Layer 2 · In depth

Objections, critique and what remains unknown

The strongest arguments against the plan and what survives each, the political-feasibility weakness, proposals examined from earlier work, the tests any one-cause explanation must pass, and the gaps in the evidence.

Updated 21 Nov 2026 · 22 min read

On this page
  1. The short version
  2. Thirteen objections, and what survives each
  3. 1. There is no designer
  4. 2. Buying groups off entrenches them
  5. 3. “Deals” keep Bangladesh in a deals economy
  6. 4. Reform comes from dominant executives, not consensus
  7. 5. The scores are invented
  8. 6. Evidence from elsewhere may not travel
  9. 7. The pain comes first and the government may not survive it
  10. 8. The world may not cooperate
  11. 9. It is a cyclical shock, not a structural crisis
  12. 10. The deals can be captured
  13. 11. Pilots stay pilots
  14. 12. It underrates religion
  15. 13. A version of the first deal has just been tried and thrown out
  16. The political-feasibility weakness
  17. Proposals examined from earlier work
  18. Four structural problems, restated as one system
  19. Four cases revisited
  20. Proposal: mindset change and individual freedom
  21. Proposal: staged anti-corruption enforcement
  22. Proposal: electoral design
  23. What the earlier plan was missing
  24. Objections to the compact itself
  25. “Rules on paper do not bind anyone”
  26. “The law can be weaponised”
  27. “Secularism by stealth”
  28. “It forgets the victims”
  29. Tests that any one-cause explanation must pass
  30. What remains unknown
  31. What survives overall

This page is the detailed reasoning behind risks and open questions. A plan that has not been attacked has not been tested. Here the strongest objections are stated at full strength, each followed by what survives. It then looks at the political-feasibility problem on its own, examines proposals made in earlier work, sets out the tests that any one-cause explanation has to pass, and lists what is unknown.

The short version

The weakest part of this work is the question of who carries it out: the plan asks the most powerful actor in the system to redesign the system. The second weakest is measurement: extortion, party finance, madrasa graduates’ earnings and the true size of GDP are all poorly known, so several judgements here rest on patterns, not totals. What holds up best under attack is the ordering (candour and banks first), the use of outside deadlines, and the habit of piloting the untested deals with published numbers.

Why publish the objections

Readers who disagree with parts of this analysis are not opponents. The plan is stronger when the people who doubt it test it, and where the evidence is disputed, the site says so. The aim is the same on every side: a prosperous Bangladesh.

Thirteen objections, and what survives each

1. There is no designer

Incentive design assumes someone neutral who sets the payoffs. In Bangladesh the designer is the governing party, which is itself one of the forces. Its record this year (reforms on the independence of judges, police and the anti-corruption body repealed or lapsed; the central bank governor removed; political appointments, per CPD) shows which way it currently leans.1

What survives: the plan does not need a neutral designer. It needs the leadership’s own interests (a full term, a safe exit, a legacy, a budget that funds its promises) to line up with a few of the deals, and it needs outside deadlines for the rest. That is a thinner base than a first reading suggests. Confidence in political feasibility is low (see below).

2. Buying groups off entrenches them

Paying the army in businesses, clerics in welfare funds and parties in public money creates claims that never end. Pakistan’s military economy is the warning.

What survives: sunset clauses and published accounts help, but this risk is real and is the main reason each deal lists its price.

3. “Deals” keep Bangladesh in a deals economy

The country needs rules that apply to everyone, and this plan negotiates group by group. A prominent political-economy argument goes further: growth came from managed rents and predictable misgovernance, conventional good-governance reform gives elites nothing workable, and a deal-making strategy that offers elites new rents in new export sectors in exchange for discipline would be the realistic alternative.2

What survives: the earlier growth ran on ordered deals;3 nobody has shown a route to rules that skips deals. The test of each deal is whether it leaves a rule behind: a bankruptcy law, a term limit, a posting rule, a published fee. The same framework also implies that rents work only when disciplined, and the bad-loan ratio is the signature of undisciplined rents. Bank supervision is therefore a precondition for performance-conditioned industrial policy, not an alternative to it.4

4. Reform comes from dominant executives, not consensus

Georgia’s anti-corruption success came from a centralised executive that fired staff and neutralised spoilers; Korea’s industrial push was authoritarian. The governing party’s supermajority could be an asset, and insisting on constraints may waste it.4

What survives: in Georgia democratic control remained weak and political interference was not curtailed.5 Bangladesh’s own history shows that supermajorities used to entrench themselves (2011) produce the violent cycles this plan seeks to end. The defensible synthesis is to let the majority drive technocratic economic reform fast, but make the political rules a cross-party deal. Separately, one study finds democratisations raise GDP per capita about 20% in the long run.6

5. The scores are invented

Power and direction in the incentive map are judgement. Someone else could score the military as negative, the ulama as strongly negative, or exporters as rent-seekers. A alternative scoring from another source differs on several groups (see that page).

