The argument over police reform in Nepal did not begin with the Gen Z movement of September 2025. It did not begin with the return of democracy in 1990, the peace agreement of 2006 or the Constitution of 2015 either. It is much older, and it is easy to say out loud. How does a state move from a police that protects the ruler to a police that protects the people, their rights and the constitutional order?

That question runs through almost all of modern Nepali history, from the Rana palaces to the streets of Baneshwor. The path leads to one uncomfortable conclusion. Nepal has changed its political system again and again, but it has never fully rebuilt the idea behind its policing. An institution can outlive a regime and quietly carry the regime’s habits with it: Not only the buildings, ranks and procedures, but the old assumptions about power, authority and obedience.

So the test of a democratic police is not the uniform it wears, or the law that governs it, or even the government that sits above it. The test is plainer. Is police power guided by law, evidence and accountable command, or by obedience to whoever happens to hold power? A democratic police is one that must never be a powerless one. It has to be effective, professionally independent, bound by the Constitution, answerable to the public and trusted by citizens, all at the same time. That is hard to build. Nepal has been trying for three-quarters of a century.

An inheritance from the Ranas

Nepal Police’s own historical material traces policing roles back to the Lichchhavi period, with offices such as Sarvadandanayak and Mahasamanta, and to the Malla period, with posts such as Dware, Chautara and Sipahi. By the Rana years, there were kotwals, faujdars, darogas and chowkidars, along with police stations (thana), police posts (chowki), the Amini and the Ram Dal, a military police force. The Ranas gave modern policing its family tree.

But a family tree is not a model to follow. Nepal Police’s own account admits that under Rana rule, forces such as the Amini, the Ram Dal and the militia were used to serve the ruling order and to crush democratic movements. When coercive institutions are built to protect a ruling order, the people who inherit them receive more than offices and ranks. They receive a way of thinking.

A democracy still needs a police

When the Rana regime fell in 1951, the new rulers faced a practical problem. They could not simply abolish the machinery of force. A democratic government still needs people to prevent crime, investigate it, keep public order and protect lives and property. The Congress-led Mukti Sena was turned into the Rakshya Dal and given policing duties. A police headquarters was set up, with Toran Shumsher JB Rana as its first Inspector General, and the Ram Dal, civil police, militia and constabulary were slowly folded together.

The Police Act of 1955 gave this arrangement a legal home. Its preamble spoke of preventing and detecting crime and keeping peace and order. The same law also placed the police under the government’s supervision and control, and Section 4 required officers to follow its orders. Both halves are right. A democracy needs civilian command of its police. But civilian command must itself have limits: the Constitution, the courts, Parliament, human rights bodies and professional standards. Without them, ‘civilian control’ can shrink into something smaller and uglier, which is simply political control.

When protecting the regime became the job

In December 1960, King Mahendra dismissed the elected government, arrested Prime Minister BP Koirala and other leaders, and began the party-less Panchayat era. The police survived, but its purpose changed. Nepal Police’s own history describes the force in those years as focused on safeguarding the Panchayat system and suppressing party activity and democratic movements.

It would be unfair to say that every officer abused power, and this series does not say so. But independent records point to a pattern, not a handful of bad apples. Amnesty International documented the detention of students, journalists, trade unionists and political opponents, and reported torture and ill-treatment in police custody. Later accounts described unofficial detention sites, detainees moved from one police facility to another, and the near-impossibility of getting a court to look into what had happened. Police power was turned against student movements, against the unrest of 1985 and against the People’s Movement of 1989–90. One historical account even holds that police brutality against a peaceful protest outside the Pakistan Embassy, held over the execution of Zulfiqar Ali Bhutto, ended up feeding public anger against the Panchayat system itself.

The courts were not entirely absent. Habeas corpus existed. But a right is hard to enforce when the institution you are complaining about holds the evidence. If officials can simply deny that they have a person in custody, move him elsewhere or hide where he is kept, a right on paper offers very little in real life. That problem would return after 1990, and again after 2006.

New rights, old habits

The return of multiparty democracy in 1990 changed the question. It was no longer “how do we keep order?” but “how do we keep order without turning the police against dissent?” The new Constitution strengthened protection of liberty, due process and freedom from torture. Yet those rights were now to be carried out by an institution shaped under Panchayat rule. It was a new house of rights, run by the old staff.

The Police Reform Commission of 1992 was a serious attempt to change that. Nepal Police itself describes it as part of an effort to modernise the force and bring it closer to the norms of democracy: professionalism, better investigation, human rights awareness, training and public-oriented policing. But the change stayed half-finished, and royal interventions in 2002 and the royal takeover of 2005 pushed it further off course.

Underneath lay a question that no commission fully answered. The issue was never whether the government should control the police. In a democracy, it must. The issue was whether an officer’s career, meaning transfers, promotions, postings and investigations, could be shielded from party interest. When transfers become punishments, promotions become bargaining chips, cases are opened for some and ignored for others, and top posts go to those with the right access, a police force can stay civilian in name and be political in practice.

That weakness was about to meet its hardest test. In 1996, a war began, and the police were sent to its front line.

Next: Part 2. The war, the disappeared, and the question of who investigates the police.