neutralize an opposition party that was threatening the hegemony of the ruling party by invoking resurgent 25 In fact dropping both Secularism and Hardline Islam from the model would slightly increase its adjusted r-squared, from 23.6% to 25.5% (Cox and Snell), or from 31.4% to 34% (Nagelkerke). Family values then becomes almost significant at the .05 level (p= 0.051). In a model excluding Family values and Secularism, Hardline Islam is still statistically insignificant.
17. Islamic values (Harding 2010: 503). In Algeria, a similar strategy could reinforce regime efforts to keep moderate Islamists in the ruling coalition. Egypt after the Arab Spring Egypt is one of the first countries to have permitted the operation of Islamic banks, as Table 1 indicated. President Anwar Sadat supported special legislation to enable the Saudi Prince Muhammad al Faisal to establish the Faisal Islamic Bank of Egypt (FIBE) in 1977. Together with a private Egyptian-owned Islamic bank, they attracted roughly 10 per cent of commercial bank deposits by 1985, but rogue money management companies, also claiming to be “Islamic,” attracted more substantial shares of deposits and remittances until their Ponzi schemes collapsed in 1988, discrediting the entire idea of Islamic banking in Egypt. Further problems arose in 1991 with the collapse of Abu Dhabi’s Bank of Credit and Commercial Investment (BCCI). FIBE, unable to find suitable investments for its customers, had parked some of the investment funds with BCCI, which offered higher returns than the Egyptian government. FIBE consequently lost one-quarter of its balance sheet. It would later recuperate much of it, but Islamic market shares stagnated in Egypt. Bolstered by Islamic windows in Banque Misr and other conventional banks, the Islamic sector finally recovered in the 2000s, almost again reaching 10 per cent by 2011. There may still be some untapped potential. The Arab Barometer survey was conducted in Egypt after the Revolution of January 25 had deposed President Mubarak and facilitated a climate of research where surveying political attitudes became possible. The results recorded in Table 10 represent public opinion in mid-year, at a time of uncertainty before the elections conducted in December 2011 and January 2012 brought forth a parliament dominated by the Muslim Brotherhood (46 %) and the Salafists (24%). In March 2011 the military’s Supreme Council of the Armed Forces (SCAF) had orchestrated a successful referendum that outflanked the more secularly inclined revolutionaries who were insisting on more time to deliberate about proposed constitutional changes. By the summer of 2011 Egypt was still very much at the beginning of a painful transition from Mubarak to the election of President Mohamed Morsi in June 2012 and his ouster on July 3, 2013. The survey thus occurred before sharper political polarization resulted in prolonged violence and massive government efforts to discredit the Muslim Brotherhood. During the relatively short time Morsi was in power, efforts were made to strengthen Islamic finance but the legislation concerning sukuk remains a work in progress. The survey results are of interest precisely because they convey, as in Algeria (but not Iraq), the separation of potential sympathies for Islamic finance from any version of political Islam. In addition to the general political orientations being analysed, the Egyptian survey included questions about party affiliation. Of the 1219 Egyptians only eight were actual members of a political party but an additional 95 intended to join one. Of these only 46 expressed opposition to conventional banks, a bit above the national average of 33.5% noted in Table 4. Among future recruits to the Freedom and Justice Party (Muslim Brotherhood) opinion was divided, with 15 opposed, 11 in favor, and two on the fence, not expressing consistent responses to the questions concerning interest-based banks, whereas all three alNour (Salafist) Party sympathizers opposed them and therefore might support Islamic finance.
