Uneven by Design: How Bangladesh Delivers Welfare Through Programmatic Asymmetry

Bangladesh is one of the most centralized states in South Asia, yet its welfare system behaves in ways that political scientists usually associate with federations. A new study published in Discover Global Society argues that the country’s Social Safety Net Programs (SSNPs), while administered through a unitary governance structure, produce deliberately uneven patterns of resource distribution that the researchers call programmatic asymmetry. Rather than spreading funds equally across all territories, the state channels more support to regions facing deeper poverty, greater climate vulnerability, and sharper demographic disadvantage. The finding challenges a long-standing assumption that asymmetrical governance requires constitutional differentiation, and it reframes how scholars and policymakers might understand welfare allocation in centralized developing countries.

The scale of the system is enormous. For the 2024–2025 fiscal year, Bangladesh allocated more than 136,026 crore taka to a wide portfolio of safety net programs, ranging from old age allowances and school stipends to food assistance and employment schemes. This makes the country’s social protection architecture one of the largest in the region. Yet the way these resources move through the administrative hierarchy is anything but uniform. The Finance Division, in consultation with the Cabinet, sets the national budget, which is then divided among line ministries such as the Ministry of Social Welfare and the Ministry of Disaster Management and Relief. Each ministry sets beneficiary quotas for sub-districts using formulas built on poverty maps derived from the national Household Income and Expenditure Survey, census data on demographic composition, and, for some programs, indicators of climate risk and disability.

Once quotas reach the local level, the process becomes participatory. Upazila administrations, led by the Upazila Nirbahi Officer, receive the allocations and oversee implementation through committees that include public representatives. At the grassroots, Union Parishads, the lowest tier of local government, identify preliminary beneficiary lists based on centrally defined criteria such as age, income, and disability status. These lists are verified upward before payments are initiated, increasingly through direct digital transfers via mobile financial services designed to reduce leakage. The result is a hybrid structure: centralized, data-driven decisions about how much support each area receives, combined with decentralized, community-based identification of who actually receives it. The study’s authors describe this as a form of shared asymmetry, where central direction and local knowledge jointly shape welfare delivery.

To test whether this differentiated allocation actually tracks need, the researchers focused on the Employment Generation Programme for the Poorest (EGPP), the program for which divisional data were most readily available. Their descriptive analysis of beneficiary coverage, poverty rates, and unemployment figures across Bangladesh’s eight divisions reveals a relationship that is partially aligned but imperfect. The strongest evidence of pro-poor targeting comes from Rangpur and Barishal. Rangpur recorded the highest poverty rate in 2010 at 46.2 percent and the highest unemployment rate at 8.0 percent in 2015–16, and correspondingly received the highest concentration of EGPP beneficiaries, with 1.00 percent of its total population enrolled. Barishal, with the second-highest poverty rate of 39.4 percent, received the second-highest concentration at 0.95 percent. This alignment suggests that allocation decisions broadly follow the principle of vertical equity, whereby regions experiencing greater deprivation receive proportionately greater public support.

Dhaka presents a more nuanced case that illustrates why absolute numbers can mislead. The division around the capital has the highest total number of EGPP beneficiaries, 218,521 people, despite a mid-range poverty rate of 30.5 percent and low unemployment of 3.3 percent. At first glance this looks like misallocation. But when normalized for Dhaka’s immense population, the beneficiary concentration is a relatively low 0.42 percent. The authors interpret this as a deliberate policy choice: the large absolute number addresses the sheer volume of poor people in the country’s most populous division, while the lower concentration prioritizes regions where poverty is more intense and widespread. It is a sophisticated balancing act between the scale and the intensity of deprivation, and it demonstrates that relative coverage, not raw headcounts, is the more accurate measure of allocation equity.

