Layer 2 · In depth
History and comparisons
Where Bangladesh started in 1972, how it caught up on a narrow model, and how its income, health and education compare with India, Pakistan, Vietnam, West Bengal and others.
On this page
The first layer of the explanation of poverty is history: a very low starting point. The second is the way the country grew from that point. This page records both, then the comparisons that help test explanations. It supports Root causes and Culture and religion.
In one paragraph
Bangladesh began independent life in 1972 with about 62% of Pakistan’s income per head, after colonial land systems, the split of the jute economy and a war, with a famine to follow in 1974. It then grew faster than most comparators but from the bottom and along a narrow track of garments and remittances. Its health and schooling results beat what its income and governance would predict. Today its income per head is about 119% of Pakistan’s and, in nominal terms, forecast to pass India’s in 2026. The comparisons argue against any single-cause explanation.
A low starting point
- to 1947Colonial period
- Landlord-tenure systems in parts of Bengal
- Export-industry output fell roughly ninefold between 1801 and 1851
- 1947Partition
- East Pakistan had over 75% of the jute-growing land, but the mills and the export port were in India
- 1947-71Within Pakistan
- A military-bureaucratic order distributed import-control rents (1958-71)
- East Pakistan reportedly received about 40% of development spending despite a majority of the population (tertiary source)
- 1971-75War and famine
- The 1974 famine: government figure 26,000 deaths; one scholar's estimate about 1.5 million
- 1975-2014Ordered deals
- Weak rules, but deals between politicians and businesses were honoured
- 2014-2024One dominant party
- Deals stopped being disciplined by competition
- 2024-26Reset
- Uprising, interim government, election in February 2026
- Was Bengal rich? A popular story says Bengal was once richer than Britain and was then looted into poverty. That is overstated. All-India income per head was about 71% of Britain’s in 1600, falling to 30% by 1801, and Bengal was “substantially lower” than England by the late 18th century. Nineteenth-century deindustrialisation was real: export-industry output fell roughly ninefold between 1801 and 1851.1
- Colonial land institutions. In India, after independence, districts that had not been under the landlord-tenure model had about 16% higher yields, 40% lower infant mortality and higher public investment than landlord-tenure districts. The authors’ preferred mechanism is weak collective action in landlord areas. This is suggestive for East Bengal but was not estimated for it (confidence: low to moderate).2
- Partition cut the jute economy in two: East Pakistan had more than 75% of the jute-growing land, and every mill and the export port of Calcutta went to India.3
- Pakistan era. East Pakistan reportedly received about 40% of development spending despite a majority of the population. The figure comes from a tertiary source whose original is unverified.4 One scholar describes 1958-71 as a “praetorian” order distributing import-control rents to military-bureaucratic elites, and 1971-75 as zero-sum asset-stripping.5
- The 1974 famine killed between 26,000 (government figure) and about 1.5 million people (Alamgir’s estimate). Sen judged it not to have been a collapse of food availability.6
- The result in 1972: $373 per head (constant 2010 US dollars), 62% of Pakistan’s $605.78
- Gaps: Hindu professional emigration, war damage in dollars, and the share of industry nationalised in 1972.
Famine deaths, 1974
26,000 (government) to about 1.5 million (Alamgir estimate).6
Fast catch-up on a narrow model
GDP per head, constant 2010 US dollars (World Bank WDI via FRED); multiples are our arithmetic.7891011
| Country | 1972 | 1990 | 2010 | Latest | Multiple 1990 to latest |
|---|---|---|---|---|---|
| Bangladesh | 373 | 473 | 944 | 1,941 (2024) | 4.1x |
| India | 357 | 538 | 1,235 | 2,523 (2025) | 4.7x |
| Pakistan | 605 | 951 | 1,257 | 1,669 (2025) | 1.8x |
| Vietnam | n/a | 688 | 2,028 | 4,308 (2025) | 6.3x |
| China | 303 | 917 | 5,765 | 13,793 (2025) | 15x |
Current-price comparisons: nominal GDP per head is about $2,886 in 2026 (PPP $10,884) on IMF figures, and the IMF projects Bangladesh’s nominal per-head GDP above India’s in 2026, though not in PPP terms.12 Vietnam’s is about $4,700.
How the 1990-to-latest multiples compare
On these data Bangladesh has multiplied its income per head 4.1 times since 1990, close to India and far ahead of Pakistan, but behind Vietnam and China. Level comparisons carry real measurement error, because the IMF rated Bangladesh’s national accounts “C” and critics allege that 2009-2019 growth was overstated (What the numbers miss).
