Abstract
This paper examines traditional Islamic school (i.e. madrasah) attendance as a determinant of social attitudes among secondary-schooled adolescents in rural Bangladesh. Although both recognized and traditional madrasah-enrolled adolescents show greater support for charity, we find no evidence that traditional madrasah attendance promotes “trust in others”. Attendance at recognized madrasahs, which use a state-approved curriculum, however, significantly increases social trust. The madrasah–trust connection is not explained by the role of teachers’ attitudes towards trust or professional background. Nor do we find evidence that the absence of an effect of traditional madrasah attendance on social trust is driven by their geographic locations. Given the differences between the two groups in terms of stated social preferences, our analysis warns against the practice of lumping state-recognized madrasahs and traditional madrasahs into one category.
Notes
I gratefully acknowledge the financial support from the Leverhulme Trust, UK. The first draft of the paper was written when I was visiting the Research and Evaluation Division (RED) of BRAC, the largest NGO in Bangladesh, as a Distinguished Fellow. The usual disclaimers apply.
1. The effects of religion and religiosity on educational attainment and other economic and demographic outcomes have received considerable attention in the social science literature (e.g. see Chang, Citation2010; Gruber, Citation2005; Lehrer, Citation2009; Lenski, Citation1963; Weber, Citation1930). See McCleary and Barro (Citation2006) for a review of the literature on how religiosity affects individual characteristics such as work ethic, honesty and thrift, and thereby influences economic performance.
2. There are similar studies (albeit using survey data) on the impact of faith schooling on civic participation, values and the behaviour of youths; see for instance Dee (Citation2005) and Hofmann-Towfigh (Citation2007).
3. For evidence on the low level of student learning in rural Bangladesh, see Asadullah and Chaudhury (Citation2015).
4. Two studies have formally examined the determinants of learning outcomes in recognized secondary schools and madrasahs in rural Bangladesh. Asadullah et al. (Citation2007) assessed performance of grade 8 students in mathematics using a sample of 321 recognized secondary madrasahs and schools. The World Bank (Citation2010) used a similar sample but additionally tested children in English.
5. The evidence of gender-based interactions between students and teachers (e.g. role-model effect) is well documented in the developing countries literature. The most recent evidence is by Dee (Citation2007) who used longitudinal data from the USA and concludes that assignment to a same gender teacher positively influences student achievement and student engagement.
6. For evidence that madrasah attendance has a positive effect on trust behaviour, see Delavande and Zafar (Citation2015). Using a field experiment in Pakistan, the authors find that madrasah students expect others to be trustworthy. Sample respondents come from four bachelors-equivalent degree-awarding madrasahs and three non-madrasah higher education institutions (an Islamic university and two liberal universities).
7. In Bangladesh, a union is an administrative unit larger than a village but smaller than a sub-district.
8. In the case where multiple sections (or classrooms) in grade 8 existed, one section was randomly chosen.
9. For further details on the study design, see World Bank (Citation2010).
10. For a discussion on the relative goodness of survey-based measures of trust, see Sapienza, Toldra‐Simats, and Zingales (Citation2013) and Thöni, Tyran, and Wengström (Citation2012).
11. Such controls for socio-economic–demographic characteristics are common in studies on the determinants of adolescent attitude and behaviour (see, for instance, Donahue & Benson, Citation1995).
12. Indeed, using individual-level data drawn from the US localities, Alesina and La Ferrara (Citation2002) find that a recent history of traumatic experiences is one of the strongest determinants of low trust.
13. We do find madrasah students to be more knowledgeable in religious matters and this varies among youths across school types. However, religious education is also a proxy of a student’s cognitive capacity and hence is not used as a measure of religiosity.
14. There is an economics literature on the intergenerational transmission of cultural traits which explains acquisition of preferences by children through an adaptation and imitation process that depends on their parents’ socialization actions and on the cultural and social environments in which children live (see e.g. Bisin & Verdier, Citation2001; Loehlin, Citation2005).
15. However, teacher attitudes do differ significantly by school type, for example, in the case of “support for charity” where traditional madrasah teachers appear to be most supportive of the redistribution of wealth.
16. This finding contrasts with Chang (Citation2010) who finds a negative relationship between religious attendance and preferences for redistributive policies in Taiwan.
17. Examples of studies that don’t distinguish between recognized and unrecognized madrasahs include Armanios (Citation2003) and Ciovacco and Wazed (Citation2008).