Why the Police Keep Choosing the State Over Their Own People

The state does not need police to love the rich; but only to fear disorder more than they resent inequality.

It’s not that the police are traitors to their class, but that class is only one of the identities competing for a person's loyalty.

(Photo: Vibhushita Singh/The Quint)

On July 20, 2026, at Jantar Mantar in Delhi, an Additional Deputy Commissioner of Police named Sandeep Lamba was filmed slapping a woman protester during the Cockroach Janta Party's Chalo Sansad march. He was one face in a week of them: a Rapid Action Force officer photographed sneering at demonstrators, students describing tear gas and baton charges over a demand as unglamorous as an inquiry into paper-leaked medical entrance exams.

The Chief Justice of India would eventually decline to fast-track a hearing into the allegations, and the government's position, when it finally answered in court, was that no one had actually filed a formal complaint.

What makes scenes like this so durable, decade after decade and country after country, is a fact that sits awkwardly at their center: the people doing the hitting are not the status quo. Constable pay in Delhi starts modestly; the officers in riot gear at Jantar Mantar go home to the same crowded neighborhoods, the same inflation, the same struggle to get a child into a decent school, as the students they were beating.

A constable and a protesting student are, by any economic measure, closer to each other than either is to the minister whose resignation was being demanded. And yet the baton, almost without exception, comes down on the side of the minister.

How the State Manufactures Loyalty

This is not a paradox unique to India in 2026. It is close to a universal feature of modern states, and it deserves an actual explanation rather than a shrug.

The most useful place to start is the theory of the state itself, and specifically the French Marxist philosopher Louis Althusser's distinction between two ways a state secures obedience. One is the Repressive State Apparatus: the army, the police, the courts, the prison, which functions, as Althusser put it, predominantly by violence.

The other is what he called Ideological State Apparatuses: schools, churches, media, the family, trade unions, even sport, which function predominantly by ideology, by shaping what people believe is normal and unquestionable before any baton is ever raised.

Althusser's genuinely useful move is to insist that the repressive apparatus is never purely repressive. It, too, runs on ideology, on getting the people who staff it to believe in the mission before they believe in the paycheck. Police academies do not simply teach marksmanship and law; they teach self-conception. Recruits are trained to see themselves as professionals whose loyalty is to the law in the abstract rather than to any protester's particular grievance. 

This is what Althusser called interpellation: the process by which an institution takes an individual and turns them into a subject who recognizes themselves in the role offered to them. A recruit walks in as a citizen with a paycheck to earn; a few years of academy discipline, uniforms, oaths, and canteen culture later, he answers to the name ‘officer’ before he answers to the name of his own class position.

Crucially, this identity does not require police to think of themselves as defenders of the rich. Quite the opposite, since most police forces cultivate an ideology of service, sacrifice, and impartial law enforcement that feels, from the inside, like the furthest thing from class loyalty.

That is precisely what makes it effective. An officer who believed he was protecting billionaires would feel like a mercenary. An officer who believes he is protecting order from the mob feels like a public servant. The state does not need police to love the rich; it only needs them to fear disorder more than they resent inequality.

Hierarchy Before Solidarity

Layered on top of that ideological formation are structures that have nothing to do with belief at all.

The first is the chain of command. A constable who refuses a lawful order to disperse a crowd is not making a philosophical statement. He is committing a disciplinary offense that can end his career, his pension, and in some jurisdictions his liberty. Insubordination is punished with a speed and certainty that inequality itself never is. This alone does enormous work in explaining compliance that looks, from outside, like conviction.

The second is what the political theorist Nicos Poulantzas, writing partly in dialogue with and partly against Althusser, called the ‘relative autonomy’ of the state: the idea that the state is not simply an instrument wielded directly by capitalists but a set of institutions that organizes the interests of capitalism as a whole. Police forces are hierarchical precisely so that the state does not have to persuade every individual constable of anything. It only has to control the officers who control the officers.

