In this blog, Fazley Elahi Mahmud discusses Bangladesh’s National Social Security Strategy (NSSS) and its aim to provide lifecycle-based social protection. Despite progress, many eligible poor are excluded. He examines the shortcomings of targeted social protection and advocates for a shift to universal coverage, offering strategies and fiscal considerations for making this transition.
Ensuring the right to social protection is a constitutional obligation in Bangladesh. To fulfil this obligation, the Government of Bangladesh adopted the National Social Security Strategy (NSSS) in 2015, aiming to establish a lifecycle-based social protection system. The objective is to provide support to all deserving citizens in addressing risks encountered at various stages of life, including pregnancy and early childhood, school age, working age, and old age (Figure 1). While significant progress in implementing life cycle programs has been made in recent years, there remains a high level of errors in the selection process. According to 2016 Household Income and Expenditure Survey (HIES) data, an estimated 46.5 per cent of those who received benefits were wrongly included, while, more worryingly, 71 per cent of the eligible poor were excluded from the benefits.1 A recent BRAC study indicates an even more troubling concern that the poorest households tend to be disproportionately excluded compared to those who are relatively better off.2
| Figure 1: Key life cycle programmes of Bangladesh |
The current coverage is sufficient to reach all poor households
Coverage of life cycle programmes has been expanding significantly in recent years. The seven programmes shown in Figure 1 provide ongoing transfers for lifecycle phases to approximately 40 million individuals, nearly one-fourth of the population.3 When the stipend programme is discounted, the other six programmes reach 19 million individuals.4 These programmes are the Mother and Child Benefit Programme (MCBP), the disability allowance, the seasonal food security programme, the Vulnerable Women’s Benefit (VWB) programme, the Widow Allowance programme and the Old Age Allowance (OAA) programme.
Currently, approximately 7.7 million households in Bangladesh live below the poverty line.5 If each household were to receive benefits from only one of these programmes, they could reach all the households below the poverty line three times over. However, families having multiple members eligible for social protection, e.g., two elderly persons, can access more than one programme. Even if, on average, each family accesses two programmes, the coverage is more than sufficient for all households below the poverty line.
The poorest are disproportionately excluded
The BRAC study examined the access of newly enrolled participants in its Ultra Poor Graduation (UPG) programme6 to the aforementioned six social protection programmes across four Northern districts which are prone to climatic shocks, particularly drought. Typically, UPG participants represent the poorest households in the project villages, meticulously selected through community-based wealth ranking and verified through house-to-house surveys.7 Consequently, UPG participants are presumed to be eligible for national social protection programmes. Despite sufficient coverage, the study revealed that a staggering 75% of UPG participants did not access any of these programmes, indicating the disproportionate exclusion of the poorest members of society compared to the relatively less poor or non-poor. While the study employed a qualitative methodology alongside a quantitative survey among a small participant group, its findings align with national data on exclusion errors and international evidence on higher exclusion among the poorest.8 Further, focus group discussions (FGDs) within extremely poor communities and a review of related literature indicate the following causes of exclusion, which support the above finding.
i. Poverty-targeted selection approach: Poverty-targeted approaches measure household poverty, often using the Proxy Means Test (PMT) method to select the recipients of social protection. This approach is severely flawed, leading to significant exclusion errors. Kidd and Athias (2020) evaluated the selection outcomes of 19 PMT-based programmes, revealing exclusion error rates ranging from 29 to 96 per cent. They found that renowned programmes such as Indonesia’s Programme Keluarga Harapan, the Philippines’ Pantawid Pamilyang Pilipino Programme, and Pakistan’s Benazir Income Support Programme had exclusion error rates of 82 per cent, 48 per cent and 73 per cent, respectively. In Bangladesh, the problem is compounded by a lack of capacity of the local government units responsible for selection, lack of monitoring and lack of accountability. Additionally, the selection indicators are not effective as they have not been regularly adjusted in line with socio-economic improvements. For instance, the annual income threshold for the Old Age Allowance (OAA) programme is only BDT10,000 (US$90), significantly lower than the per capita annual income of those living below the poverty line, which is BDT42,936 (US$ 386).9 As a result, only 12% of OAA recipients fulfil the income eligibility indicator.10 In theory, these indicators should facilitate the selection of the most impoverished but, in practice, the ineffectiveness of the indicators grants local government members discretionary power to select the recipients.11 This practice perpetuates patronage and nepotism, favouring those who are better connected with the local government members, thereby depriving the poorest.