What survives: the method, not the numbers. If you disagree with a score, change it and see whether the order of the deals changes. For most plausible re-scorings, banks and the insurance pact still come first because they act on the largest forces.

6. Evidence from elsewhere may not travel

Audits in Indonesian villages, payment cards in Andhra Pradesh and police reform in Georgia were done in different states with different politics.

What survives: they show that officials respond to monitoring and payment design. They do not show the size of the effect in Bangladesh. Hence pilots.

7. The pain comes first and the government may not survive it

Bank resolution, austerity and subsidy reform arrive while poverty is rising (national poverty has risen four years in a row). Approval of the government already drops sharply among people in financial trouble.7

What survives: targeted transfers through one card, mid-day meals, and candour about the timeline. There is no version of this that is painless, and delay raises the bill: government borrowing from banks already takes 81.5% of its financing and private credit is barely growing, so the slump is partly caused by the unresolved banking system.4

8. The world may not cooperate

The plan assumes energy prices ease, Gulf labour demand recovers and trade partners sign deals. The IMF baseline has the Strait of Hormuz reopening only in March 2027.8

What survives: the domestic half (banks, extortion, credibility) is worth doing in any external scenario, which is the argument for putting it first.

9. It is a cyclical shock, not a structural crisis

Bangladesh grew at a ten-year average of 5.6% with weak institutions. The present slump reflects an uprising, an election and an energy shock; remittances are at records. Wait, and growth returns.4

What survives: partly true for the short run, and the evidence cannot cleanly separate cycle from structure. But the 2029 graduation clock, depleting gas reserves (about 2036), garment concentration and a falling female labour force are structural and dated. Waiting means meeting them with the same model. Measured growth before 2019 may also have been overstated, which weakens “it worked before” (see what the numbers miss).

10. The deals can be captured

A settlement window can be used to forgive friends and prosecute enemies. Public party funding can be pocketed while extortion continues. A simple tax can become one more collection.

What survives: only publication and outside audit protect against this, and both depend on a press and civil society that are themselves under pressure. Hence the stress on feedback loops: right to information, independent statistics, whistleblower protection.

11. Pilots stay pilots

Bangladesh has a long record of projects that worked in one district and were never scaled. The “pilot” label can also be an excuse not to act.

What survives: tie scaling to the 2029 graduation date and to budget lines, not to donor projects. Publish pilot results quarterly and write the scale-up trigger into the pilot’s design.

12. It underrates religion

A sceptic who gives religion more weight would note that one survey in 2013 found 82% of Bangladeshi Muslims favouring sharia as official law, that Jamaat won 31.77% of the vote, that madrasa enrolment is rising while secondary enrolment falls, and that Hefazat can put 20,000 people on the street against gender reform. The sceptic would say the comparisons chosen (West Bengal, the diaspora) are comfortable and trend data are ignored. If religiosity is hardening, or if Jamaat’s vote grows and its positions on women’s work stiffen, the madrasa compact (deal 7) could feed the force it is meant to settle.4

What survives: the evidence that madrasas have responded to incentives is real but dates from the 1980s-2000s. There is no recent trend data on religiosity and no data on madrasa graduates’ earnings. This is the judgement held with the least confidence, and the one most worth testing with new data. The judgement that religion is second-tier is about the past and present, not a forecast: it could change within a decade if the Islamist veto extends to economic policy. That is why the plan front-loads women’s economic freedom while it is still politically achievable. See culture and religion.

13. A version of the first deal has just been tried and thrown out

Section 18(a) offered former bank owners a paid route back and was repealed within five months. Any settlement with large defaulters will be attacked the same way, and the attack will often be right.

What survives: the fiscal problem described by one economist does not go away because the clause did. The design on strategy and deals differs on the points that sank 18(a): owners do not return, terms cannot be met with bank money, and the process is public. Whether that is enough to win consent is untested.