Social outcomes beat income. In 2006-10 Bangladesh had 19 fewer infant deaths per 1,000 and about two fewer births per woman than its income predicted. The authors attribute this to BRAC’s low-cost community programmes, family planning, female health workers and the female secondary stipend. Governance was significantly worse than in comparator countries, and aid per head was lower.13 The development economist Stefan Dercon calls this a pragmatic “developmental bargain”.14
How the growth model worked. Several scholars describe a pattern in which weak formal rules were offset by reliable informal deals:
- Garments grew on “accidental learning rents” from Multi-Fibre Arrangement quotas and on high-level political support for back-to-back letters of credit and bonded warehouses. Under “competitive clientelism”, conventional good-governance reform “does not give elites a workable reform agenda”.5
- Garments thrived because misgovernance was predictable and the state did not intervene.15
- Since 1975 the economy ran on “ordered deals”: when a politician made a deal with a business, it was honoured. After 2014 the system shifted to one dominant party.1617 One reading (our judgement) is that under one-party dominance the deals stopped being disciplined by competition, which is how bank lending turned into bank losses. Selim Raihan now describes the constraint as concentrated oligarchic power and calls for a new political settlement.18
- Mirza Hassan and Selim Raihan also describe “partyarchy”: parties capture formal institutions, and the elites cannot agree on how power changes hands.19
The constraints that bind now
The formal diagnoses agree on a short list. The World Bank’s 2022 Country Economic Memorandum names trade protection, the financial sector and urbanisation (figures on The economy in late 2026). The interim government’s Task Force named an anti-export tariff bias, banking governance and energy pricing.20 The IFC lists unreliable power, limited finance, corruption, informal competition and slow customs.21 Learning is the slow-acting constraint (Education and skills).
Comparisons used to test explanations
| Comparison | Result | Source |
|---|---|---|
| Bangladesh against Pakistan | 62% of Pakistan’s income in 1972; about 119% in 2025 (both 2025 values) | FRED78 |
| Bangladesh against Hindu-majority West Bengal | under-five mortality cut 51% (early 1990s-2007) against 39%, at roughly half West Bengal’s income; West Bengal still led on female literacy | CGD22 |
| Bangladesh against India | 77% of India’s income (constant dollars); nominal per head forecast above India’s in 2026 | FRED; Al Jazeera |
| British pupils of Bangladeshi heritage | Attainment 8 score 51.9 against 44.9 for White British pupils in state-funded mainstream schools in England, 2022-23 (migrant selection is an unmeasured factor) | UK government23 |
| Female labour participation | Indonesia 53.7%; Bangladesh 39%; India 32%; Sri Lanka 31% (modelled ILO 2025) | reported second-hand from World Bank gender data; series, age range and year not confirmed24 |
| Personal freedom (Cato 2025) | Bangladesh 5.16; Pakistan 5.00; Indonesia 6.47; Malaysia 6.79; Nepal 7.27 | Cato25 |
| Saudi Arabia | female labour participation about 36% by 2024-25 after state-led reform (no pre-reform baseline given) | GASTAT26 |
| Hours worked | see Culture and religion | |
| Malaysia | $470bn economy (unverified: no primary source located) |
Two comparisons are listed as gaps in the inputs: diaspora employment and female economic inactivity in the UK, and a comparison of Malaysia’s and Indonesia’s income paths with Bangladesh’s.
Does Bangladesh lead India and Sri Lanka on women’s participation?
A second-hand report of the World Bank gender data says yes on a modelled ILO 2025 series (39% against 32% and 31%); we have not confirmed the exact series, age range and countries on the linked indicator, which defaults to total participation ages 15-64.24 On the national definition, Bangladesh’s rate has fallen to 38.4% (2024), and on the stricter ILO definition it was 21.3% in 2022 (Culture and religion). The series are not comparable with each other.
The lesson drawn from these comparisons is on Culture and religion: Muslim-majority countries differ widely, Bangladesh outperformed Pakistan, and no one-variable explanation fits.
What other countries did about corruption (Singapore, Hong Kong, Georgia, Indonesia) is used as evidence in Strategy and deals.
Gaps
- Pakistan and West Bengal time series, and Malaysia and Indonesia development comparisons, are not in the inputs.
- The origin of the “40% of development spending” figure.
-
Broadberry and Gupta, working paper (CAGE, Warwick). ↩
-
Banerjee and Iyer, History, institutions and economic performance. ↩
-
Daily Star, Unthreading partition: the politics of jute sharing between two Bengals. ↩
-
Wikipedia, Bangladesh Liberation War. ↩
-
Mushtaq Khan, 2010 paper. ↩↩
-
bdnews24, 1974 famine in Bangladesh: economic context and myths. ↩↩
-
Al Jazeera (IMF figures), Bangladesh per capita GDP forecast to outstrip India’s. ↩
-
Asadullah, Savoia and Mahmud, IGC working paper. ↩
-
Ahmed and co-authors, The paradox of export growth in areas of weak governance: the case of the ready-made garment sector. ↩
-
ESID, Navigating the labyrinths of the deals world; Daily Star, analysis. ↩
-
ESID briefing paper 27, Bangladesh paradox. ↩
-
TBS, Economists call for urgent reforms and a new political settlement. ↩
-
BIGD, Violence and the breakdown of the political settlement. ↩
-
Center for Global Development, West Bengal and Bangladesh child health. ↩
-
UK Ethnicity Facts and Figures, GCSE Attainment 8. ↩
-
World Bank Gender Data Portal, labour force participation. ↩↩
-
Cato Institute, Human Freedom Index 2025. ↩
-
GASTAT, Saudi labour force statistics. ↩
Go deeper
- Key statisticsData & sourcesThe important figures from the research, grouped by theme in tables, each with its value, date, source and link. Where sources disagree, see the disputed-figures ledger.
- Disputed figuresData & sourcesThe ledger of figures where our sources disagree, with both values, the likely reason and how this site handles each, plus a list of known gaps.
- SourcesData & sourcesThe consolidated bibliography of every external source cited in our research, grouped by type and deduplicated, with the links as given.
Summarised in
- Why progress has stalled Overview
Where this comes from
Synthesised from 5 research reports, checked against the sources cited above and reviewed before publication. How we work · All sources.