Layer 2 · In depth

Culture and religion

A careful, evidence-only look at whether culture or religious conservatism explains Bangladesh's economic outcomes - eleven specific propositions tested one by one, what is causal, what is correlation and what is opinion.

Updated 21 Nov 2026 · 14 min read

On this page
  1. Levels that should not be run together
  2. Eleven propositions, tested one by one
  3. Evidence on each, in a little more detail
  4. What is causal, what is correlational, and what is opinion
  5. The channels that matter economically
  6. Personal freedom, social life and investment
  7. Calibrated judgement
  8. Overlapping consensus and the individual inside the community
  9. Voices from within the tradition
  10. Gaps

Faith is central to the lives of most Bangladeshis, and it is also a topic where careless claims do harm. This page treats it with the same rule as every other topic: state what the evidence shows, say where it is thin, and do not blame any people or religion. Nothing here is a claim that Muslims, or any group, are less capable, less industrious or less open-minded. Every proposition below is a hypothesis about institutions and enforced arrangements, not about the character of individuals.

In one paragraph

Religious identity does not by itself explain Bangladesh’s development outcomes. Muslim-majority countries differ enormously. Bangladesh passed Pakistan in income per head, and its own record on health, schooling and garment jobs for women shows that norms respond to opportunity. What does matter, on the evidence, are specific arrangements that exclude women from paid work, limit peaceful dissent, or enforce one orthodoxy by threat. These can be real obstacles whatever the religion that is invoked. The popular claims about lost working hours from prayer or the afterlife do not meet the same standard of evidence.

Levels that should not be run together

A rigorous account separates:

  1. Religious doctrine: texts and theological arguments.
  2. Interpretation and practice: legal schools, local traditions, clerical movements and believers’ own choices.
  3. Political Islam or Islamism: movements that seek to organise political institutions around a conception of Islam.
  4. Social organisation: family authority, patriarchy, kinship, class and local custom, some of which predate or exist independently of any religion.
  5. Enforcement: the difference between private observance, community pressure and criminal or administrative coercion by the state.
  6. Differences across countries: Bangladesh, Pakistan, Afghanistan, Indonesia, Malaysia and the Gulf states differ markedly in institutions, conflict exposure and gender outcomes.
  7. Similar patterns outside Islam: religiously conservative and authoritarian movements exist in other faiths and in secular settings.

These cannot be collapsed into “religion causes poverty”. The narrowest defensible proposition is that particular interpretations, when combined with coercive laws, patriarchal arrangements and weak accountability, can reduce participation, experimentation, criticism and institutional adaptation.

How belief becomes economically consequential

Inputs

  • Religious teachings and traditions
  • Local patriarchy, hierarchy and kinship
  • Party politics, state incentives, coercive power

Interpretation

  • Interpretation by people and institutionsthis step changes over time

Enforcement

  • Laws, social expectations and who enforces them

Effects

  • Participation, freedom of inquiry, incentives
  • Human capabilities and development outcomesfeed back into politics and into interpretation
Beliefs act through interpretation and institutional enforcement, and those can themselves change.

Eleven propositions, tested one by one

For each, the assessment says how far the evidence supports it. The question at the end of each is the most useful test.

# Proposition Assessment
1 Islam is the least reformed major religion; radicalism is uniquely widespread Not testable as stated: “reformed” is not a measurable variable, and Islamic history includes jurisprudential diversity, Sufi and modernist reform and secular governance as well as coercive orthodoxy. Partly supported in specific legal-political settings (such as Pakistan and Afghanistan on some rights), too sweeping as a universal comparison
2 Belief in an afterlife reduces effort to improve worldly conditions Plausible as an individual mechanism, weak as a macroeconomic explanation. It can equally encourage discipline, saving, charity and service. Productivity gaps are better explained by capital, technology, management, institutions and human capital
3 Five daily prayers substantially reduce productivity Weak as a national explanation. Time displaced varies by shift, occupation and scheduling; no credible estimate of lost output exists
4 Friday prayers remove an extra workday Generally overstated. The Friday-Saturday weekend is a scheduling choice, and rest has value
5 The Friday-Saturday weekend creates a three-day trade shutdown with Western countries A real coordination issue, not a literal shutdown. With a Friday-Saturday weekend, Bangladeshi working days are Sunday to Thursday against Monday to Friday in much of the West, so regular overlap is Monday to Thursday. This can add hand-off and settlement delays in specific processes; factories, ports and digital work do not all stop. The cost has not been measured
6 Conservative interpretations suppress women’s work and autonomy Among the strongest mechanisms when expressed as enforceable restrictions and social constraints. The key distinction is individual choice versus enforced exclusion
7 Gender segregation and restrictions on intimacy cause widespread psychological harm and economic under-performance Credible concern for autonomy and wellbeing for some people; sweeping causal claims about national productivity are unproven. The firmer pathway is social exclusion: less networking, mobility and women’s participation
8 Restrictions on gender interaction stunt entertainment and so development Potentially relevant to expression and creative industries; likely secondary as a macroeconomic cause
9 Respect for seniority discourages criticism and improvement Plausible and important, but not specifically Islamic. Variables that matter: psychological safety, professional autonomy, accountability and whether evidence can override rank
10 Religion filters all public questions and intensifies conflict Plausible under identity-based political mobilisation, not universal. The issue is whether religious identity is one of many voluntary affiliations or a compulsory test of civic legitimacy
11 Blasphemy taboos and suppression of opposing views block intellectual progress A strong institutional concern where threats, violence or laws enforce orthodoxy. This is an argument for protecting peaceful criticism and conscience under equal laws, not for believers to abandon belief

