Sub-component 2.1: PMT based Stipends and Tuition to the Poor Boys and Girls (Total: BDT63894, 00 lakh; US$ 92.6 million)
The objective of this component is to increase access and retention of poor girls and boys, and ensure their completion of secondary schooling through provision of stipends and tuition, based on poverty and educational criteria.
1. After a pioneering and highly successful gender-targeted Female Secondary Stipends Program, GoB has now decided to introduce a poverty-targeted stipend program to address large disparities across different income groups in secondary school access and quality. The experience from the long implemented female stipends program reveals that the usual beneficiaries of the program are mostly from the well off families and the poor are most unlikely to benefit from the stipends as they fail to meet the eligibility compliance criteria. During the Mid-term Review (MTR) of FSSAP-II a Pro-poor Self Targeting (PPST) scheme was introduced on a pilot basis to focus on both the poverty and educational criteria for being eligible for stipend. The objective of implementing the pilot PPST was to explore the feasibility of introducing a poverty targeted stipend program by using Proxy Means Testing (PMT), a targeting tool being used world wide. PMT is a household-based pro-poor targeting method that screens applicants into beneficiaries and non-beneficiaries through a set of proxy variables that can easily be measured and verified. Compared to a simple unverified scheme, PMT is more objective, transparent, and accurate. Measuring true welfare using actual consumption or income (direct means-testing) is very time consuming, costly and administratively infeasible in a country like Bangladesh. On contrary, PMT is administratively more feasible and cost efficient. The targeting approach based on means testing that FSSAP-II piloted and other MoE project has been implementing has a grave error of exclusion of potential beneficiaries. As this approach concentrates on a cut off line of beneficiaries with out taking into consideration the local poverty line, hence, it might create a sort of inequity in the society especially, among the poor. On the other hand, the advantage of PMT based targeting approach is that it includes all the target beneficiaries eligible for a social program and, thereby differentiates the non-poor from the poor.
2. In view of the above, both GoB and IDA have decided to implement a PMT based stipend program under the SEQAEP. It would be worth mentioning that the PMT based stipend approach is very much in the light of the PRSP that envisage to creating a social safety net for the poor. After successful implementation and effective evaluation of the program MoE will expand the PMT based stipend program through out the country.
3. PMT is a household-based targeting method that generates a welfare score for applicant individual/household based on a set of observable household characteristics. Proxy variables and their weights in the welfare score are determined using a statistical regression analysis of consumption expenditure of a detailed survey data on household welfare such as consumption expenditure .
Individual applicants will be ranked according to their respective household predicted welfare scores and their eligibility determined according to a threshold or cut-off score.
4. There are essentially two issues in the PMT approach:
(a) the selection of appropriate variables in the regression model; and
(b) the choice of cut-off point for eligibility. The choice of PMT proxy variables should ensure two main criteria:
(i) the approach sufficiently explains the variation in welfare levels across households so that it identifies the poor with reasonable accuracy (i.e. levels of exclusion and inclusion errors, and incidence of beneficiaries are acceptable ); and
(ii) the model variables are relatively few, are easy to measure and verify, but hard to manipulate. The proxy variables will be finalized in consultation with the experts in the field and approval by the project Steering Committee.
5. The primary objective of the proposed PMT-based stipend and tuition program is to maximize the coverage of the poor in targeting the secondary school students. Equally important is the need to keep the level of the stipend and tuition sufficiently high to attract and retain those poor children in secondary education. The annual stipend and tuition level is based on per student household education expenditure in a typical poor household , as estimated from HIES 2005 data. Grade-specific stipend rates, as shown in Table 2, are largely consistent with the projected rates in other pro-poor stipend programs in the country and in conformity with the country's fiscal sustainability. Tuition rates are based on existing rates prescribed by MoE.
