Internal security: the difference between a secure state and a police state
In August 2026, Iraq’s Ministry of Interior and the United Nations Development Programme discussed how to address technology-facilitated violence against women and girls, from receiving a report to protection, investigation and referral. This was a preparatory workshop, not an announcement that the problem had been solved or evidence of a decline. Yet it offers a practical test of what security means: can a woman facing blackmail seek state protection without having her private information leaked or becoming the subject of accusation herself? That question is closer to people’s lives than comparisons of agency or camera counts. UNDP statement, 20 August 2026.
A government may possess extensive powers of detention and surveillance while citizens struggle to secure consistent protection against violence and extortion. Conversely, restrictions on force may be clear while the institution responsible for protection remains weak or slow. Choosing between more authority and more freedom therefore does not adequately explain internal security. We need to establish what authority actually does, whom it serves and who can review its decisions. A secure state, as the term is used here, reduces dangers and protects rights through accountable legal capacity. A police state’s coercive logic begins when expanding control becomes the measure of success, regardless of the harm prevented or the right violated.
Protection is a function that can be tested
Articles 15 and 37 of Iraq’s Constitution distinguish protecting life, security and liberty from restricting them arbitrarily; they prohibit torture and reject coerced confessions. These are not benefits an institution can defer until security has been established. They define the outcome the state was assigned to produce. If private abuse ends only to be replaced by official coercion that the victim cannot challenge, the operation cannot properly be recorded as an unqualified security improvement. The source of danger has changed while the citizen remains unable to resist it.
Institutional assessment should separate capacity from its use. Capacity includes qualified personnel, verifiable information, resources and a timely response. Legitimacy concerns the legal grounds for intervention, its limits and the proportionality between action and danger. An institution may fail through inadequate capacity despite a lawful mandate, while an efficient institution may employ its tools for an unlawful purpose. Calling both situations “weakness” obscures the remedies: the former requires professional development and disciplined funding; the latter requires changes to incentives and oversight, alongside accountability for those who diverted the mandate.
The distinction becomes visible when a service is designed around the victim. Receiving a report is not the same as ending harm, and transferring someone between agencies does not become an integrated response simply because official correspondence exists. I would assess whether responsibility is clear, information protected, appropriate support accessible and the complainant able to understand the progress of the case within investigative limits. These are proposed service measures, not claims that national data on them are already published. Their value is to make the citizen’s experience examinable rather than leaving an institution to measure activity by the paperwork it circulates.
When incentives redefine success
Ferguson in the United States illustrates how an institution can deviate within a constitutional system. The Justice Department’s investigation, published in March 2015, found that an excessive focus on generating revenue, together with racial bias, contributed to policing and court practices that violated constitutional rights. This case establishes nothing about the prevalence of the same practice in Iraq and is not a judgement on all American policing. It identifies a specific mechanism: when a financial outcome becomes an institutional objective, authority can be redirected away from public safety. Official investigation findings.
The same distortion can arise through non-financial indicators. If officials are rewarded for arrest numbers alone, ease of arrest may take precedence over evidential strength or harm prevention. If they are punished for rising complaints, an incentive may emerge to discourage recording rather than address causes. This does not make every increase in enforcement abusive or every decline an improvement. The more precise conclusion is that an indicator requires accompanying evidence to reveal its meaning. Detainee numbers alone, for example, tell us little about case quality, waiting times or the outcome of judicial review.
Reducing enforcement is not an objective in itself either. An increase in reports during a particular period may reflect greater willingness to seek protection, or a genuine increase in victimisation. Distinguishing between them requires comparing recorded reports with victimisation surveys, service users’ experiences and case outcomes, rather than choosing the interpretation most comfortable for management. This is a methodological proposal, not a statistical estimate for Iraq. Public security suffers when attractive figures are rewarded and uncomfortable information punished, because institutions lose the ability to see changes beyond their reports.
Early oversight costs less than correcting injustice
A joint United Nations Assistance Mission and human rights office report published in August 2021 provided an Iraqi example of the gap between law and implementation. It drew on interviews with 235 people deprived of liberty and covered July 2019 to April 2021, identifying problems with access to defence, medical screening and detention oversight. This is a defined sample and period, not a contemporary national torture rate for 2026. Its continuing analytical value lies in identifying the points of failure through which abuse can occur despite written safeguards. UN report and summary.
Early intervention at these points serves both justice and investigative quality. A case built on testable evidence differs from one that depends on continued isolation of a person to hold together. This does not imply that every confession is false; it means its value cannot be separated from the circumstances in which it was obtained and the supporting evidence. Correcting an error years later does not automatically restore lost time or damaged trust. Oversight during a procedure therefore matters more than merely announcing that a complaint will be possible after the damage is complete.
Oversight also requires a clear distribution of responsibilities. Management can review discipline and performance, but internal review alone is insufficient when the complaint concerns the chain of command itself. Courts can rule on legality within their jurisdiction, yet their effectiveness also depends on timely access to files and relevant information. Public scrutiny needs aggregate data that do not expose victims or investigative details. The choice is not between total secrecy and publishing everything: sensitive information can be protected while complaint numbers, resolution processes, reasons for delay and corrective measures are disclosed.
This structure protects professional staff as well. Where orders are documented, responsibilities known and procedures reviewable, there is less scope to blame a junior employee for a decision they could not change. Accountability for abuse becomes fairer than collective accusations against an institution. A police officer should have a claim to serious training, appropriate tools and clear lawful instructions, in return for observing the limits attached to the role. Employee rights and citizens’ rights may conflict in particular circumstances, but they do not require two different institutions.
What should change in the security test?
In my assessment, Iraqi reform should begin with a bounded trial in a group of stations selected through published criteria, connecting response quality to rights protection. The Ministry of Interior would establish a service baseline; judicial and oversight bodies, each within their jurisdiction, would monitor procedural safeguards; and an independent mechanism would assess service users’ experiences. Comparisons would cover report-handling time, repeat harm, substantiated complaints and access to services. Results should be interpreted against differences in populations, resources and case types, so comparison does not become a punishment for stations confronting harder conditions.
This requires neither a promise of an immediate crime reduction nor importing an entire foreign model. The task is to test whether changes in service, oversight and incentives improve specified outcomes, then expand what proves useful and adjust what fails. It forms part of the broader inquiry into internal security, intelligence and law enforcement: making security an institutional responsibility with a purpose, limits and evidence, rather than a label that protects any procedure from scrutiny.
Returning to the victim of blackmail, state success is measured by its ability to protect her, establish the offence and preserve her dignity within the law, rather than by how many agencies can inspect her phone. A secure state needs power it can exercise, limits it can observe and institutions capable of correcting their mistakes. Without the last condition, reassurance depends on the goodwill of whoever holds authority. That is too fragile a foundation for lasting security.
Ali Zuweid — Researcher in strategic affairs and national security
Cite this article
Ali Zuweid. Internal security: the difference between a secure state and a police state. 9 November 2026. https://alizuweid.com/articles/internal-security-safe-state-and-police-state/?lang=en