Evidence on each, in a little more detail

Propositions 2, 3 and 4 (belief and time). The strongest causal study is on fasting. Using the fact that the lunar calendar moves Ramadan through the seasons, it finds that longer fasting hours lower output growth in Muslim-majority countries and raise Muslims’ reported well-being. A one-standard-deviation increase in fasting hours (about 10%) lowers growth by about a sixth of a standard deviation. The authors say the effect “cannot be primarily explained by a direct reduction in labor productivity” and attribute it to shifts in labour supply and occupational choice.12 Fasting by pregnant women also harms prenatal development.3 Barro and McCleary find that growth rises with religious belief but falls with attendance, which suggests that the costs come from the time and institutions of practice, not from faith.4

Figures on Ramadan and prayer that are not supported

An earlier analysis quoted a “0.7% GDP drop per additional fasting hour”. That number is back-of-envelope arithmetic from the paper’s illustration (a 1.4-hour fasting gap and a roughly one-point growth gap between Bangladesh and Turkey); it is not a reported coefficient and it describes a growth effect, not a fall in GDP.2 Figures of “42 hours lost per employee” and a “35-50% productivity decrease during Ramadan” have no source anywhere in this research and cut against the paper’s own conclusion. We do not use them. Workers in Bangladesh average about 2,383 hours a year per person engaged (2023), so the evidence does not support the idea that poverty mainly reflects unwillingness to work.5

Propositions 1, 10 and 11 (religion and the state). Across countries the literature finds no robust “Islam penalty”: one review reports that “virtually every statistically significant coefficient” on Muslim population share is positive.6 Kuran locates the historical drag in legal institutions (inheritance rules, rigid endowments, delayed impersonal exchange, states using Islam as a governing tool).3 Rubin argues that the culprit was religious legitimacy giving clerics a veto over innovations such as printing and interest-bearing lending, not Islam as such.7 Even the Protestant advantage worked mostly through literacy.8 None of these studies Bangladesh directly. In practice, the cases of Pakistan and Afghanistan show severe restrictions in some domains, but war, aid dependence and weak institutions are indispensable to explaining Afghanistan, and Europe’s own history includes centuries of religious establishment and censorship before pluralism.910 Religious traditions are neither permanently liberal nor permanently illiberal.

Propositions 6, 7 and 8 (women and social life). On women’s work and autonomy the evidence is strongest. Restricting women’s mobility, employment, occupation, property or association reduces household income, narrows the pool of entrepreneurs and professionals and keeps women dependent. The estimated economic cost is on Root causes. The economic payoff of nightlife, alcohol rules or a different weekend is not measured in the inputs. See “Personal freedom” below.

Proposition 9 (hierarchy). Hierarchy is not unique to any faith; Japan and Korea combine high power distance with wealth. On Hofstede’s scores Bangladesh is cited as high on power distance (80) and low on individualism (20) and indulgence (20), but these are old national averages, and the research could not retrieve World Values Survey trust data (Root causes).

What is causal, what is correlational, and what is opinion

Level Findings
Causal The Ramadan study (above); Heath and Mobarak’s study of garment jobs; Kuran’s prenatal-fasting result
Correlational No robust Islam penalty across countries; Bangladesh versus West Bengal and Pakistan; pupils of Bangladeshi heritage in England (see History and comparisons)
Opinion, with no Bangladesh-specific support in the inputs That deference suppresses innovation in Bangladesh; that a lack of nightlife, alcohol or open romance depresses growth; the size of tourism and creative sectors; the cost of moral policing

Norms follow opportunity. When garment factories opened within reach of a village, girls stayed in school longer and delayed marriage and childbirth. The effect on schooling was larger than a big conditional cash-transfer programme.11 So at least part of the conservatism observed is a consequence of poverty and scarce jobs, not only a cause. When the state offered teacher salaries and stipends for girls, many traditional madrasas registered, added secular subjects and admitted girls.12 Gender norms that look religious can respond to economic opportunity within a few years.

