Punishment Without Learning: Why Pakistan Must Interview Convicted Child-Sex Offenders

Independently written and published by Shahbaz Shah Legal Journal.

The unresolved question after conviction

Pakistan's response to the sexual assault of a very young child follows a familiar sequence: public horror, arrest, trial, sentence and the closing of a file. The punishment may be severe. The offender may remain in prison for years. Yet when another child is assaulted by a different offender, the State begins again from almost the same point of ignorance.

That repetition does not prove that punishment is pointless. Punishment condemns the wrong, recognizes the child's injury, incapacitates a dangerous person and, where enforcement is credible, contributes to deterrence. The problem is narrower and more important: punishment is not a research method. A judgment answers whether the prosecution proved the charge and what legal consequence follows. It rarely reconstructs, in a form that can be compared across cases, how an adult moved from thought to intention, from intention to access, and from access to the decision to violate a child.

The central policy question is therefore not simply, "Why are these offenders evil?" Moral condemnation is already clear. The usable question is: what sequence of beliefs, emotions, opportunities, calculations and failures of inhibition allowed a person to grant himself permission to rape a two-year-old, a five-year-old or a ten-year-old child? What did he tell himself? Why did he believe the child could be reached, controlled or silenced? Which barrier was absent, and which barrier might have stopped him?

After a conviction becomes final, the State should invite the offender to take part in a voluntary, confidential and structured offence-pathway interview conducted for research and prevention — not for interrogation, publicity, sentence reduction or sympathy.

Interview convicted offenders — systematically

The proposal is deliberately confined to convicted persons. Pending accused persons must be presumed innocent and should not be turned into compulsory research subjects. The first national study should focus even more narrowly on adult men whose final convictions concern rape or another serious contact sexual offence against a child aged two to ten. Juvenile offenders require a separate protocol because development, peer influence, culpability and rehabilitation differ materially.

The interview would not ask an offender to repeat graphic facts merely to satisfy curiosity. It would reconstruct the offence of conviction as a decision process. A trained forensic interviewer would examine the period before contact, the choice of victim, the offender's relationship to the child, the acquisition of access, the use of trust or secrecy, the moment at which internal restraints failed, the offender's expectation of detection, and the conduct used to conceal or rationalize the offence afterwards.

This is not an invitation to accept the offender's story as truth. Convicted offenders may deny, minimize, blame, forget or present themselves strategically. Their accounts must be compared with judgments, investigation records, digital evidence, prison files and other lawfully available material. Contradiction is itself data: repeated forms of denial or minimization can show how offenders protect a self-image and neutralize the child's humanity. But an interview remains a report by the offender, not a scientific window into the mind and not proof of a single cause.

International precedent shows the method is workable

Pakistan would not be inventing an untested method. Queensland's Project Axis examined men serving sentences for sexual offences against children. Official records identified 323 prospective convicted participants; 182 completed a detailed self-report questionnaire, 96 indicated willingness for further participation, and a smaller selected group completed semi-structured interviews. The project was designed to improve prevention, detection, investigation and correctional responses, while preserving independence from prison authorities and protecting identifying information.

A related criminological method, commonly called crime-script analysis, maps the stages through which an offence is prepared, committed and concealed. Researchers can translate recurring steps into situational prevention measures: remove unsupervised access, strengthen boundary rules, increase the likelihood of disclosure, preserve digital traces, or interrupt grooming before physical contact occurs. The value lies not in producing a dramatic psychological label, but in locating practical points at which a future sequence can be broken.

International research also rejects the idea of one universal child-sex-offender profile. People convicted of offences against children are heterogeneous. Some may have a persistent sexual interest in pre-pubertal children; others may offend through broader antisociality, entitlement, opportunism, sexual preoccupation, distorted beliefs, poor self-regulation, or a combination of factors. Not every person who abuses a child meets diagnostic criteria for paedophilic disorder, and a diagnosis by itself does not explain opportunity, victim selection or the decision to act. That diversity is precisely why Pakistan needs comparative interviews rather than one assumed explanation.