The political-feasibility weakness

Every deal asks someone to give up something they hold. The five points below say honestly where that is most likely and least likely.

  • Where interests already line up. The Finance Ministry needs working banks and revenue to fund its own budget and pay scale. Exporters need tariff, port and gas reform before November 2029. Outside lenders and buyers have deadlines. A leader who hopes to complete a term and hand over peacefully needs the economy not to fail. Foundational learning has a motivated sponsor in the World Bank programme, parents and NGOs, with only the teacher unions (on pay grade) and the coaching industry against it.
  • Where they do not. Reforms that constrain the executive (police, courts, anti-corruption independence) need the executive’s consent; the governing party’s actions in 2026 suggest it currently values control over insurance. Campus depoliticisation meets the student wings of both governing and opposition parties. The coalition that would carry these reforms is the post-2024 student generation, professionals, women’s groups and parts of the business community, which is broad but not organised around one reform.9
  • Outside levers are real but limited. The 2029 graduation date aligns exporters, the EU and the government. The governing party’s insurance interest aligns it with caretaker and commission reforms. The IMF has been rebuffed at times, and FDI investors are too small to lobby.
  • If both fail: the realistic fallback is islands of effectiveness: a few ports, zones, courts and the central bank run well while the broader system stays clientelist. That would still be worth doing, and it is where the plan starts.
  • A note of fairness. The governing party also did things this year that point the other way: members gave up government plots and duty-free cars, parliament repealed the clause allowing former bank owners to reclaim banks, and the army is being withdrawn from street policing.110 These show that change is possible when the public is watching, and they are the opening the plan tries to widen.

Confidence in political feasibility is low

This is the single judgement on this page that most readers should hold loosely. Even if every analytical claim on the site is right, the plan can fail on who carries it out. The aim of publishing the incentive map openly is to widen the number of groups who see something to gain.

Proposals examined from earlier work

Earlier proposals on freedom, anti-corruption and electoral design, and four structural problems, were examined against comparative evidence. The examination broadly agreed with the aims, sharpened the mechanisms and listed what was missing. The conclusions are summarised neutrally here and used throughout the site.

Four structural problems, restated as one system

The four problems usually named (a slow and discretionary bureaucracy, education that rewards certificates, religious-traditionalist resistance to change, and winner-take-all politics) describe real phenomena, but they are the visible surface of one self-reinforcing equilibrium, not independent causes.

  • The bureaucracy is better read as a legacy of an institution designed for revenue extraction under earlier rulers, onto which political patronage was added, than as incompetence: its dysfunction serves the incumbent coalition.
  • Education is a patronage asset (appointments, admissions, hall rooms, tenders), which is why low quality is tolerated.
  • Religion has sincere grassroots belief, but at the political level it can serve as a legitimation of the existing distribution of power.
  • Winner-take-all politics is a contest for the state’s rent machinery, which is why losing is so costly.

Any reform that hits only one corner is absorbed. The loops are on circular dependencies and the root causes on root causes.

Four cases revisited

Case What the examination found Where it landed
Student politics The problem is capture of student organisation by parent parties, not student political engagement (which produced the 2024 uprising). Prefer “separate party machinery from educational institutions and protect autonomous student governance” to “ban student politics”; add a parallel rule for faculty “colour groups” Deal 4
Chattogram port Container handling is a case where global operators usually outperform, but the process was a weak vehicle: no fresh open tender, a naval-adjacent site and verbal labour guarantees; public scepticism was not irrational Deal 5
Rights and anti-corruption laws Bangladesh already has most of the laws; what is missing is an enforcement equilibrium. “Anti-corruption agency theatre” results when new agencies are created without changing it Design rules, Deal 3
Women’s rights The economic argument is the strongest evidence in development economics and avoids a framing trap; use the garment industry and microfinance as proof of concept; treat safety, transport and childcare as labour-market and GDP policy; engage reformist Islamic thought on its own terms Roadmap

Proposal: mindset change and individual freedom

The commitment to individual freedom (conscience, speech, dress, association, career, life choice) is the right normative anchor, close to John Stuart Mill’s harm principle: power may rightly be used over a person against their will only to prevent harm to others. The examination added three points.

