Autopsy of Police Recruitment: The Stakes and Flaws of a System

Seventeen thousand eight hundred eighty. That is the number of young people who, in March 2026, applied to join the 37th class of the Haitian National Police (PNH), hoping to land one of the 1,200 places offered by the National Police Academy (ENP). With one applicant admitted for every fifteen, the PNH boasts a selectivity worthy of the most prestigious elite schools, which, on the face of it, should inspire confidence in the quality of its recruits.

Yet this figure masks a far more complex reality. The question is not how many applicants are turned away, but what the filter actually selects. A high selection rate loses all meaning if the stated criterion, merit, is not the real basis for sorting. When places become scarce, their market value rises, and that attracts the worst candidates as much as the best.

What is seen on the ground contrasts sharply with the rigorous selection process described on paper. Citizens’ accounts are troublingly consistent: newly graduated officers firing into the air in their neighborhoods to assert their authority, motorists shaken down at the roadside instead of being ticketed, checkpoints turned into tollbooths for market women and transport operators.

Faced with these abuses, the reflex is to blame the officers’ lack of discipline. But that approach comes too late. By the time anyone tries to sanction an officer, he already has the gun, the uniform, and the authority of the state. The problem lies not in discipline but in admission. How can a system that is supposed to be so selective let such profiles through?

The thesis is clear: under the pressure of mass intake, the recruitment system is not simply faulty; it is “inverted.” A faulty filter would let candidates through at random, mixing honest people and predators. An inverted filter is far more dangerous: its very structure preferentially selects those it ought to exclude. Tightening a filter of this kind is counterproductive, because it only raises the price of getting through, a price that only candidates determined to recoup it at the population’s expense are willing to pay.

Urgency as a trap

Police understaffing is a reality that complicates the security response. While the official headcount is 15,498 officers according to the United Nations Development Programme (UNDP), the real number on active duty is reportedly much lower. In mid-2025, The Haitian Times estimated the number of officers on active duty at about 13,500, a figure eroded by departures abroad and by the killings documented by the National Human Rights Defense Network (RNDDH): 35 officers killed between June 2025 and June 2026. The police-to-population ratio, estimated at 1.3 per 1,000 inhabitants on the basis of the official headcount, remains well below the benchmark of 2.2 often cited in international comparisons.

It is this equation, fewer police officers facing growing threats, that led to the launch of the P4000 program. The goal is ambitious: to train 4,000 officers in 18 months, nearly a third of the active force.

Sources: RNDDH, report of 12 June 2026 on the PNH; BINUH; Le Nouvelliste; PNH press releases.

The urgency is real, but it is a formidable trap for the selection process. Everywhere in the world, the argument of urgency is the most effective way to lower standards. The pressure works through two levers. First, the scarcity of places, which become a commodity if selection is not strictly on merit. Second, the compression of training time, from seven months of residential training to about three. That time at the academy was not devoted solely to instruction; it served as a second screen for observing recruits and weeding out the unfit. By eliminating that period, the entire burden of screening was shifted onto entry, making that filter both crucial and vulnerable.

Anatomy of a porous filter

On paper, the process looks robust: written, physical, and medical tests, and character investigations. A closer look, however, reveals major flaws.

The written and physical tests, based on scores and timed results, are hard to falsify on a large scale and in effect account for all of the process’s apparent rigor. The problem lies elsewhere, in the character-screening mechanisms.

The certificate of “good conduct and morals,” the centerpiece of the application, has become an administrative formality obtained for a fee. It involves no real inquiry among neighbors and no thorough verification. In reality, the document attests only to the absence of any known conviction, which is not enough to assess the integrity of a future officer.

At the same time, existing police records are fragmented and go unused. The data lost when 41 police stations were destroyed, weak information systems, and the mass jailbreaks of 2024 have made it extremely difficult to consult the files of the Central Directorate of the Judicial Police (DCPJ) and of the prisons. In gang-controlled areas, violence often no longer leaves any official trace. The state is not blind; it is nearsighted, and that nearsightedness can be cured by better interconnection of data.

Another essential lever is the community reference. Since September 2025, two letters of recommendation have been required, one of them from a ranking police officer. Yet nothing in the public record indicates that these references are systematically checked, even though applicants must provide the signatories’ phone numbers. A reference that nobody calls is a zero-cost signal, and therefore has no selective value.

In the days of the United Nations Stabilization Mission in Haiti (MINUSTAH), this screening was carried out by outside third parties such as the RNDDH or the Vetting Unit, which made it possible to weed out unsuitable profiles (26 dismissals in the 17th class). Since the departure of the UN forces, this external screening function has never really been brought in-house or replaced by a dedicated structure, leaving the process vulnerable.