Ideologically, the poverty-targeted approach for selection conflicts with Bangladesh’s rights-based approach. This poverty approach perceives social protection as a means for the poor to survive rather than viewing it as an entitlement. Besides its high exclusion errors, this approach also proves incompatible with the lifecycle approach from an operational perspective. By determining eligibility based on household-level poverty, it overlooks vulnerable individuals within families, whereas the lifecycle approach extends support to individuals, such as persons with disabilities and older people.
ii. Lack of social capital: The poorer the families, the weaker their social capital, including their social networks – as well as cooperation and trust – which play a vital role in influencing the decision of local government members regarding the selection of social protection schemes. Sometimes, the local government members promise to help extremely poor families by submitting applications for them while taking copies of their national identity cards, yet those families often never receive the benefits. The poor families lack the ability to lobby or protest. Due to their weak networks, they cannot find any influential person to advocate for them or hold the local government accountable. Many poor families lack trust in the system due to the limited likelihood of positive outcomes if they apply. Therefore, they prefer not to incur transportation costs and sacrifice wages for the day to visit the local government office to apply. Isolated indigenous people demonstrate another glaring example of exclusion of the poorest, due to a lack of social capital.
Time for universal social protection
The universal approach inherently considers social protection as entitlements. For example, providing pensions to all older people would recognise their right to a pension, as opposed to providing it as charity through a poverty-targeted approach. Evidence shows that exclusion errors in the universal approach are very minor. According to Kidd and Athias (2020), universal programmes such as Georgia’s Old Age Pension, Mongolia’s Child Money programme, Bolivia’s social pension and a school stipend programme had exclusion errors, ranging only from 1 to 8 per cent.12 Bangladesh has made strides toward implementing universal social protection by adopting the universal Primary School Stipend Programme, which is provided to all children at primary schools regardless of their economic status. Data13 shows that the disability allowance programme reached a coverage exceeding the number of persons with disabilities estimated by the 2022 population census. It is time for the programme to offer the benefits universally to anyone who qualifies due to their disability, regardless of economic conditions. In line with the NSSS recommendation, the Government introduced a contributory pension scheme for both the formal and informal sectors in 2023. A logical next step would be to make the OAA programme universal to be complemented by the contributory scheme meant for the non-poor. Next, the Mother and Children’s Benefit Programme (MCBP) needs to be considered for universal access, to help close the lifecycle gaps. Options for universal benefits to vulnerable working age groups such as widows and single women, and seasonally unemployed day labourers need to be explored to ensure that all deserving persons are included.
It is worth noting here that a conceptual problem lies in the NSSS and other policy documents around the term ‘universal’. By ‘universal’ these documents mean ‘universal for eligible persons’, which led to the adoption of the term ‘targeted universal’ for selection. This term does not make any sense because ‘targeted’ and ‘universal’ are opposite to each other. The use of the term contributed significantly to the errors in selection. The current NSSS will complete ten years in 2025, with a new strategy expected in 2026, which must overcome this conceptual limitation. Launching a lifecycle system was a remarkable achievement of the first NSSS. The next phase of the NSSS could be marked by launching a universal (not ‘targeted universal’) social protection system.
Can Bangladesh afford a universal approach?
With social protection coverage having moved forward progressively, Bangladesh stands poised to transition towards a universal social protection system, through a phased strategy and budget restructuring. For example, the OAA could be offered universally at a higher age threshold initially, perhaps at 70 or 75 years. Similarly, the MCBP could be offered universally from pregnancy to a child’s first year, with the age threshold gradually extending to 4 years. Alternatively, universal access could be introduced exclusively in rural areas during the initial phase. With regard to fiscal space, maintaining the current allocation level of 2.5 per cent of GDP and reforming schemes would allow significant funds to be redirected towards achieving universal coverage. For instance, phasing out the pensions for retired government employees and interest subsidies for savings certificates from the social protection budget could potentially suffice for universal coverage of the OAA and the MCBP. A thorough analysis of cost options is imperative to inform the strategy to implement universal social protection.
Fazley Elahi Mahmud is an international development professional with 30 years of experience in Bangladesh and Indonesia. His expertise spans social protection, livelihood development, climate change, and disaster management. Currently, Fazley is an independent consultant on social protection.
[1] PRI (2019). Policy Insights.
[2] Mahmud, E. F. et al. (2021) Policy Research on Climate Vulnerable Indigenous and non-indigenous Poor Population Living in the Barind Tract of Bangladesh, BRAC, 2024.
[3] Government of Bangladesh, Ministry of Finance (2023). Social Security Programmes: 2023-24 (Budget).
[4] Ibid.
[5] Calculated based on HIES data on poverty and the national census (2022) data on population.
[6] BRAC. Overview Ulra-poor Graduation Programme, Bangladesh.
[7] UPG programme is a comprehensive livelihood programme and strives to link with national social protection programmes. It is a highly targeted programme. The reference to UPG programme is made in this blog to use the study finding as an indicator of exclusion of extremely poor people from national social protection programmes. Details of the programme are not discussed.
[8] Kidd, S. and Athias, D. (2020). Hit and Miss: An Assessment of targeting effectiveness in social protection.
[9] Government of Bangladesh (2022). Household Income and Expenditure Survey.
[10] World Bank (2021). Bangladesh Social Protection Public Expenditure Review.
[11] Ibid.
[12] Ibid.
[13] Government of Bangladesh, Ministry of Finance (2023). Social Security Programmes: 2023-24 (Budget).