  • Sequencing. The evidence is that mindsets change when institutional incentives for coercion change, not the reverse. In Tunisia, in Indonesia after 1998 and in several West African Muslim-majority states, an institutional settlement protecting freedom of conscience came first and the tolerance norm consolidated around it. Bangladesh’s own constitutional history (a secular 1972 text, later amendments inserting a religious invocation and a state religion, and a partial restoration of secularism by a 2010 court ruling) shows the question is not peripheral.9
  • The Millian qualification. The harm principle says the state cannot coerce on purely moral or religious grounds. People remain free to criticise, argue, persuade and disapprove. Criticism is protected; coercion by state, mob or extra-legal violence is not. Stating this explicitly answers the objection that “you are trying to stop us from criticising”.
  • What freedom of conscience requires institutionally. At minimum: narrowing the religious-insult provision and the speech provisions of the digital and cyber laws that have been used against writers; effective state protection of minorities and dissenters (the real test is whether the state stands between a mob and its target); reform of personal-status laws that discriminate by religion in inheritance, marriage and divorce; decriminalising consensual adult conduct that harms no one; and protecting heterodox or satirical speech as speech.

Practical pathway: do not lead with the most divisive points. Lead with freedom of inquiry and the press, women’s rights and the anti-mob-violence point, which are easier to defend and build the protections that make harder ones possible later. The plan here (a liberal floor now; wider social liberalisation zonal and later) follows this.

Proposal: staged anti-corruption enforcement

The insight that strict enforcement cannot be switched on when a whole system is non-compliant, and that pay reform must remove the underlying incentive, is correct. The examination found the comparative evidence does not favour gradual bottom-up tightening where corruption is networked, and recommended:

  1. start at the top, with the first major prosecutions including figures from the sitting coalition;
  2. rebuild a single anti-corruption institution on an independent model (investigation, prevention and education; constitutional independence; a budget line to parliament; open selection with confirmation and a fixed, non-renewable term);
  3. target pay increases at high-risk posts, not across the board;
  4. redesign systems, not only enforce (digitise land registration, licensing, vehicle and passport services, customs);
  5. protect whistleblowers and complainants;
  6. make enforcement visible (a regular prosecution bulletin, mandatory published asset disclosure);
  7. accept the political cost: reform that reaches only the outgoing party is a cycle of political prosecution.

Later analysis accepts items 2 to 6 and item 7 as a test, but favours pilots and islands before top-first prosecutions for the first two years. The disagreement is recorded on strategy and deals.

Proposal: electoral design

The target is right: reduce the stakes of any single election. The three mechanisms sketched (a shorter term, opposition figures in bodies, staggered elections) had mixed support. The effect of a shorter term is small; keeping defeated leaders in various bodies is brittle unless written into law; and staggering half the lower house every two years is the weakest, because a parliamentary government needs a majority at investiture. The more durable package (proportional element, entrenched autonomy for key bodies with cross-party selection and fixed terms, a constructive no-confidence vote, super-majorities for constitutional and electoral change, statutory opposition seats in selection committees and staggered local elections) is set out in the instruments table on strategy and deals.

What the earlier plan was missing

Our assessment is that the four problems and three proposals cover perhaps 40% of what a development plan needs. The missing pieces, and where they are now covered:

Missing piece What was said Where covered
Economic structural transformation East Asian developmental states paired strong institutions with an economic strategy: a competent pilot agency, “embedded autonomy”, export discipline, heavy investment in secondary and technical education, finance aligned with industrial priorities and selective protection with sunsets. Bangladesh has not yet climbed from low-end garments to higher-value manufacturing Deal 5 (performance-conditioned industrial policy after bank supervision works), Roadmap
Press freedom and civil society No inclusive settlement survives without them; a right-to-information commissioner, editorial independence for public broadcasters, narrower defamation and incitement law, source protection, rebalanced state advertising Supporting reforms
Judicial independence Everything depends on a judiciary that can rule against the sitting government; design choices matter Supporting reforms
Climate and demography Climate vulnerability and a demographic window closing around 2030 raise urgency Roadmap (lightly covered; gap)
The religion chapter Not the opening chapter; engage Islamic modernist and reformist thought; be diagnostic, not prescriptive, on faith; tell the state what it is not entitled to do on others’ behalf Culture and religion; editorial rules
Coalitions for reform Name who wins and loses from each reform Who blocks and who could carry each reform
The international dimension India, China, the US and EU, and the Gulf each shape the path Group incentives

Objections to the compact itself

The ground rules (reasoning on the compact across differences) are a proposal, not an existing national agreement. They deserve the same testing as the deals.

