Bangladesh: Research Overview, BRAC-IED

In Bangladesh studies of communities and schools (ComSS) in six locations, one from each administrative division in Bangladesh, were conducted over a three year period with two rounds of data collected in 2007 and 2009 from households and schools in order to gain insights into the dynamics of participation and exclusion of children in schooling. The study covered 6,696 households with 9,045 children age 4-15 years from 18 school catchment areas (12 government primary schools and 6 registered non government primary schools).

The survey was designed to explore access to basic education in these districts using CREATE’s conceptual model of ‘zones of exclusion’. A baseline survey was conducted in 2007 and after two years, in 2009, a follow up survey was done with the same households. The survey identified children who have been excluded from education in both 2007 and 2009 who are in ‘zone 1’ – never enrolled. We also identified children who dropped out from primary school (zone 2), and those who were enrolled in 2009 but not in 2007. Those enrolled but attending irregularly, repeating years of schooling and learning little (zone 3) were also identified as were those not transitioning to lower secondary level after completing primary schooling (zone 4).

Sampling was driven from all households from within each district, and children were tracked back into schools in the districts selected. It was not possible to follow children who attended schools outside the case study districts.

Some Findings

The ComSS confirms that non-enrolment of school-age children (exclusion zone 1) remains a significant problem. Official national statistics indicate a non-enrolment rate of around 10% within the primary age group. The data from the sites was consistent with this though there were difficulties in establishing non-enrolment rates definitively in specific communities – some apparently out of school children migrated, others were not declared by households, others were temporarily out of school etc. Children who did not enrol at all were more likely to be disabled and less likely to play normally than children in school and those who dropped out. These excluded children were from poorer families, both economically and educationally, than families of children who dropped out (Ahmed and Hossain, 2010).

The ComSS highlights the extent to which dropout after enrolment (exclusion zone 2) is a critical problem both at primary and secondary level. Dropout is associated with poverty, reflected in the food-security status of families, parental education and household income. The capacity of poor households to support and guide their children through schooling is limited and linked to drop out. Issues around how education is delivered (school-related factors) are also crucial. Schools serving children from poor households fail to compensate or remedy disadvantage associated with poverty and themselves have few resources to do so.

Stipend income is directed to households rather than schools which have minimal non-salary budgets. Those who dropped out were on average older, came from lower income families, had parents with lower levels of education, had more household responsibilities, were in poor health, suffered from malnutrition and received significantly less support from parents for their school work (Sabates et al, 2010).

Poverty, low levels of parental education and unskilled parental occupations are associated with late entry into education and repetition of years of schooling. Children who start school late and repeated school grades become overage and the more over age children were, the more likely they were to drop out (Hossain, 2010; Sabates et al., 2010).

Around half of the children who dropped out were participating in rural-urban migration, as a livelihood strategy for families. Child migration on this scale has important implications for policy and strategy regarding access, continuation in school, completion of the primary stage and transition into secondary school (Ahmed and Hossain, 2010).

The category of silent exclusion or zone 3 is difficult to quantify with precision. ComSS data indicate that this category that includes over age, repeaters, poor attenders and low achievers, at a minimum comprises 20-25% of primary students depending on the definition used. Silent exclusion is also associated with the poor who cannot afford school materials or pay for private tuition (Hossain and Zeitlyn, 2010; Ahmed and Hossain, 2010).

About 24% of the children who completed Grade 5 in the sample did not enrol in Grade 6, the first year of the secondary stage. This number, however, underestimated the zone 4 (transition to secondary) problem. Drop out in Grade 5 before transition was about a quarter of all drop out from Grades 1 to 5 with substantial numbers leaving before completion This implies that the transition rate in the sample is even lower than it appears to be. Slightly more than half of all children successfully transit into secondary school (Ahmed and Hossain, 2010).

About 60% of the not enrolled children were in the 6-8 years age group. Some may enrol later but their chances reduce rapidly as their age increases. The culture of enrolling children in school consistently at age six is not embedded in the sample communities. The absence of birth registration and birth records also encourages a casual approach to the age for starting school. The consequences of late enrolment are manifested in dropout in later years, and high opportunity costs for school attendance as older children are seen as ready for paid or unpaid work. For girls, there is also concern among parents about safety and security of older girls walking to school and family and community pressure to marry (Hossain, 2010; Ahmed and Hossain, 2010).