18. The logistic regression results reported in Table 10 show that secularism and Islam bear no direct or indirect association with interest rejection . They remain statistically insignificant even if only one of them is included in the regression. But at least during Table 10: Logistic Regression of Interest Rejection – Egypt B S.E. Wald df Sig. Exp(B) Correlations Urban 0.501 0.154 10.573 1 0.001 1.651 .134** Age -0.008 0.006 1.854 1 0.173 0.992 -0.04 Female 0.054 0.202 0.071 1 0.79 1.055 0.026 Education -0.05 0.047 1.151 1 0.283 0.951 0.047 Income$100s 0.067 0.045 2.272 1 0.132 1.07 .068* Unemployed 0.011 0.186 0.003 1 0.954 1.011 0.009 Married 0.057 0.197 0.084 1 0.772 1.059 0.008 Christian -1.828 0.637 8.24 1 0.004 0.161 -.099** Religiosity 0.284 0.166 2.928 1 0.087 1.328 .062* Trips West -0.874 0.257 11.555 1 0.001 0.417 -.118** Interest in politics 0.216 0.102 4.502 1 0.034 1.241 .074* Secularism -0.178 0.136 1.724 1 0.189 0.837 -0.027 Hardline Islam 0.029 0.162 0.033 1 0.857 1.03 0 Family values -0.158 0.166 0.905 1 0.342 0.854 -0.006 Distrust of state 0.425 0.136 9.822 1 0.002 1.53 .151** Constant 0.674 1.339 0.254 1 0.615 1.962 **. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed). the summer of 2011 those who rejected conventional banks were significantly interested in politics. Elsewhere, in most of the Arab world except Yemen and Sudan where Islamic finance was relatively well established, potential customers seemed uninterested in politics, in Saudi Arabia significantly so (Table 7). In Egypt political interest was associated with secularism (r= 0.16, p<0.01) and opposed to hardline Islam (r=0.305, p<.01). The Egyptians who unconditionally rejected bank interest also distrusted the transitional government. Government initiatives, such as moving forward with the sukuk legislation, are driven by hopes of attracting more GCC funding. From the survey it seems there may also be potential clients at home who are interested in public affairs and apparently have little sympathy for Islamism in any of its political forms. Another possible side effect of further government initiatives could be to shore up Salafist support for the incumbent regime. Tunisia in Transition When promised important Saudi investments to drain and develop swamps by the Lake of Tunis, President Habib Bourguiba admitted an off shore Islamic bank as part of the deal. BEST Bank (renamed Al Baraka Bank Tunisia) even briefly acquired limited onshore status in 1988 but under President Ben Ali’s regime the bank was not permitted to open new branches. His son-in-law Mohamed Sakhr El Materi opened Ez-Zitouna Bank in 2010 as part of Ben Ali’s strategy to placate any potential Islamist opposition that had survived the vicious repression of the Islamist Nahda Party in 1991-1992, in the wake of the Gulf War. Table 1
19. shows that it immediately gained market share before the Tunisian Revolution of January 14 overthrew Ben Ali and the state took over the assets of his 113 family members and cronies , including the Zitouna Bank. The bank continued, however, to flourish, with market share increasing from 2012 to 2013. Despite the country’s turbulent transition to democracy as well as Zitouna’s dubious origins, Islamic finance steadily grew. The bank was in government receivership, yet its deposits increased by 45.7 per cent in 2013. 26 As in most of the Arab countries, a devout Table 11: Logistic Regression of Interest Rejection – Tunisia B S.E. Wald df Sig. Exp(B) correlates Urban -0.22 0.196 1.293 1 0.255 0.8 -0.049 Age 0.004 0.009 0.22 1 0.639 1.004 0.054 Female 0.228 0.198 1.331 1 0.249 1.256 -0.021 Education -0.05 0.068 0.52 1 0.471 0.952 -0.021 Income$100s -0.01 0.014 0.7 1 0.403 0.988 -0.052 Unemployed -0.53 0.178 8.945 1 0.003 0.588 -0.065 Married -0.1 0.235 0.171 1 0.679 0.908 0.034 Religiosity 0.317 0.118 7.269 1 0.007 1.373 .136** Trips West -0.02 0.112 0.018 1 0.894 0.985 -0.032 Interest in politics -0.04 0.107 0.135 1 0.714 0.961 0.013 Secularism 0.03 0.157 0.038 1 0.846 1.031 -0.015 Hardline Islam -0.09 0.198 0.208 1 0.649 0.914 0.007 Family values 0.22 0.213 1.061 1 0.303 1.246 .091** Distrust of state 0.312 0.121 6.632 1 0.01 1.366 0.043 Constant -1.6 1.156 1.918 1 0.166 0.202 **. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed). constituency seemed ready to support Islamic finance. The logistic regression reported in Table 11 shows that religiosity, coupled with being employed and distrusting state institutions, retained statistical significance in the multivariate analysis. Behavioral Religiosity trumps conservative Family values as well as Secularism and Hardline Islam; that is to say, none of the ideational variables is a significant correlate even if the others are removed from the model. In short, the Tunisians who might support Islamic finance seem quite removed from any partisan politics. Within the sample the 47 respondents who claimed they were planning to join the Nahda Party appeared a bit more hardline Islamist than the national average, but they were also less unconditionally opposed to conventional bank interest. During Tunisia’s political transition Islamic banking was not a political issue; better known as “participatory finance,” it seems likely now steadily to expand its market share. The less Nahda promotes it, the more promising its prospects may be for including religious Tunisians who otherwise reject banks altogether. 26 Zitouna deposits exceeded those of BEST (renamed Al Baraka) Bank by the end of the year. See APBT, 2012: 54 and 2013: 52.
