Crime indicators to enhance police management control

In a modern state, management control is not merely an administrative formality; it is the fundamental pillar for evaluating the effectiveness and efficiency of public policies. This principle becomes critically relevant in the realm of public security, where the state primarily depends not only on its ability to prevent and prosecute crime but also on the very legitimacy of the institutions responsible for protecting citizens. It is in this context that the implementation of Law 21.332, known as the Carabineros STOP Law, must be evaluated.[1], whose regulations, published nearly two years late in June 2023, present serious deficiencies that, rather than optimizing management, could be hindering the state’s preventive and prosecutorial actions.

The objective of the law is, in theory, impeccable: to establish a transparent and data-driven police management system aimed at crime prevention through the analysis of criminal trends and the monitoring of deployed actions. However, the regulatory practice reflected in the law’s regulations has distorted this objective, crystallizing at least five structural shortcomings:

1. Improper aggregation of dissimilar categories (incivilities and crimes). The system groups 21 radically different classifications or families of crimes: crimes, offenses, and incivilities. This homogenizing approach is a serious conceptual error in criminal policy, where the gradation of social impact is essential. A homicide cannot be metrically equated to a public consumption infraction; its impact on the victim and the community, as well as the state response required, are incomparable. Analyzing them together distorts any indicator of criminal variation at the communal, regional, and national levels.
2. Fusion of divergent preventive strategies. The regulations amalgamate crimes whose prevention and prosecution strategies are specific and, therefore, must be measured separately. Two examples are telling: on one hand, homicides are grouped with femicides, whose scenarios (public space/intramural) and perpetrator profiles are diametrically opposed. On the other hand, “vehicle theft and its accessories” consolidates high-violence crimes like carjacking with a theft of a vehicle mirror. The impact on citizens, the sense of insecurity, and police action for each are different; measuring them separately would reflect this complexity.
3. Confusion between preventive input and criminal output. The platform combines metrics of police actions (inputs) with data on crime commission (outputs). In the case of incivilities, inspections are measured (positive actions that are desired to be maximized since they are preventive efforts and often end in an infraction), while in crimes, events are counted, a negative outcome that is sought to be minimized as it indicates a higher level of crime.
4. Delayed and incomplete implementation.  The regulations establish the inclusion of victims based on their sex and age group—fundamental for preventive policies—and the incorporation of minutes from sessions at the jurisdictional level of police stations and prefectures that are held (art. 8), which is essential for ex post management control of commitments made in STOP sessions. Unfortunately, this information is still not available on the platform despite its relevance for criminal dynamics and the monitoring of agreements made in monthly sessions.
5. Duplication of platforms and potential distrust in data. The coexistence of the old STOP system and the new “STOP Law” is unproductive and counterproductive. It generates operational confusion and, more seriously, sows distrust in the official data due to disparities in information loading times (with delays of up to 24 hours). The lack of a single source of statistical truth undermines the credibility of the system for decision-making.

Transparency and effectiveness in security management are non-negotiable public goods. The STOP Law was born with the potential to be a beacon in the formulation of evidence-based policies. However, its implementation threatens to obscure its purpose and diminish its utility.

Recommendations:

It is imperative, therefore, that the competent authorities review the current regulations. Correcting these shortcomings is not merely a technical issue; it is a condition sine qua non for transforming data into intelligent decisions. To this end, it is proposed, at a minimum:

1. Urgent reclassification of nomenclatures.A clear taxonomy must be established that distinguishes between crimes of greater social significance (homicides, rapes, violent thefts) and offenses of incivilities. Each category must have specific and separate performance indicators.
2. Disaggregation of critical crimes.It is a priority to disaggregate the statistics of femicides and domestic violence from other homicides and injuries, and to segment vehicle thefts by modality (carjacking vs. theft of accessories or vehicles parked in public spaces). This will allow for the design of targeted police strategies and the evaluation of their real impact.
3. Separation of process and outcome metrics.The platform must clearly differentiate management indicators (inspections, preventive patrols) from outcome indicators (variation in crime frequency). Only then can it be assessed whether police actions are generating the desired preventive effect.

The demand for public security requires, first, a management control that meets the challenge. Perfecting the STOP Law is the first indispensable step to deploying resources efficiently and ultimately advancing in the reduction of crime that rightly concerns citizens.

---

[1] Prior to the issuance of the regulations, the STOP platform was used; today there is a parallel platform called STOP Law. Although some crimes are present in both platforms, data often does not match due to a 24-hour delay. It is worth noting that the STOP platform has been non-functional for months.