“Rules on paper do not bind anyone”

Principles are easy to sign and costly to keep. A group that expects its rivals to defect will not obey first, and where courts and police are not independent, a call for respect for law can become a shield for incumbents.

What survives: the compact is not offered as a substitute for institutions. It works as a public yardstick that makes breaches visible (the repeal of the bank-reclaim clause within five months shows that public reaction can bind a government), and it is tied to mechanisms: a cross-party minimum compact, an independent quarterly scorecard, pilots, and handover records. Where the referees are weak, the rival test (would the sponsors accept this under their strongest rival?) is the check. Signing is a signal; the legal powers and budgets still have to exist.

“The law can be weaponised”

“One law for everyone” can be turned into selective prosecution, mass cases or the use of vague public-order and development grounds to silence critics. Unconditional obedience to a bad law is not a virtue.

What survives: the rule is that the law is constitutional law, applied through reasons, remedies and review; allegation, arrest and guilt are different things; restrictions on rights must be lawful, necessary and proportionate; peaceful protest, which may disrupt ordinary activity, and lawful claims that delay projects are protected. Enforcement priorities are published. This is a safeguard that has to be built into every deal, not an assumption.

“Secularism by stealth”

Devout communities may hear “persuade, never coerce” and “protect individuals within communities” as a way of ranking growth above faith or of weakening religious authority.

What survives: the compact asks for the same thing from every tradition: conscience, worship, teaching and peaceful advocacy are protected, and coercion and exclusion face equal rules. Each tradition is invited to endorse the rules from within its own values, so that no faith is placed beneath economics. The cost is real: individuals within communities keep the right to decline and dissent, and that limits some forms of communal authority. Whether religious leaders will endorse the rules is untested, and the trend evidence is thin (see the gaps below).

“It forgets the victims”

Asking everyone to put the past aside can deny victims a remedy, and shifts into silence what Spain’s 1977 pact deferred for decades.

What survives: the compact does not ask anyone to forget. Credible allegations are investigated impartially across all affiliations, evidence is preserved, victims and witnesses are protected, remedies are accessible, and responsibility is individual and established in fair proceedings. “Timely” means that proceedings are timely, not that victims’ claims expire or that anyone leaving office gains immunity. What is ruled out is reprisal and collective punishment.

Confidence in the compact

The compact makes the plan more robust where it is hardest, on political feasibility, but it does not solve it: it relies on sponsors being willing to accept rules that bind them. We hold this with moderate to low confidence, for the same reason as above.

Tests that any one-cause explanation must pass

Six tests apply equally to the claims on this site. They also explain why the site does not treat religion, corruption or culture as “the” cause.

  1. Cause or outcome? Corruption can cause low capacity and low capacity permits corruption. Reciprocal causation has to be modelled explicitly.
  2. No ecological inference. Aggregate outcomes do not establish the motives, intelligence or diligence of individuals. Poor professional performance may reflect selection, education, incentives and management, not a population-wide deficiency.
  3. Demand counterfactuals. If prayer or afterlife beliefs were major explanations, comparable groups with different observance should show large, robust differences.
  4. Account for omitted variables. Geography, colonial inheritance, trade policy, conflict, energy, infrastructure, capital access and global shocks must sit beside culture.
  5. Examine variation. Differences among Muslim-majority countries, across regions of Bangladesh and over time are evidence, not a nuisance.
  6. Separate private belief from public coercion. A person may pray or dress conservatively without limiting another person’s capabilities. The development concern grows where rules are imposed on those who do not choose them.

Under these tests the mechanisms that survive are gender exclusion, censorship and threats against dissent, clientelism, non-meritocratic recruitment and organisational dysfunction. Claims about lost hours from prayer, a general afterlife effect on effort or universal psychological damage do not meet the same standard in our assessment. The evidence is set out on the culture and religion page.