Across the board there were clear relationships between non-enrolment, dropout, non-transition and socio-economic status, represented by the food-security status of families, household income, and parent’s education, as might be expected. Two-thirds of the never-enrolled children were from families with “always in food deficit” (ultra-poor) and “sometimes in food deficit” (poor) status in respect of staple grains supply for family. In the case of dropouts, 55% of the children came from households with food-deficit, though 45% of the population were in this category (Hossain and Zeitlyn, 2010).

A large proportion of school age children suffer from health problems (about a quarter of children of school age, in and out of schools, sampled were sick in the previous 30 days). When ill health or episodes of sickness are combined with other disadvantages of children prone to be in exclusion zones, their chances of effective participation in education are further diminished. Their quality of educational experience ultimately suffers as children with health problems often enrol in school late, have high rates of absenteeism, lower cognitive development and increased risk of drop out (Hossain and Zeitlyn, 2010; Ahmed and Hossain, 2010).

Household perception of reasons for non enrolment and dropout suggested supply-side constraints: schools are located too far from homes, and school education is perceived as of little value to children. This varies by school and community and perceptions and preferences change. There is some evidence of a drift away from government schools to registered non-government schools and madrasas (Ahmed et al 2010). This may reflect dissatisfaction with standards and the quality of teaching.

Some Policy Messages

• Absence of birth registration leads to major uncertainties in statistics regarding enrolment, completion and dropout necessary for proper planning and management of the system. Measures should be taken through local government agencies to enforce birth registration policies. Retro-active birth registration of 5-6 year olds should be undertaken as part of the government plan to bring all 6-year old children into school by school-year 2011. All children should be located and encouraged to attend schools from the age of 6 years by local authorities who should be obliged to monitor and report on their actions and outcomes.
• Child migration with or without other household members, has been identified as an important dimension of school dropout. Its nature and prevalence needs further research to understand the scale of the phenomenon, the reasons for migration, and the activities of migrants, and the access to school of migrant children. There is a very limited amount of research on child migration in Bangladesh (exceptions are Giani, 2007 and Heissler, 2008). Migration is likely to increase with urbanisation and currently there is no systematic policy on how to deliver educational services to migrant children.
• The peaking of dropout in Grade 5 and reduced primary completion associated with the newly introduced public examinations which encourage some schools not to enter low achieving candidates need serious attention. The supply of books and school materials need to be improved so that all children have books, and essential school equipment, since lack of these is associated with drop out and poor performance.
• Silent exclusion, where children attend but learn little, is clearly a serious problem affecting a large proportion of children, probably more than a quarter of the students in primary schools identified in this study. Agreement is needed on minimum thresholds for silent exclusion e.g. less than 90% attendance, more than two years over age, achievement tow years or more below the norm for the grade. More research is needed to detail and sensitise schools and teachers to the extent to which children are failing to learn and the reasons why. More school based work is needed to further investigate the processes and practices of silent exclusion in different locations. Responses to address such exclusion have to be based on school and community-based actions informed by evidence on its prevalence and persistence. Upazila-based planning and management must be sensitised to the issues.
• Clear relationships have been found between exclusion from education and socio-economic variables, represented by the food-security status of families, household income, and parent’s education. Stipends to poor students (conditional cash transfer) in primary schools linked to PEDP II remains a government strategy to promote equity in educational access. The supply-side constraints perceived by parents, and poor targeting of stipends which mean that the poorest may not receive support, suggest that the funds spent for stipends could be better used in providing essential quality-enhancing inputs in schools, including school meals. This question needs to be examined rigorously, especially because of the major budget implications of choices made (Hossain and Zeitlyn, 2010).
• The rapidly growing population of poor slum dwellers have been under-served by both governments and NGOs. There is a subset of households who are extremely poor (roughly, the poorest quintile) and for whom even low private education costs are prohibitive. The rapidly growing poor slum dwellers have been under-served by both governments and NGOs (Cameron, 2011). In areas of high urban migration which are often illegally settled, public provision of schools is largely lacking. Expanding services for the urban poor, including subsidies and other support is essential to maintain high primary enrolments and to reach the UPE goal (Cameron, 2010).
• NGOs, given their record in providing complementary and alternative educational opportunities, should be supported to target educationally disadvantaged areas and groups, and to design and offer inclusive and responsive approaches to under-served populations within the framework of area-based and coordinated programmes.
• Substantially greater public resources should be committed within the framework of the sixth five-year plan and the new education policy in order to assure minimum necessary levels of quality with equity. Bangladesh allocates a smaller proportion of GDP to education than most low income countries and this constrains progress towards universal access.