A SINGLE district police memo tells the story. The language of ‘permission first,’ the stress on ‘orders,’ the presumption that authority flows in one direction: these are not quirks of one officer on one day. They are the reflexes of a legal order built to govern subjects rather than to serve citizens. Much of Bangladesh’s day-to-day criminal justice still runs on statutes drafted in the 19th and early 20th centuries — the Penal Code of 1860, the Code of Criminal Procedure of 1898, the Evidence Act of 1872, the Police Act of 1861 and secrecy provisions descended from the Official Secrets Act of 1923. We have translated, patched and amended these laws. We have not replaced the philosophy that animates them.
Why do they endure? The first answer is path dependence. Courts, prosecutors, police academies and law faculties have been trained on these texts for generations. Every routine, from filing a general diary to seeking remand, sits on forms, precedents and habits that presume the old codes. Rewriting a statute is easy compared to re-engineering the surrounding machinery, retraining every actor, and paying for the transition. The second answer is political. Colonial codes are convenient for any government that wants discretionary tools: vague ‘public order’ clauses, sweeping search and seizure powers, preventive detention, and criminal defamation. Reform that strengthens rights and accountability looks, from the vantage point of power, like a self-imposed constraint, so it drifts down the to-do list.
A third reason is the lure of legal certainty. However antiquated, these rules are predictable. Business groups, banks and bureaucracies often prefer an old but familiar playbook over a multi-year overhaul that could generate litigation spikes and transitional confusion. The fourth is judicial conservatism. Thousands of appellate decisions rest on these codes. Judges schooled in that lineage understandably hesitate to pull up the floorboards that give them interpretive anchors. Add to this the technocratic habits of the aid industry — digitise here, pilot there, train a cohort somewhere — and you get a decade of incremental ‘rule-of-law’ projects that treat symptoms while leaving the trunk of the tree untouched. Finally, language matters. Authoritative versions often live in English and plain-Bangla versions are inconsistent, so many frontline officials navigate law through circulars and unwritten practice rather than through clear, citizen-facing statutes. The culture that flows from all this is unmistakably colonial: command and control, permission before rights, procedure before service.
Keeping the Raj’s rulebook has costs. Rights become fragile because vague offences, ‘annoyance,’ ‘mischief,’ ‘seditious tendencies’ — and expansive police powers chill speech and assembly. Even when the courts eventually correct excesses, the process is the punishment. Policing remains reactive and force-centred rather than service-centred; an 1861-style force was designed to quell, not to protect. Justice drags because old procedures plus massive case backlogs produce years of under-trial detention and routine remand. Poor defendants plead simply to end the ordeal, while the well-advised litigate to delay. The economy pays too. Outdated evidence rules and secrecy laws frustrate digital services, media innovation and data-driven industries. Compliance uncertainty becomes a hidden tax on growth.
If this diagnosis is right, what would a cure look like that doesn’t break the state? We do not need a revolution; we need a disciplined rewrite with clear guardrails. The first step is institutional. Parliament should create, by statute, an independent Law Reform Commission with multi-party nominations, secure tenure, and a mandate to draft complete replacements — not cosmetic amendments — for the Police Act, the Code of Criminal Procedure, the speech-related chapters of the Penal Code and the cluster of secrecy and communications laws. The commission’s job would be legislative carpentry, not political theatre: produce exposure drafts on a fixed timetable, consult widely, publish clause-by-clause explanations and table final bills for an open vote.
Every draft should embed a proportionality test. Any restriction on fundamental rights must be lawful, necessary for a legitimate aim, and the least restrictive means available. That test, already part of modern constitutional practice across Commonwealth courts, should be written into every provision touching public order, speech, search, seizure and surveillance. Alongside this, Bangladesh should finally decriminalise defamation, moving reputation disputes to civil law with capped damages and a fast-track right of reply. Archaic speech offences that criminalise insult or ‘annoyance’ should retire from the statute-book. None of this would prevent the state from prosecuting direct incitement to violence or targeted harassment; it would only end the criminalisation of robust democratic disagreement.
The policing framework needs a 21st-century charter. A replacement for the 1861 Act should recast policing as a public service under democratic control and civilian oversight. That means an independent complaints body with binding powers; clear, reviewable rules for custody, interrogation, and use of force; body-worn cameras and chain-of-custody protocols; statutory duties for community policing, victim care and crime prevention; and transparent, merit-based postings and promotions. Good officers would benefit most because professionalism, not proximity to power, would drive careers.
Criminal procedure must be rebuilt for speed and fairness. The default in non-violent offences should be bail, not incarceration. Remand and custodial interrogation should be strictly limited and judicially reviewable. Charge-framing and disclosure should be time-bound so that trials begin while memories, and evidence, are fresh. E-warrants and e-summons should be standard to cut delay and discretion. The Evidence Act should be modernised for a digital world, with clear rules for electronic records, chain-of-custody and forensic science, and with a statutory exclusion of coerced confessions rather than a fragile reliance on precedent alone.
Secrecy-first laws should give way to a harm-based National Security Information Act that fits alongside the Right to Information regime. Definitions must be tight, not elastic; oversight should be independent; and whistle-blowers who expose wrongdoing in the public interest should be protected, not punished. None of this weakens legitimate national security; it strengthens it by making secrecy credible rather than convenient.
Reform will fail if law remains a priestly language. Government should publish a free, authoritative ‘Bangladesh statutes’ website with continuously updated texts in Bangla and English, annotated with illustrations and flowcharts that explain procedures to citizens and frontline officials alike. Law schools should teach these new codes, police academies should retrain their cohorts, judicial education should adapt, and legal aid should expand so that the benefits reach real defendants rather than staying on conference panels.
Return to the memo that prompted this discussion. Its tone is not uniquely ‘bad.’ It is entirely predictable when the rules are designed for command, not consent. When we keep 1860s laws, we get 1860s behaviour. If we want different conduct from our institutions, we must give them different rules to live by. Bangladesh has already modernised payments, logistics, and digital identity in ways that impressed the world. Law is the next infrastructure. A smart, rights-respecting rewrite will not weaken the state; it will make it more legitimate, predictable, and effective — for citizens, for honest officers and for investors.
Colonial laws remain because they are useful to power, familiar to institutions, and invisible to most of us most of the time. They will disappear only if we replace them deliberately and transparently, with the humility to design a system for citizens rather than for rulers. That is a political choice, not a technical quibble. A 21st-century republic cannot be run on a 19th-century rulebook.
Ìý
Md Obaidullah, a visiting scholar in the department of development studies at Daffodil International University, is a graduate assistant of political science at the University of Southern Mississippi, USA.