What remains unknown

Gap Why it matters
Figures from secondary summaries with untraceable references Its numbers (for example the net loss of the banking sector in 2025, “distressed” loans near 60%, productivity per hour, scenario probabilities) are marked unverified unless found elsewhere11
Two economists’ writing Only part of the writing of each could be read (some articles did not open; Facebook and YouTube output not readable)
Primary data BBS, World Bank, IMF and ADB figures were taken mostly from reports and press coverage of them, not from the databases; ADB was not consulted
Total size of extortion Only sector fragments exist
Party finance Some evidence exists (a November 2026 study by Transparency International Bangladesh reports party accounts and estimated operating costs, and TIB has recorded campaign-spending violations by MPs), but no comprehensive total of what elections cost or how parties fund themselves, and informal spending is concealed12
How many cadres there are, and what they earn Unknown; it sets the cost of the cadre-to-career deal
Madrasa graduates’ earnings and employment One imam survey and one district study; no national data
Religiosity trend A 2013 survey is the only comparable trend point found; a January 2025 Pew survey on religion and governance in Muslim-majority countries is listed in the sources but was not reviewed, and its comparability with the 2013 questions is not established13
Cost of each deal Only the bank clean-up and the pay scale have figures
Apprenticeship and public-works evidence in Bangladesh Not searched
Agriculture, climate, health, urbanisation, automation in garments Outside most of the research
Policy status still to confirm The post-October IMF position (partly updated: talks continue), the status of the Commercial Court Ordinance and the police bill
Culture measures Hofstede and World Values Survey trust and values data; diaspora employment and female inactivity
Social restrictions Alcohol and nightlife law and its economic footprint; tourism and creative-sector size; cost of the Friday-Saturday weekend
Diet and heat Rice and oil shares of the diet; monthly wet-bulb data for Dhaka
Comparators Malaysia and Indonesia development comparisons; Pakistan and West Bengal time series
Works cited from memory Baumol (1990), Murphy, Shleifer and Vishny (1991), Olson (1993), Tay and Diener (2011): standard references, not re-checked

Open research questions include: the productivity cost of the female-participation gap in current terms; the constituency-level mechanics of the party-wing recruitment-to-rent pipeline; how much of the bad-loan stock is recoverable and the total fiscal cost; what happened at the universities that banned party-affiliated politics, two years on; the current doctrinal state of the religion-in-the-constitution question; what Bangladesh’s own history of civil-service reform commissions says about why implementation fails; the effect of the 2024 uprising on young people’s participation; term-length effects; what design features would make the anti-corruption commission work more like Hong Kong’s than like most African agencies; and who the right local voices would be on Islamic reformist thought.9

The disputed figures that these pages rely on are collected on the disputed figures page.

What survives overall

Our overall assessment

These are our conclusions, not sourced facts, and their evidential limits are listed above.

After these attacks, the core holds. Bank and revenue repair are the binding near-term constraints. The graduation deadline is the best available commitment device. Political reform should be framed around binding future governments. Foundational learning and women’s work are the highest-return long-run investments. Social liberalisation is a consequence of development more than a cause. The parts that weaken are confidence in political feasibility, the separation of cycle from structure, and the claim to have measured religion’s role precisely.


  1. CPD, Six Months of the Government, 24 August 2026. ↩↩

  2. Khan (2010), core.ac.uk; Khan and Roy, Making anti-corruption real, SOAS ACE; summarised in claude-fable-5-2, “The case against this plan”. ↩

  3. Hassan and Raihan, Navigating the labyrinths of the deals world, ESID. ↩

  4. Prothom Alo on World Bank, FY26 outlook; Pew, 2013 survey on sharia; HRW, Hefazat rally against women’s reform commission. ↩↩↩↩↩

  5. LSE, research online record. ↩

  6. Acemoglu and co-authors, NBER w20004. ↩

  7. Dhaka Tribune on IRI, approval. ↩

  8. Jyoti Rahman, A State of Economic Emergency, Counterpoint, 20 October 2026. ↩

  9. Our assessment. Comparative case details (Singapore, Hong Kong, Georgia): Unverified: no primary source located. ↩↩↩

  10. Dhaka Tribune, army to return to barracks; Dhaka Tribune, bank resolution amendment passed. ↩

  11. Unverified: no primary source located (figures from an unverified summary). ↩

  12. TIB, Institutional governance of political parties in Bangladesh, executive summary, 2026; Dhaka Tribune, TIB: 79.46% of MPs are crorepatis. ↩

  13. Pew Research Center, What role should religion play in Muslim- and Jewish-majority countries? (2025). ↩

Go deeper

Where this comes from

Synthesised from 10 research reports, checked against the sources cited above and reviewed before publication. How we work · All sources.