Executive Summary. Bangladesh cut its upper poverty line headcount from 48.9% in 2000 to 18.7% in 2022 (BBS HIES 2022), one of the fastest sustained reductions on record, then watched part of it unwind: an independent PPRC survey fielded in May 2025 put the upper-line rate at 27.93% and extreme poverty at 9.35%, up from 5.6% in 2022. The governing thought of this piece is falsifiable: the remaining poverty is not a smaller version of the old poverty but a spatially concentrated one, so a social-protection system that allocates on population rather than current divisional need will keep missing it. Barishal (26.9%), Rangpur (24.8%), and Mymensingh (24.2%) carry HIES 2022 headcounts above twice Khulna's 14.8% floor, and Sylhet's child stunting rate of 31.2% (BDHS 2022) predicts a generational productivity gap no aggregate growth figure erases. The BNP government (sworn in February 17, 2026, following the Yunus-led interim administration from August 2024) inherited both the two-decade gain and its recent erosion.

Bangladesh halved its poverty rate in two decades. In 2000, nearly half the country, 48.9%, lived below the upper poverty line. By 2022, that number had fallen to 18.7% (BBS HIES 2022). Achieved without oil wealth and without large-scale aid dependency, this ranks among the most sustained poverty reductions in the developing world, driven by the garment sector, remittances, and agricultural modernization.

But the national average conceals a fracture. Disaggregated by division, by district, and by the rural-urban divide, the story of convergence dissolves into one of divergence. The satellites that track nightlights, built-up area, and vegetation health register it: growth has been spatially concentrated, and the places it has not reached are falling further behind.

This is Part 5, the final installment of the Satellite Bangladesh series. It examines who has been left behind, where they live, and what the data says about the structural barriers that keep them there.

The national story

The headline numbers are genuinely impressive. Bangladesh's poverty headcount under the upper poverty line fell from 48.9% in 2000 to 40.0% in 2005, 31.5% in 2010, 24.3% in 2016, and 18.7% in 2022 (BBS HIES; World Bank Global Poverty database for the 2016 figure). That is a decline of more than 30 percentage points in 22 years, and the 2016 and 2022 readings match this series' companion brief, "Spatial Poverty in Bangladesh."

The extreme poverty rate, measured under the lower poverty line, fell faster: from 34.3% in 2000 to 5.6% in 2022 (BBS HIES 2022). Extreme deprivation, the kind that kills, was reduced to a fraction of its turn-of-the-century level.

The rural-urban decomposition reveals the first crack. In 2000, rural poverty stood at 52.3% and urban poverty at 35.2%, a gap of 17.1 points. By 2022, rural poverty had fallen to 20.5% and urban to 14.7%, narrowing the gap to 5.8 points (BBS HIES 2022). Progress on both fronts, but rural Bangladesh still holds one in five people below the line, and rural poverty reduction slowed between 2016 and 2022 relative to the prior decade. The easy gains are made. The remaining poverty is harder, deeper, and more spatially concentrated.

The divisions that diverge

The Multidimensional Poverty Index, which captures deprivation across health, education, and living standards simultaneously, exposes geographic fault lines the national headcount obscures.

The OPHI/GED Bangladesh National MPI 2025, built on MICS 2019 data, places the national multidimensional poverty headcount at 24.1%. Sylhet division leads at 37.7%, followed by Mymensingh at 35.0% and Barishal at 31.6%, while Khulna sits lowest at 15.2%. The Sylhet to Khulna gap is more than two and a half to one.

The HIES income poverty data tells a parallel story. In 2022, Barishal had the highest divisional headcount at 26.9% under the upper poverty line, followed by Rangpur at 24.8% and Mymensingh at 24.2%. Khulna sits lowest at 14.8% (BBS HIES 2022), a 12.1-point spread across the eight divisions. Rangpur, the poorest division in 2016 at 47.2%, nearly halved its rate and fell to second place, so the map redrew itself in six years while most allocation formulas did not.

The satellite-derived poverty proxy index, constructed from VIIRS nightlights, GHSL built-up area, and MODIS vegetation health, adds spatial resolution to these survey-based estimates.

At the district level, the proxy ranks Kurigram highest at 0.534, followed by Bhola at 0.525, Manikganj at 0.524, Satkhira at 0.522, and Sunamganj at 0.522. At the other end, Gazipur (0.494) scores lowest, followed by Feni (0.498), Chattogram (0.498), Khagrachhari (0.498), and Joypurhat (0.499).