The channels that matter economically

Channel Evidence Materiality Confidence
Women’s economic participation Female participation 38.4% (2024), falling; Global Gender Gap economic-participation rank 146; 64.5% of women not seeking work cite housework or family duties; about 85% of jobs lost in the first half of FY25 were women’s; 49.3% of young women NEET High (estimate of the gain: see Root causes); the split between norms and lack of demand is unknown moderate
Child marriage 47.2% married before 18 (MICS 2025); 51% of women aged 20-24 (BDHS 2022); a conflicting 63% figure exists; a 2023 survey found 88-89% of caregivers believed it carried benefits High, via girls’ schooling and fertility; it responds to jobs moderate
Madrasa-stream skills Alia enrolment up to 2.84m; Qawmi estimates 1.4m to 7m (see Education and skills) Plausibly material, unmeasured; graduates’ labour-market outcomes are a gap low
Speech restrictions At least four 2026 arrests for insulting the government (Human Rights Watch); 97 shrine attacks with 11 cases filed Not established: no systematic arrest data to compare political and religious restrictions; mob enforcement can erode investor confidence low
An organised veto over reform Hefazat-e-Islam rallied about 20,000 people against the Women’s Reform Commission; Jamaat won 31.77% of the vote; Freedom House reports intimidation curbing women’s public life; mobs disrupted women’s football and cultural events in 2024-26 Moderate and rising; it can block gender-law and curriculum reform moderate

Sources: 1314151617181920

Child marriage and female labour participation

Child marriage before 18: 47.2% (MICS 2025), 51% (BDHS 2022, women aged 20-24) and 63% (a UN Women citation). Female labour participation: 38.4% (national definition, 2024), 42.8% (2022), 21.3% (ILO definition, 2022) and 39% (modelled ILO 2025). Different surveys, definitions and ages. See Disputed figures.

An unverified statement about women’s working hours

Jamaat’s ameer has spoken about regulating women’s working hours, but its manifesto does not say so.21 We note it as a fact to monitor, not as a policy position.

Personal freedom, social life and investment

Analysts treat this differently, and both views are included.

  • One view. Skilled managers, engineers and investors will not relocate to a society that is hyper-policed in private life. Malaysia, the UAE, Singapore and India attract capital while guaranteeing personal privacy and predictable commercial law, even in Muslim-majority countries. Zia Hassan’s analysis of the raid in which 85 consenting adults were arrested (see Voices) is the example cited.
  • The other view. The research found no measured growth payoff for nightlife, alcohol liberalisation or a Saturday-Sunday weekend. The economically material freedoms are women’s mobility, safety and access to work, and freedom from mob and state coercion. Cultural freedoms may matter for quality of life, talent retention and the diaspora, but the evidence for a growth effect is missing.

What can be said on facts:

  • Alcohol is tightly restricted rather than banned (Narcotics Control Act 2018; Alcohol Control Rules 2022: minimum age 21, permits largely for foreigners, non-Muslims or on medical prescription, 10:30 pm bar closings).22 The WHO estimates over 90% of Bangladeshis are alcohol-free (5.4% adult consumption in 2020), which keeps alcohol-related mortality among the lowest in the world, a genuine public-health benefit.23
  • Keeping consumable alcohol under narcotics law, despite a High Court ruling that it should be regulated separately, is linked in one input to illicit home-brewed spirits and methanol poisonings and to a weak hospitality sector.24 Bangladesh drew about 323,000 international tourist arrivals in 2019 and $217.9m in tourism receipts in 2020 (0.57% of exports).25 Whether that is mainly due to alcohol rules is not established.
  • The effect of a UAE-style weekend change is predicted but not measured.26
  • Minority violence counts are contested (522 rights-group counts against 71 classified as communal by the government in 2025).27
  • Gaps: Section 377, Section 295A and moral-policing data, the size of tourism and creative sectors, and the cost of a Friday-Saturday weekend.

Calibrated judgement

How much weight should religion carry?

