Bangladesh Social Protection Analysis
Coverage, Targeting, and the Path to a Lifecycle System
BDPolicyLab · 2026-07-05
Bangladesh runs more than 140 social safety net programmes, yet half of its poorest households still get nothing. The World Bank Poverty and Equity Assessment 2025 finds that only half of the poorest consumption quintile received any social assistance in 2022, while 35 percent of the richest quintile did, and that the programmes' progressivity fell over 2010 to 2022 as benefits leaked to better-off households. Spending is not the binding constraint: the Centre for Policy Dialogue puts the FY2026-27 social safety net budget at 2.11 percent of GDP, in line with regional peers. The constraint is targeting, benefit adequacy, and an Old Age Allowance that covers 6.2 million people on a Tk 700 monthly benefit set program-wide rather than by need. The lifecycle consolidation prescribed by the National Social Security Strategy fixes fragmentation, not exclusion. This brief sets out three moves that do: a single registry that decides eligibility, automatic indexation of flat-rate allowances, and registry-verified migration before any scheme consolidation.
Key findings
- Only half the poorest quintile gets support, while 35 percent of the richest does. World Bank Bangladesh Poverty and Equity Assessment 2025 (using BBS HIES 2022) finds that only half of the poorest consumption quintile received social assistance in 2022, against 35 percent of the richest quintile. Transfers more than doubled between 2010 and 2022, but their progressivity declined as benefits increasingly reached better-off households.
- Spending is adequate at 2.11 percent of GDP; the problem is targeting, not money. CPD's Analysis of the National Budget for FY27 (June 2026) puts the social safety net budget at 2.11 percent of GDP in FY2026-27, and at 1.40 percent of GDP once pension is excluded. Spending sits broadly in line with South Asian peers, so the binding constraints are who is selected and how much they receive.
- The system is fragmented across roughly 145 programmes with no common registry. Bangladesh runs close to 145 ongoing social safety net programmes across more than 20 ministries (Unnayan Onneshan, Rapid Assessment of the Budget FY2019-20; World Bank Country Partnership Framework FY16-20). The National Social Security Strategy prescribes consolidation along the lifecycle, but a household can still draw several benefits while an equally poor neighbour draws none.
- Old Age Allowance: 6.2 million recipients at Tk 700 per month, set program-wide not by need. The National Budget Speech FY2026-27 (11 June 2026) raised Old Age Allowance coverage to 6.2 million and the monthly benefit by Tk 50 to Tk 700. The benefit is a flat program-wide rate with no statutory inflation indexation, and the Planning Commission's analysis of HIES data found about 33 percent of recipients to be below the eligibility age, evidence that selection, not headcount, is the weak point.
The single fact that should govern social protection policy in Bangladesh is this: in 2022 only half of the poorest consumption quintile received any social assistance, while 35 percent of the richest quintile did, and over 2010 to 2022 the programmes became less progressive as benefits increasingly reached better-off households (World Bank Bangladesh Poverty and Equity Assessment 2025, using BBS HIES 2022). That is a targeting failure, not a budget failure. The Centre for Policy Dialogue puts the FY2026-27 social safety net budget at 2.11 percent of GDP, and at 1.40 percent once pension is stripped out (CPD, Analysis of the National Budget for FY27, June 2026), broadly in line with regional peers. Money is reaching the system. It is not reaching the right people.
The conventional reform answer, consolidating the roughly 145 programmes along the lifecycle as the National Social Security Strategy prescribes, addresses the wrong failure. Consolidation cuts duplication and coordination cost. It does not, on its own, enrol the poor household the union list left out, lift a benefit that has no statutory floor, or stop better-off households capturing transfers. Three changes do: a single registry that decides eligibility from survey-anchored data, automatic indexation of flat-rate allowances, and registry-verified migration before any scheme is folded into another.
The architecture causes duplication, not exclusion
Bangladesh runs close to 145 ongoing safety net programmes across more than 20 ministries (Unnayan Onneshan, Rapid Assessment of the Budget FY2019-20; World Bank Country Partnership Framework FY16-20). With no common registry, a household can draw several benefits while an equally poor neighbour draws none. Successive National Social Security Strategy action plans have prescribed the fix: a single registry management information system, sourced from a national household database, that allows cross-verification of eligibility (NSSS Action Plan Progress Report 2024, MoSW). The FY2025-26 social security budget reports a central database with single-registry features, including a beneficiary management module, in development (Social Protection Budget Report FY2025-26, MoSW).
The implication is precise. Consolidating programmes along the lifecycle cuts duplication, which is the inclusion-error problem. It does not reach the half of the poorest quintile who receive nothing, because exclusion is driven by stale beneficiary lists and nomination capture, not by the number of programmes (World Bank Bangladesh Poverty and Equity Assessment 2025). A live registry that re-runs eligibility against survey-anchored data is the operative fix; lifecycle consolidation is the housekeeping around it.