Treat this proxy as a ranking, not a measurement of poverty levels. The full range is only 0.494 to 0.534, a 0.040 spread on a unitless composite, and the index has not been calibrated against survey headcounts to attach a confidence interval to any single district's score. The honest claim is narrower than "this is a map of structural inequality": the proxy's rank order aligns with the independent HIES and MPI evidence. The top-scoring districts span the northwest (Kurigram in Rangpur), the coast (Bhola in Barishal, Satkhira in Khulna), the northeast (Sunamganj in Sylhet), and the center (Manikganj in Dhaka); the lowest cluster around industrial and urban centers (Gazipur, Chattogram). Where three independent methods, a consumption survey, a multidimensional index, and a satellite composite, point at the same geographies, the targeting signal is real even if the proxy's absolute scale is not. The proxy's value is cadence between survey rounds, not precision on any one district.

The children who suffer most

Poverty is an abstraction until it is measured in the bodies of children. The 2022 Demographic and Health Survey provides that measurement.

Sylhet again stands apart. Per the BDHS 2022 Key Indicators Report, 31.2% of children under five in Sylhet are stunted, too short for their age, an irreversible marker of chronic malnutrition that constrains cognitive development and lifetime earnings. That is well above the national rate and roughly twelve points above Khulna's rate of about 19%. Sylhet has come down from 49.6% in 2014, a real reduction from a catastrophic baseline.

The 12-point stunting gap between Sylhet and Khulna is a prediction about the next generation's capacity to escape poverty. A child stunted in Sylhet today will, on average, earn less, learn less, and produce less over a lifetime than a child in Khulna or Dhaka.

What the satellites cannot see

Nightlight composites, built-up area estimates, and vegetation indices measure proxies, not poverty. They cannot detect whether a child was breastfed exclusively for six months, whether a school in Kurigram has a functioning toilet, or whether a household in Sunamganj has clean water and a staffed health facility within reach.

What satellite data can do is identify spatial patterns of deprivation at a resolution and frequency household surveys cannot match. DHS runs every five to seven years; HIES every four to six. Satellites observe daily. The proxy is imperfect, but its coverage is total and its cadence continuous.

The real gap is not in the data. It is in the institutional response. Bangladesh has some of the best poverty and health survey infrastructure in the developing world, better than many richer countries. The question is whether the spatial patterns it produces, the same patterns visible from orbit, are translated into spatially targeted policy.

What policy must do

The satellite and survey evidence converges on three priorities, each with an owner, a measurable target, and a falsification test.

First, re-base the social-protection allocation formula on the 2022 divisional map, not population shares. Bangladesh's largest in-kind safety-net schemes, Vulnerable Group Development and Vulnerable Group Feeding, distribute resources partly on population rather than current headcount, which dilutes spending where need is now highest. The Finance Division and the Department of Social Services should weight VGD and VGF card allocations by current divisional and district headcount ratios from HIES, refreshed each survey cycle. Owner: Finance Division. Target and success signal: by the FY2027 budget, per-capita VGD/VGF allocation in Barishal (26.9%), Rangpur (24.8%), and Mymensingh (24.2%) exceeds Khulna's (14.8%), reversing a formula that did not track the 2016-to-2022 reshuffle. Falsification: if a recosted formula leaves the four high-headcount divisions at or below the national per-capita average, the reform has not happened.

Second, launch a division-level nutrition response in Sylhet. A 31.2% stunting rate is a crisis, roughly ten points above the best-performing divisions, with large per-child losses in cognition and lifetime earnings. Model the response on community-based management of acute malnutrition: supplementary feeding, micronutrient supplementation, infant and young child feeding counseling, integrated with the existing community clinic network. Owner: Directorate General of Health Services, with the Institute of Public Health Nutrition. Success signal: a measurable fall in Sylhet's under-five stunting at the next BDHS round, narrowing the gap to Khulna.

Third, wire the satellite poverty proxy into targeting between survey rounds, validated before it allocates money. The multi-year gaps between DHS and HIES rounds leave policymakers blind between survey years. An annually updated proxy cannot replace survey measurement, but it can flag districts where conditions deteriorate before the next survey confirms it. Owner: BBS, with the social-protection delivery agencies. Success signal: an interim sub-district deterioration watchlist published between survey rounds, each flag scored against the following HIES before it changes any allocation. Falsification: if proxy-flagged districts show no above-average deterioration in the next survey, drop the proxy from targeting.

The strongest counterargument is leakage: the poor are not only in poor divisions, and a formula keyed to divisional headcounts can miss poor households in richer divisions while subsidizing non-poor households in poorer ones. The answer is to keep the proxy-means test as the household-level eligibility instrument and use the divisional map only to set budget envelopes. What would change the conclusion: if a fresh census or HIES round showed districts converging, the case for geographically differentiated allocation would weaken. The PPRC 2025 reading also carries a caution. It is a single non-BBS survey of 8,067 households, so its 27.93% headcount should anchor urgency, not replace the official HIES series, until the next BBS round confirms the direction.

Five chapters, forty years of satellite data. The evidence is consistent: Bangladesh made extraordinary progress, visible in retreating flood footprints turned to farmland, in nightlights spreading across a once-dark landscape, and that progress was unevenly distributed, with millions, disproportionately in the north and northeast, living in conditions the national averages hide. The next budget cycle will show whether the allocation formula moves with the map or stays frozen on it.