SSC Exam Fees
6. The choice of the cut-off point for eligibility is essentially a policy decision (with aim to cover most poor students with the available budget). Based on a simulation of the PMT model with respect to different cut-off points, it is proposed that 50th percentile of the actual per capita consumption expenditure will form the cut-off score for eligibility for stipend and tuition program . According to this cut-off, the stipend program would benefit approximately one-third of all enrolled students in the initial years and an increasing share over time. As per GoB's decision to continue to provide at least tuition support to girls from "middle-poor" households, the second category of tuition-only beneficiaries will consist of these girls from households that lie between the 50th and 80th consumption percentiles.
7. Based on the grade-specific share of enrolments associated with these welfare cut-off scores and administrative data on enrolments from the FSSAP II project schools, it is estimated that the targeted number of students eligible for stipend and tuition and tuition only program will be as shown in Table 3. The projections also assume that: (i) the share of boys among total stipend and tuition beneficiaries will increase over time and (ii) the share of poor children in secondary school enrolment will increase over time. Both of these assumptions are, in fact, directly related to two of the key project indicators (KPIs).
Table 3– Projection of stipend and tuition and tuition only beneficiaries in PMT-based Stipends Program by Academic Year
|60 Upazillas||(121 Upazillas)|
|Category 1 – Beneficiaries of stipends and tuition (up to 50th percentile), Grades 6-10||281,000||588,000||612,700||637,500||662,400||2,782,100|
|Girls, Grades 6-10||154,800||317,500||324,700||331,500||337,800||1,466,300|
|Boys, Grades 6-10||126,700||270,500||288,000||306,000||324,600||1,315,800|
|Category 2 – Girl Beneficiaries tuition only (between 51st & 80th percentile), Grades 6-10||140,750||294,000||306,350||318,750||331,200||1,391,050|
8. Given that PMT initiative requires a new operational modality, several options were explored to identify a specialized agency to serve as a PMT administrator that would meet the following key criteria: (a) substantial presence with local level staff; (b) experience with large-scale and systematic data collection and data processing under time-bound deadlines; and (c) capacity for mobilizing the community at local level. In addition to these criteria the cost of administering the targeting program would be a prime consideration before taking up PMT based targeting program. The world wide implementation experience of PMT based pro-poor targeting reveals that the administering cost is about 8-16%. The feasibility of utilizing private firms was reviewed by IDA but given up because of the heavy cost that would come up. The feasibility of selecting BANBEIS and Education Engineering Department as the probable PMT Administrator was also reviewed but both the organizations do not have adequate field level staff. With an attempt to find out an organization with in the Government sector, LGED has been given a consideration for responding to the criteria mentioned above. LGED has a reasonable number of staff at Upazilla level and familiar with the task of community mobilization. At central level LGED has a strong MIS with sound expertise and has the experience of dealing with voluminous data. For the last few years LGED has been administering the Reaching out of School Children (ROSC) project under Primary and Mass Education.
9. A Participation Agreement will be signed between SEQAEP and LGED delineating the responsibilities of PMT Administrator, SEQAEP and the roles of DSHE and MOE as the executing and sponsoring agencies respectively. LGED will deliver services as PMT Administrator for which they will be provided with administering cost amounting to about 5% of the total cost allocated for the sub-component. The Participation Agreement would be made on a yearly renewable basis subject to full satisfaction of DSHE/MOE.
10. PMT Administrator will start its assignment from July 2008 and continue through December 2013. However, the volume of PMT processing to select the potential beneficiaries will be front-loaded in the first two years – the first year in 60 Upazillas and the second year in the remaining 61 Upazillas. Once selected through the PMT scheme, a beneficiary is expected to be eligible for stipends and tuition for the entire duration of his/her secondary school education provided s/he meets the educational compliance criteria.
11. Given the complexities involved in the PMT scheme, a PMT-based Stipends Operation Manual (both in English and Bangla) would be disseminated to stakeholders at all levels. This Manual will include details of the PMT concept, criteria and procedures for selecting the poor, beneficiary selection, stipends and tuition rates, disbursement procedures, timeliness, roles and responsibilities of all concerned, flow of funds, Participation Agreements (with the PMT administrator, schools, disbursing bank, etc.), application and confirmation forms. Besides, in order to implementing the activities under this sub-component training, workshop and orientation sessions will be organized for the concerned stakeholders.