Benefits are flat-rated with no inflation floor
The Old Age Allowance reaches 6.2 million recipients and was raised by Tk 50 to Tk 700 a month in the FY2026-27 budget (National Budget Speech FY2026-27, 11 June 2026). The benefit is a single program-wide rate revised only by discretionary budget decisions, with no statutory indexation rule. The CODI Country Report 2024 tracks the same allowance nominal versus real value over time and documents how the gap widens whenever inflation outpaces nominal revisions (CODI Country Report 2024, MoSW). A transfer with no automatic floor loses real value between revisions even when it reaches the right household.
The implication: the cheapest high-impact reform is not a new programme but an indexation rule. Annual automatic adjustment to the consumer price index holds the real benefit constant at near-zero administrative cost and removes the recurring political bargaining that lets benefits drift in real terms between ad hoc increases.
Selection, not delivery, is where transfers leak to the wrong households
The delivery layer is far along: of 30 cash-based programmes, 29 are now under the Government-to-Person digital payment system, routed through mobile financial services, electronic fund transfer, and the National Payment Switch (Social Protection Budget Report FY2025-26, MoSW; NSSS Action Plan Progress Report 2024, MoSW). Digital delivery narrows leakage in transit. It does nothing about who is selected. The Planning Commission's analysis of HIES data found about 33 percent of Old Age Allowance recipients to be below the eligibility age, and the World Bank finds richer households capturing a rising share of transfers (World Bank Bangladesh Poverty and Equity Assessment 2025).
The implication: pushing payments fully digital is a payment-integrity reform, worth completing, but it cannot correct selection. Whether a transfer reaches a poor or a non-poor household is decided at enrolment, not at disbursement. Only a registry that screens eligibility against survey-anchored data closes that gap.
Recommendations
1. Finance Division and Statistics and Informatics Division: make the single registry the system of record for all consolidated programmes, with a December 2027 deadline. Tie every programme's payment file to the national household database with mandatory cross-verification, so the registry, not a union-level list, decides eligibility. Expected effect: directly attacks the exclusion that leaves half the poorest quintile uncovered (World Bank Bangladesh Poverty and Equity Assessment 2025). Success signal: published share of bottom-quintile households enrolled rising above 50 percent in the next HIES round.
2. Ministry of Finance: legislate annual CPI indexation of the Old Age Allowance and other flat-rate allowances, starting FY2027-28. The Tk 700 benefit has no statutory floor and is revised only by discretionary budget decisions (National Budget Speech FY2026-27). Automatic indexation halts the real-value erosion the CODI Country Report 2024 documents, at near-zero administrative cost, and ends multi-year drift in real terms. Success signal: real benefit value held flat year on year in the CODI nominal-versus-real series.
3. Department of Social Services: re-verify Old Age Allowance enrolment against National ID age and the registry before the next disbursement cycle. About 33 percent of recipients appear to be below the eligibility age (Planning Commission analysis of HIES data), so age and means screening from verifiable records removes ineligible recipients and frees fiscal space for excluded poor elderly. Success signal: below-age-eligibility share in the recipient roll cut to single digits in the next DSS audit.
4. Cabinet and Finance Division: sequence lifecycle consolidation behind the registry, not ahead of it. Migrate programmes into consolidated lifecycle categories only once their beneficiary files are registry-verified. Expected effect: consolidation removes duplication without re-importing the stale lists that produced exclusion, so the reform fixes targeting rather than relabelling it. Success signal: no programme folded into a lifecycle category before its beneficiary file passes registry cross-verification.
What would change this view
Two facts would shift the assessment. First, if the single registry reaches operational coverage across the major cash-transfer programmes with survey-anchored re-verification, and the next HIES shows bottom-quintile coverage rising and richer-household capture falling, then the targeting diagnosis is being resolved and the registry mandate, not new spending, is doing the work. Second, if an audited series shows flat-rate benefits holding real value without a statutory indexation rule, the case for legislated indexation weakens. Absent both, the conclusion stands: spending is adequate, delivery is improving, and the binding constraints are targeting, adequacy, and the discretion in who gets selected.
Sources: World Bank Bangladesh Poverty and Equity Assessment 2025; BBS HIES 2022; CPD, Analysis of the National Budget for FY27 (June 2026); National Budget Speech FY2026-27 (11 June 2026, Ministry of Finance); Social Protection Budget Report FY2025-26 (MoSW); NSSS Action Plan Progress Report 2024 (MoSW); CODI Country Report 2024 (MoSW); Unnayan Onneshan, Rapid Assessment of the Budget FY2019-20; World Bank Country Partnership Framework for Bangladesh FY16-20.
Data and methodology
Coverage and targeting figures are drawn from the World Bank Bangladesh Poverty and Equity Assessment 2025, which analyses BBS Household Income and Expenditure Survey 2022 (HIES 2022). Social safety net spending shares are from the Centre for Policy Dialogue's Analysis of the National Budget for FY27 (June 2026). Old Age Allowance coverage and benefit levels are from the National Budget Speech FY2026-27 (Ministry of Finance) and the CODI Country Report 2024 (MoSW). Programme counts are from Unnayan Onneshan's Rapid Assessment of the Budget FY2019-20 and the World Bank Country Partnership Framework FY16-20.
Cite this
BDPolicyLab Research. (2026). Bangladesh Social Protection Analysis. BDPolicyLab. https://bdpolicylab.com/publications/bangladesh-social-protection-analysis