Not every case fits so neatly. Chattogram had the second-highest unemployment rate in 2015–16 at 4.5 percent, yet received the lowest concentration of EGPP support, with only 0.23 percent of its population enrolled. The researchers attribute this discrepancy to a possible over-reliance on 2010 poverty data, which had ranked Chattogram as the least poor division. If allocation formulas weighted that older metric heavily, they may have failed to capture the division’s more recent employment crisis. This mismatch exposes a structural weakness in the data-to-policy pipeline: when welfare decisions rest on outdated or unsynchronized indicators, the resulting asymmetry may not reflect actual need. The study also notes temporal inconsistencies in the available datasets themselves, combining poverty estimates from 2010 with unemployment and program data from 2015–16, which means the findings should be read as descriptive approximations rather than causal measurements.

The political dimension of allocation cannot be ignored. Previous empirical work cited in the study, including research by Ali and colleagues and by Karim and Noy, shows that social assistance and disaster-related spending in Bangladesh are shaped not only by vulnerability but also by political considerations and administrative discretion. A report by the interim government’s task force on economic strategy restructuring found that nearly 54 percent of poor and vulnerable households remain outside the coverage of social protection, while 62 percent of households currently receiving benefits are neither poor nor at risk. These inclusion and exclusion errors, documented repeatedly across South Asian social protection systems, suggest that even well-designed need-based formulas can be distorted by local elite dynamics and patronage. The authors caution that asymmetry without transparency risks being perceived as political favoritism, undermining the public trust on which welfare legitimacy depends.

To frame these patterns conceptually, the study draws on India’s experience with constitutional asymmetry. India’s federal architecture grants differentiated autonomy and protections to particular states and regions through provisions such as Articles 370 and 371, including special safeguards for Nagaland under Article 371A and Mizoram under Article 371G, along with the Fifth and Sixth Schedules that create special administrative frameworks for tribal areas and Autonomous District Councils in the northeast. Scholars such as Alfred Stepan have argued that this asymmetry has been central to India’s democratic stability by allowing local institutions to adapt national policies to specific sociocultural contexts. Bangladesh grants no such constitutional differentiation, but the study contends that the underlying logic, accommodating territorial diversity through differentiated intervention, operates in its welfare system all the same. The Indian case serves as a conceptual reference rather than an institutional model, demonstrating that asymmetry is not an anomaly but a governance strategy for heterogeneous societies.

The policy implications are substantial. The authors recommend developing transparent, standardized allocation frameworks that incorporate regularly updated poverty, unemployment, demographic, and vulnerability indicators; exploring region-sensitive program designs that reflect the distinct shocks facing coastal districts, char regions subject to river erosion, and haor wetlands prone to seasonal flooding; and strengthening beneficiary identification by combining digital databases with community-based verification. They also suggest district-level wage rates for public works programs to match local economic conditions, interoperability among fragmented ministry databases to reduce duplication, and pilot projects in highly vulnerable districts to test decentralized approaches before national rollout. Each recommendation rests on the same premise: that fairness in practice requires differentiated, evidence-based distribution rather than uniform allocation that treats unequal regions as if they were identical.

The study’s contribution lies in extending the concept of asymmetrical governance beyond federalism. By demonstrating that differentiated welfare allocation can emerge within a unitary state through administrative and programmatic mechanisms, the research opens a new analytical window on social protection in centralized developing countries. The authors are careful to note the limits of their evidence: the analysis relies entirely on secondary data, focuses primarily on a single program, and cannot establish causal relationships. Future work using district-level longitudinal datasets and mixed methods could test whether programmatic asymmetry consistently improves outcomes. But the core insight stands. In a country where cyclones batter the coast, rivers devour char settlements, and poverty clusters unevenly across divisions, treating every territory the same would not be equality. It would be blindness. Bangladesh’s welfare system, with all its imperfections, is learning to see.

Subject of Research: Programmatic asymmetry in social safety net allocation within Bangladesh's unitary governance system

Article Title: Asymmetrical governance and territorial welfare allocation in Bangladesh

Article References: Asymmetrical governance and territorial welfare allocation in Bangladesh. (n.d.). https://doi.org/10.1007/s44282-026-00603-3

Image Credits: AI Generated

DOI: 10.1007/s44282-026-00603-3

Keywords: Bangladesh, social safety nets, welfare allocation, asymmetrical governance, programmatic asymmetry, unitary state, federalism, poverty targeting, EGPP, social protection, India comparison, governance