20. In sum , the North African countries with the largest financial surfaces for which Arab Barometer survey data are available offer bright prospects for Islamic finance. Neither wealth nor gender turned out even in Algeria significantly to condition attitudes toward bank interest, so that Islamic banks may appeal across a broad range of populations in these countries. Everywhere it may find ready constituencies not particularly interested in politics but for whom family values and religion are important aspects of their daily lives. Portfolio diversification in Saudi Arabia Further analysis of the Saudi data may suggest why Islamic finance achieved its head start in the GCC countries. The hypothesis of portfolio diversification can be further examined in light of the ideational as well as personal characteristics already discussed. The logistic regression with the full set of variables produced surprising results. Table 12: Logistic Regression of Interest Rejection – Saudi Arabia B S.E. Wald df Sig. Exp(B) correlations Urban -0.23 0.384 0.36 1 0.549 0.794 -0.049 Age 0.032 0.01 10.932 1 0.001 1.033 .101** Female 0.06 0.247 0.059 1 0.808 1.062 0.001 Education 0.026 0.071 0.133 1 0.715 1.026 0.044 Income$100s 0.02 0.006 11.298 1 0.001 1.02 .127** Unemployed 0.075 0.226 0.112 1 0.738 1.078 -0.025 Married -0.22 0.254 0.739 1 0.39 0.804 0.018 Religiosity 0.771 0.204 14.278 1 0 2.163 0.012 Trips West 0.045 0.091 0.238 1 0.626 1.046 0.015 Interest in politics -0.42 0.098 18.597 1 0 0.656 -.210** Secularism -0.26 0.158 2.633 1 0.105 0.774 0.056 Hardline Islam -0.78 0.177 19.492 1 0 0.459 -.173** Family values -0.37 0.208 3.201 1 0.074 0.689 -.155** Distrust of state 0.023 0.16 0.021 1 0.884 1.024 .065* Constant 1.131 1.304 0.751 1 0.386 3.098 **. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed). Age, income, positive disinterest in politics, Hardline Islam, and family values all kept their statistical significance, and personal religiosity became a strong predictor of interest rejection, while any distrust in state institutions vanished in this multivariate analysis. As Table 7 presenting the ideational correlates of Interest Rejection suggested, the hardline version of Islam remained negatively correlated, holding the other variables constant.27 In other words, a more liberal Muslim was more likely to reject interest than a more hardline fundamentalist. Religiosity, expressed in daily behavioural routines, still strongly correlated with the rejection of interest when the effects of the other variables were held constant in the 27 When Secularism was removed from the model, hardline Islam and family values gained in statistical significance to the .000 and .033 levels of significance, respectively, with slightly smaller Betas of -.581 and .425.