The third is the psychological wall that forms between us and the crowd. Riot policing puts an individual constable inside a formation, behind a shield, facing people who are, in that moment, shouting at him personally, sometimes throwing bottles at him personally - as Delhi Police alleged happened at Jantar Mantar, with an officer's head left bleeding. Whatever the protesters' economic position, in that instant they are not comrades; they are the threat in front of him. Solidarity is a slow, cultivated thing. Adrenaline and a shield line are immediate.

Bangladesh and the Limits of State Power

None of this makes police loyalty to the state a law of nature, and the exceptions are as instructive as the rule.

The clearest recent case is Bangladesh in 2024. Through July of that year, as students protested a quota system for government jobs, the police did what riot police almost always do: they fired tear gas, rubber bullets, and in many documented cases live ammunition at unarmed demonstrators, killing hundreds over the course of the movement in one of the most violent crackdowns in the country's history, according to the UN human rights office's own investigation.

On the night of August 4, with Prime Minister Sheikh Hasina ordering troops to enforce a curfew and crush the protests, the army's generals made a different decision. They would not fire on civilians, and told her so. Within a day, Hasina had fled to India by helicopter. The police themselves fractured only afterward, and in a way that is arguably more revealing. As public fury turned on them, twenty-five officers were reportedly killed on the day of Hasina's resignation alone, and dozens more in the weeks around it. 

Police across the country simply stopped showing up for duty. It took until mid-August, and a promise from the interim government that their grievances would be heard and the discredited top leadership replaced, before all 639 police stations nationwide were operating normally again.

The mutiny wasn't solidarity with the students, by most accounts it was closer to rank-and-file officers refusing to keep absorbing the consequences of a command structure that had ordered them to kill on behalf of a leader who was, by then, gone. It illustrates Poulantzas's point precisely. The repressive apparatus is not one seamless will but a hierarchy that can be held together only as long as its lower ranks believe the hierarchy itself is still standing.

Why Defections Are Rare

Serbia in October 2000 offers a cleaner version of a genuine, in-the-moment defection. As hundreds of thousands of protesters converged on Belgrade after Slobodan Milošević tried to nullify an election he had lost, riot police initially fired tear gas to hold the parliament building.

Then, as crowds broke through, most of the police simply stopped resisting and joined the protesters, according to multiple contemporary accounts of that day, a switch made easier by the fact that some police units had reportedly maintained sympathetic contact with opposition organizers beforehand, and that miners and soldiers sent to break an earlier strike had likewise declined to act against people whose economic complaints they shared. The army, notably, stayed in its barracks and let the outcome be decided in the street.

What Serbia and Bangladesh share is a state that had already, visibly, lost the argument. An election annulled in one case, a prime minister who had fled in the other. Defection is a bet on who will hold power tomorrow, and it becomes rational only once the answer looks obvious. That is a colder, more instrumental story than the one where workers in uniform rediscover their class, but it fits the evidence better.

None of that theoretical machinery excuses what happened to the students at Jantar Mantar, and none of it is offered as a defence of Officer Lamba or the RAF constable who called protesters cockroaches back. It is an attempt to explain why the outcome was so predictable. The government did not fall, the courts moved cautiously rather than urgently, and the police who cracked down face, at most, individual complaints.

Under those conditions, Althusser's machinery runs exactly as designed, ideology supplies the officer's sense of righteous duty, hierarchy supplies the threat that keeps him in line if the ideology ever falters, and the shield line supplies the adrenaline that makes solidarity, in the moment, feel like the more foreign idea.

It’s not that the police are traitors to their class, but that class is only one of the identities competing for a person's loyalty. Modern states have gotten extremely good, across a hundred years of institution-building, at making sure that when the moment of choosing arrives, it isn't the one that wins.

(Pratyusha Pan is an author covering culture, history, and society in Kolkata, West Bengal. This is an opinion piece and the views expressed are the author's own. The Quint does not endorse or is responsible for them.)