Finally, the discretionary stages, medical and administrative, have become a breeding ground for allegations of corruption. The ORDEDH report of 29 May 2026, relayed by Vant Bèf Info, speaks of a “heist” of the P4000 program, with candidates pushed aside in favor of those paying up to 75,000 gourdes or benefiting from string-pulling. These allegations call for an independent audit, but they highlight a structural weakness: when decisions are neither reasoned nor published, abuses flourish.

The inversion mechanism: why the predator wins

To understand why the system tips over, we need to analyze the value of a police job. It offers three kinds of reward. First, the salary, a regular income that is precious in an economy in crisis. Second, road money: shakedowns, tolls levied on market women, deals with criminals. Finally, power, that is, the ability to exercise authority over others and to protect one’s own, a rare form of social importance in a country where the state is absent.

Someone who joins the police to serve settles for the salary; for that person, the value of the job is capped. Someone who joins to extract or to dominate collects all three currencies, and for that person the value of the job has no ceiling. In a system where access can be bought, the one who sees the job as a profitable investment, the predator, is willing to pay top dollar to get in. The honest candidate, meanwhile, is discouraged by the cost or by the need for string-pulling, because he cannot make the job “pay for itself.” That is how the inversion happens: candidates are sorted by their appetite for domination.

This mechanism is not theoretical; it feeds on itself. The more recruits there are seeking to recoup their outlay, the more visible predation becomes, which raises the value of the job in the eyes of future predators, and the higher the price at which the filter sells. A careless filter would produce scattered errors; an inverted filter concentrates misconduct where the money is: on the main roads and at checkpoints. The PNH’s Inspectorate General (IGPNH) could confirm this drift by publishing misconduct data broken down by class.

Lessons from history: screening in the middle of a war

The idea that it is impossible to screen during a crisis is contradicted by international experience. Liberia, Bosnia, and Georgia managed to reform their police under extreme conditions, by treating screening as an indispensable wartime operation to prevent infiltration from within.

Four pillars underpinned these successes. First, background investigation: dedicated investigators, each handling a limited number of files, check every reference and visit applicants’ homes; character screening becomes a profession, not a formality. Second, community testimony: in Liberia, publicly posting the names and photos of applicants allowed citizens to report violent pasts, under the supervision of a joint vetting board. Third, registration and certification: in Bosnia, universal registration created the missing records and allowed case-by-case certification, with avenues of appeal to prevent legal abuses. Finally, time: probation, as practiced in North America, keeps the door ajar after training, with actual conduct in the field determining whether or not the officer is confirmed in the post.

Georgia showed that it is sometimes necessary to dissolve a corrupt institution in order to rebuild on sound foundations, while Kenya illustrated the dangers of reform that comes too late, applied to personnel already in place. Between these two markers, the message is clear: screening must happen at the point of entry. The Leahy Law, which prohibits US security assistance to units about which there is credible information of gross human rights violations, is moreover a lever that donors can pull to demand integrity in screening.

Rebuilding: the chain and the academy

To break out of this vicious circle, reform must go through five auditable stations, at each of which every decision can be traced.

The academy must recover its role as a filter. With training compressed to three months, sorting during the course has all but disappeared: in the first two P4000 cohorts, fewer than 2% of those admitted failed to graduate, and the academy has become a mere “sieve.” That observation time must be restored, perhaps by sequencing the training: three months of core curriculum followed by six months of field placement supervised by certified field training officers. This would preserve throughput while restoring the quality of the sorting.

The battle for the door

Haiti suffers not only from the police force it has, but from the police force it is producing at the fastest pace in its history. Two kinds of public force can emerge from these 18 months.

The first, produced by a guarded filter, would arrive in the neighborhoods with legitimacy of its own, and every police check would reinforce the line between the state and the gang. The second, produced by a filter that was sold, would arrive weighed down by the price it needs to recoup, and every roadblock would erase the difference from predation, turning the uniform into a mere racketeer’s costume.

Nothing needed for the first option is out of reach. The pool of applicants is huge, the precedent of MINUSTAH vetting exists, and the methods of reform are known. The donors’ lever is already written into their own law. What is missing is a firm political decision: to treat the institution’s front door as a war front.

The reconquest of order will begin neither on the Kenscoff front lines nor in the Artibonite, but in an assessment room at the National Police Academy. It will play out over a file that someone will choose, conscientiously, either to justify in writing or to monetize. It is at this seemingly trivial scale that the future is decided: the day the market woman sees a uniform approaching on the road, she will know, or will not know, whether the outstretched hand is that of a protector. There is still time to decide what that hand is reaching for.