21. multivariate analysis . But family values also worked in a surprising direction. More liberal family values, not conservative ones, were significantly correlated with rejecting interest. In short, the picture of the Saudi clientele depicted by the survey was one of wealthy and aging individuals, not particularly dissatisfied with state institutions but turned off of politics and basically liberal in outlook at least in the context of Saudi Arabia. Their broadly liberal outlook was perfectly compatible with high degrees of religiosity, or religious practices in daily life. Working with Islamic rather than conventional banks was perhaps part of their religious way of life, although in Saudi Arabia virtually the entire retail sector is sharia compliant. The survey sheds light on the individual motivations behind the high demand for sukuk and other Islamic financial instruments open to private investors. It seems to be the relatively wealthy investors from GCC countries who motivated the big international banks to expand their sharia-compliant menu of investment options. These in turn will finance the Saudi and other GCC treasuries dependent on oil revenues that continued their decline in 2015. Conclusion The biggest finding from the survey data is the broad rejection of conventional bank interest and consequent potential of Islamic finance in the region. Roughly half of the populations sampled have some affinity for one of the leading ideas behind Islamic finance, the rejection of making money out of money. And for the most part they have little interest in politics, much less political Islam. Before drawing conclusions about the apolitical nature of Islamic finance, however, it seems only fair to turn briefly to the other countries represented in the Arab Barometer survey to check for possible correlates of the potential Islamic finance constituencies with political variants of Islamism. Accordingly the logistic regressions for each country are Table 13: Logistic Regressions of Interest Rejection Jordan Lebanon Palestine Sudan Yemen Exp(B) Sig. Exp(B) Sig. Exp(B) Sig. Exp(B) Sig. Exp(B) Sig. Urban 1.049 0.786 1.471 0.391 1.369 0.034 1.518 0.029 1.232 0.255 Age 1.008 0.202 0.995 0.553 1.02 0.004 1.018 0.096 1.006 0.605 Female 0.727 0.063 1.5 0.037 1.347 0.113 0.628 0.028 1.551 0.03 Education 1.114 0.054 0.917 0.173 1.092 0.125 1.333 0.000 1.158 0.02 Income$100s 0.993 0.566 0.985 0.222 0.996 0.715 0.973 0.269 0.976 0.376 Unemployed 0.964 0.827 1.096 0.622 0.661 0.026 1.21 0.428 0.794 0.291 Married 0.923 0.608 1.087 0.69 0.74 0.132 1.422 0.183 1.32 0.207 Christian 0.535 0.194 0.169 0.000 0.299 0.077 Religiosity 1.403 0.003 1.278 0.021 1.382 0.008 1 0.998 1.115 0.533 Trips West 0.954 0.538 0.86 0.131 0.799 0.026 0.822 0.1 0.914 0.66 Interest in politics 1.007 0.931 1.053 0.558 0.996 0.962 0.996 0.973 1.226 0.042 Secularism 0.901 0.39 1.318 0.106 1.014 0.92 1.002 0.99 1.035 0.826 Hardline Islam 0.854 0.242 1.055 0.804 1.716 0.000 1.136 0.496 1.838 0.002 Family values 2.251 0.000 1.271 0.17 1.965 0.000 1.628 0.02 1.533 0.036 Distrust of state 1.06 0.586 1.182 0.128 1.021 0.789 1.209 0.122 1.491 0.001 Constant 0.138 0.044 0.223 0.232 0.071 0.018 0.024 0.005 0.003 0.000
22. briefly summarized in Table 13 , presenting the exponents of the B coefficients and their statistical significance.28 In this multivariate analysis Hardline Islam still retains statistical significance in the Palestinian and Yemeni samples. In Palestine, however, any putative association of hardliners with Islamic finance elicited little political interest. That is to say, interest in politics was statistically unrelated to interest aversion, whatever one’s understanding of Islam. Among the Palestinians, moreover, interest in politics was significantly related to secularism (r=.093, p<.002), and definitely not to Hardline Islam (r=-.090, p<.002). In this respect the Palestinians resembled the Egyptians discussed earlier. Only in Yemen does interest in politics go hand in hand with Hardline Islam, family values, and distrust of state institutions, constituting a powerful Islamist clientele for Islamic finance. Yemen, however, remains one of the most underbanked countries in the Arab region, despite Islamic banking’s 42 per cent market share (Table 1). While the survey cannot specify banking behaviours, its rich political data collected in the wake of the Arab Spring do help to place other information about Islamic banking in perspective. Except in Algeria, Yemen and possibly in Palestine the potential constituencies for these banks really seem to be apolitical. Algeria is potentially one of the largest financial surfaces in the MENA, and the government might benefit from channelling its interest in politics into