Note on 2025 poverty data. After this narrative was written, the Power and Participation Research Centre (PPRC) released "Economic Dynamics and Mood at Household Level in Mid-2025," based on a May 2025 survey of 8,067 households across all 64 districts. It found the upper poverty line headcount at 27.93% (from 18.7% in BBS HIES 2022) and extreme poverty at 9.35% (from 5.6% in 2022), with the national Gini rising to 0.436 against the HIES 2022 consumption Gini of 0.334. Separately, the World Bank's November 2025 assessment estimated that nearly 2 million more people fell into poverty in 2025. These are not BBS HIES figures and are not incorporated into the charts above, which use HIES 2022 as the most recent official BBS round. They reinforce rather than contradict the piece's central argument: Bangladesh's inequality is widening, not narrowing.

Sources

• HIES poverty headcount: Bangladesh Bureau of Statistics, Household Income and Expenditure Survey 2022. https://bbs.portal.gov.bd/. 2016 upper-line headcount (24.3%) cross-checked against the World Bank Global Poverty database for Bangladesh. Internal output: bd_gis/outputs/poverty/hies_national_hcr_timeseries.csv.
• HIES 2022 divisional headcounts: BBS HIES 2022 (Barishal 26.9% highest, Rangpur 24.8%, Mymensingh 24.2%, Khulna 14.8% lowest, upper poverty line). https://bbs.gov.bd/site/page/648dd9f5-067b-4bcc-ba38-45bfb9b12394/Income,-Expenditure-&-Poverty.
• Bangladesh National MPI 2025: Oxford Poverty and Human Development Initiative (OPHI) / Government of Bangladesh General Economics Division, built on MICS 2019. https://ophi.org.uk/bangladesh.
• BDHS 2022: Bangladesh Demographic and Health Survey 2022, Key Indicators Report. NIPORT and ICF. https://dhsprogram.com/publications/publication-FR380-DHS-Final-Reports.cfm
• Satellite poverty proxy: Composite of VIIRS nightlights (2020), GHSL built-up (2020), MODIS NDVI (2020). Uncalibrated ranking, not a level estimate. Internal output: bd_gis/outputs/poverty/poverty_district_ranking.csv.
• Child nutrition by division: Internal output: bd_gis/outputs/health/dhs_child_nutrition.csv, derived from BDHS 2022.
• 2025 poverty backsliding: Power and Participation Research Centre (PPRC), "Economic Dynamics and Mood at Household Level in Mid-2025," May 2025 survey, 8,067 households (UPL 27.93%, LPL 9.35%, Gini 0.436). Reported in The Daily Star, "Poverty rate jumps to 27.9%; extreme poverty nearly doubles to 9.3%," 2025. https://www.thedailystar.net/business/news/bangladeshs-poverty-jumps-3970246
• World Bank poverty estimate (2025): World Bank, "Job Creation Key to Equality and Faster Poverty Reduction in Bangladesh," November 2025 (nearly 2 million additional people in poverty in 2025). https://www.worldbank.org/en/news/press-release/2025/11/25/job-creation-key-to-equality-and-faster-poverty-reduction-in-bangladesh-says-world-bank
• HIES 2022 consumption Gini (0.334): BBS HIES 2022, consumption-based Gini; the PPRC 2025 Gini of 0.436 is benchmarked against this figure.

Data Sources

• Poverty headcount (HIES): Bangladesh Household Income and Expenditure Survey, BBS. Years: 2000, 2005, 2010, 2016, 2022. 2016 figure cross-checked against the World Bank Global Poverty database. Output: bd_gis/outputs/poverty/hies_national_hcr_timeseries.csv.
• Multidimensional Poverty Index (MPI): OPHI/GED Bangladesh National MPI 2025 report, built on MICS 2019 data. Division-level headcount ratios.
• Satellite poverty proxy: Composite of VIIRS nightlights (2020), GHSL built-up area (2020), MODIS NDVI (2020). 64 districts. Uncalibrated ranking, not a level estimate. Output: bd_gis/outputs/poverty/poverty_district_ranking.csv.
• Child nutrition (BDHS 2022): Bangladesh Demographic and Health Survey 2022 (ninth round), Key Indicators Report. Division-level stunting, wasting, underweight.
• HIES division-level poverty: BBS HIES 2022, division-level headcount ratios (upper poverty line); Barishal 26.9% highest, Khulna 14.8% lowest, Rangpur fell from 47.2% in 2016 to 24.8%.
• 2025 poverty (PPRC): Power and Participation Research Centre, "Economic Dynamics and Mood at Household Level in Mid-2025," May 2025 survey of 8,067 households (UPL 27.93%, LPL 9.35%, Gini 0.436 vs HIES 2022 consumption Gini 0.334). Non-BBS survey; not incorporated into HIES-based charts.