Religious conservatism is a real and economically material factor, but on the evidence it is a second-tier cause. It ranks below the political settlement and its institutions (captured banks, courts and police), trade protection, the weak fiscal state and the historical starting point. Its largest effects run through women’s work and marriage, where it interacts with, and is partly produced by, a narrow labour market. It also runs through an organised veto over gender and curriculum reform, which may become the most important channel if its political support grows. The evidence does not support saying that Islam is the reason Bangladesh is poor, or that secularisation is a precondition for growth. It does support saying that a development strategy that ignores gender norms and organised vetoes will under-perform. Confidence: moderate. It would fall if diaspora employment data or madrasa outcome data showed large penalties, and it is a judgement about the past and present, not a forecast. Pew’s 2013 survey found 82% of Bangladeshi Muslims favoured sharia as official law, and there is no trend data on religiosity since.28

A conditional reading

If growth stalls and women’s jobs keep disappearing, organised vetoes could widen. If the economy reopens and women’s employment recovers, the garment evidence suggests norms could move with it, as they did before. This is a conditional judgement, not a forecast. What follows for the strategy is on Strategy and deals and the political risks are on Risks and open questions.

Overlapping consensus and the individual inside the community

A practical way to hold these findings together is to ask each tradition to endorse shared public rules from inside its own values, rather than to place faith below the economy. Political philosophers call this an overlapping consensus. It is proposed, not established; the reasoning and the comparisons are on the compact across differences. It rests on Rule 3 of the proposed ground rules: people may persuade and practise freely, but never coerce, and individuals inside any community keep the right to decline a programme, to dissent and to act independently of religious leaders. Inclusion of religious institutions in public programmes should therefore use neutral eligibility, and basic protection should not depend on a community’s endorsement.

Voices from within the tradition

One argument (unverified; no primary source located) is that the most persuasive case for freedom of conscience in a Muslim-majority society comes from within the tradition, and names Islamic modernist and reformist thinkers (Fazlur Rahman, Nasr Abu Zayd, Mohammed Arkoun, Khaled Abou El Fadl, Amina Wadud, Ziba Mir-Hosseini and, from Bangladesh, Akbar Ali Khan). It also suggests leading with economic points (inquiry, women’s work, an end to mob violence) before the most divisive questions. Those are recommendations for strategy and tone; they are discussed on Objections and critique.

Gaps

  • Trend data on religiosity after 2013.
  • Graduates’ earnings from madrasas.
  • Diaspora employment and female economic inactivity.
  • Comparative income paths of Malaysia and Indonesia.
  • Hofstede-type and World Values Survey measures with sources for Bangladesh.

  1. Campante and Yanagizawa-Drott, Does religion affect economic growth and happiness? Evidence from Ramadan (NBER w19768). ↩

  2. Same paper, QJE version. ↩↩

  3. Kuran, Islam and economic performance, Journal of Economic Literature; related synthesis. ↩↩

  4. Barro and McCleary, NBER w9682. ↩

  5. FRED (Conference Board), Annual hours worked per person engaged, Bangladesh. ↩

  6. Noland, Religion, culture and economic performance (PIIE). ↩

  7. Rubin, Why the Middle East fell behind. ↩

  8. Becker and Woessmann, DIW. ↩

  9. Pew Research Center, Restrictions in the 25 most populous countries in 2022. ↩

  10. Freedom House, Pakistan: Freedom in the World 2025. ↩

  11. Heath and Mobarak, Manufacturing growth and the lives of Bangladeshi women. ↩

  12. Asadullah and Chaudhury, Holy alliances; Bano, summary paper. ↩

  13. Prothom Alo on the World Bank Bangladesh Development Update, https://en.prothomalo.com/business/local/qgylemzmn1. ↩

  14. UN Women, State of the Bangladesh economy from a gender lens. ↩

  15. Daily Sun, child marriage, MICS 2025. ↩

  16. Human Rights Watch, Bangladeshi women’s rights opposed by hardline religious groups. ↩

  17. Freedom House, Bangladesh 2026. ↩

  18. Daily Star, 97 attacks on shrines. ↩

  19. Human Rights Watch, Bangladesh: 4 arrested for insulting government. ↩

  20. For the 2022 BDHS, 2023 caregiver survey and mob disruptions, citing UNICEF, Licas News and the BTI 2026 country report. ↩

  21. TBS, Jamaat election manifesto. ↩

  22. TRW, Movendi); commercial and advocacy sites, treat as leads. ↩

  23. WHO, alcohol and health: Bangladesh country snapshot. ↩

  24. TBS, Hotel and bar owners urge separate laws for alcohol and narcotics. ↩

  25. CEIC, Bangladesh: international tourism arrivals. ↩

  26. Zawya, Weekend shift in UAE. ↩

  27. IANS, 522 communal violence incidents in 2025; ThePrint, government says most incidents were non-communal. ↩

  28. Pew Research Center, The world’s Muslims: religion, politics and society (2013). ↩

Go deeper

Where this comes from

Synthesised from 11 research reports, checked against the sources cited above and reviewed before publication. How we work · All sources.