Islamic finance. Its potential constituency generally supports state institutions, so that promoting Islamic finance might be one way of gaining support for restructuring and diversifying Algeria’s economy. Comparisons with Iraq are enlightening. Both oil economies, of similar overall GDP, have underdeveloped financial surfaces, the product of years of state bank monopolies as well as subsequent brutal war and violence. Table 3 indicated Iraq’s banking surface to be less than half Algeria’s. Yet Iraq’s Islamic banking has already gained an important foothold in its fragile commercial banking system, whereas Islamic banking has stagnated in Algeria. The survey suggests that it may be an important vehicle for recuperating Sunni support in Iraq. In Algeria’s underbanked system it might also offer greater political and social as well as financial inclusion. As for Tunisia, the very apolitical nature of its potential constituency for Islamic or “participatory” finance is a distinct advantage. Both Egypt and Tunisia witnessed revolutions being “hijacked,” in the view of many, by the Muslim Brotherhood and its Tunisian cousins. While the political transitions took different directions, one common outcome was increased polarization, in Tunisia as well as Egypt, between political Islamists and their religiously more neutral compatriots. By remaining carefully aloof from politics the banks have better chances of improving their market shares. Previous authoritarian regimes kept a tight lid on Islamic finance, for fear of any association between Islamic finance and Islamist political oppositions. But the Arab Barometer survey data clearly point to the absence of such a relationship. On the ground in Tunisia, moreover, Ben Ali’ cynical use of Islamic banking may have compromised it in the eyes of the Nahda Party. Its sympathizers in the survey showed less concern about conventional bank interest than their compatriots on average. In Egypt it was the Salafists rather than Muslim Brothers who seemed most concerned. And 28 Any Exp(B) below 1 connotes a negative B. Among the Jordanians, for example, being a female was negatively correlated with Interest Rejection: a female is (1-.727=) 27% less likely to reject bank interest. The most statistically significant correlates of interest rejection were Religiosity and Family values. Holding everything else constant, an increase of religiosity of one unit on a four point scale was associated with a 40% greater likelihood of rejecting conventional interest-based banking.
23. despite the political interest associated in Egypt with Islamic finance , the authorities need not fear any connection between financial and political Islam. The governments of Egypt and Tunisia are likely to encourage Islamic banking and finance not so much to satisfy domestic demand as to encourage more investment from the GCC countries. Here again, the survey results point to potential demand for sukuk and other forms of sharia-compliant investment from wealthy Gulf families. But as to whether Islamic finance may become really inclusive and self-sustaining across the Arab world, the surveys can only point to a vast potential clientele, part of an arc stretching from Sub-Saharan Africa across to Southeast Asia. As part of a community of practicing Muslims, principally defined in the survey by “religiosity” and “family values,” it may overcome the religious objections of some and bring more people out of informal economies. Better time series data are needed to understand the conditions under which Islamic finance may actually mobilize wealth that would not otherwise circulate and put it to constructive uses. References Association Professionelle Tunisiennes des Banques et des Etablissements Financiers (APTB), Rapports annuels, 2011, 2012, 2013. Ben Naceur, Sami, Adolfo Barajas, and Alexander Massara, 2015. Can Islamic Banking Increase Financial Inclusion? Working Paper 15/31 (February), Washington, DC: International Monetary Fund. Demirguc-Kunt, Asli, Leora Klapper, and Douglas Randall, 2013. Islamic Finance and Financial Inclusion: Measuring Use of and Demand for Formal Financial Services among Muslim Adults, Policy Research Paper Working Paper 6642, Washington, DC: World Bank. Ernst and Young (2013), World Islamic Banking Competitiveness Report 2013-2014: http://www.ey.com/Publication/vwLUAssets/World_Islamic_Banking_Competitivene ss_Report_201314/$FILE/World%20Islamic%20Banking%20Competitiveness%20Report%20201314.pdf (Accessed 21 May 2015) Harding, Andrew, 2010. Sharia and national law in Malaysia, in Jan Michiel Otto, ed., Sharia Incorporated: A Comparative Overview of the Legal Systems of Twelve Muslim Countries in Past and Present. Leiden, Holland: Leiden University Press.