Ask how the boundary was crossed

The interview schedule should be standardized enough to compare cases, but open enough to identify pathways the researchers did not anticipate. At minimum, it should examine nine domains:

• Formation of intention. When did the offender first contemplate sexual contact with the child or with a child of that age? Was the thought fleeting, recurrent or progressively elaborated? What converted a thought into a plan?
• Mental permission. What did he tell himself immediately before the offence? Did he deny the child's suffering, minimize the act, claim entitlement, externalize blame, imagine consent where consent was impossible, or treat the child as an object rather than a person?
• Victim selection. Why was this particular child selected? The inquiry must focus on the offender's choice — age, dependency, familiarity, isolation, disability, trust or perceived inability to disclose — and must never shift responsibility to the child.
• Access and opportunity. How did the offender obtain proximity and privacy? Did family trust, employment, teaching, religious authority, transport, domestic work, online contact, shared housing or weak supervision create the opportunity?
• Grooming and control. Were gifts, affection, threats, secrecy, desensitization, testing of boundaries or manipulation of caregivers used? For very young children, was simple physical control or the child's limited language considered sufficient?
• The decision point. What was happening immediately before the offence? Which restraints were present, which failed, and did substance use, anger, stress or opportunity affect timing without becoming an excuse?
• Expected detection. Did the offender think the child would not understand, would not disclose, would not be believed, or would be silenced by the family? Did he expect poor investigation, delayed reporting or weak evidence?
• Earlier warning behaviour. Were there prior boundary violations, attempts to isolate children, suspicious digital activity, sexualized comments, stalking, previous complaints or other conduct that adults observed but did not connect?
• Possible interruption. What concrete measure might have stopped the offence at that time — another adult entering the space, a locked access rule, credible monitoring, a child's safe disclosure channel, faster police action, loss of digital anonymity, or the expectation of certain detection?

The final question is especially important. Project Axis observed that prevention may depend not only on asking what causes an offender to act, but also on asking what failed to stop him. That reframing moves the study from speculation about character to identifiable barriers, opportunities and institutional choices.

A workable national design

Pakistan should begin with a disciplined pilot rather than announce an enormous database before the method is tested. There is no magical sample size. A sound design would proceed in phases:

• Phase one — qualitative pilot. Conduct approximately 40 to 60 in-depth interviews across selected prisons, continuing until the principal offence-pathway themes reach saturation. Test the wording, consent process, interviewer safety and coding framework.
• Phase two — national structured study. Use the pilot to build a validated questionnaire and scale to several hundred convicted participants from Punjab, Sindh, Khyber Pakhtunkhwa, Balochistan and the federal territory, with transparent sampling and refusal data.
• Phase three — prevention translation. Convert recurring pathways into operational recommendations for police, prosecution, prisons, probation, child-protection authorities, schools, hospitals, religious and residential institutions, online-safety teams and families.
• Phase four — follow-up. Where lawfully possible, link anonymized records to prison conduct, treatment participation, release supervision and later reconviction data. Longitudinal results should test which indicators actually predict outcomes rather than merely appearing persuasive in interviews.

The sample should be stratified before analysis. At minimum, researchers should distinguish victims aged two to five from those aged six to ten; family or household offenders from acquaintances and strangers; planned from apparently opportunistic offences; lone from group offending; and first detected convictions from known repeat offending. Categories must be used to compare pathways, not to create a crude label that claims to predict guilt.

The research team should include forensic psychologists or psychiatrists experienced in sexual offending, criminologists, qualitative researchers, statisticians, prison specialists, investigators, prosecutors, legal ethicists and child-protection experts. Survivor-informed oversight is essential, but no survivor should be required to confront or participate with the offender. The study concerns offender decision-making; the child's dignity, privacy and safety remain its reason for existing.

The safeguards are part of the science

Prisoners live in a coercive environment. A request from the State can feel compulsory even when a form says "voluntary". International research standards therefore apply additional safeguards to prisoner participation. Pakistan should treat those protections as a minimum design requirement, not administrative decoration.

• Final convictions and a defined scope. The core interview concerns the offence of conviction. Participants with pending appeals or related proceedings require separate legal review; the safest initial cohort is offenders whose convictions are final.
• Genuine consent. Participation must have no effect on sentence, remission, parole, prison classification, family visits or access to services. Refusal and withdrawal must carry no penalty.
• Independent governance. An independent ethics committee with prisoner-rights and child-protection representation should approve the protocol. Researchers should not report to the participant's prison command or prosecution team.
• Defined confidentiality. The law and consent form must state in advance what is confidential and what must be disclosed, such as an imminent threat or an identifiable child currently at risk. Researchers must never improvise after an interview begins.
• Data separation. Names and prison identifiers should be stored separately from research data; publications must be anonymized; access should be logged; and retention and deletion periods should be fixed.
• No inducement through leniency. The study must not become a route to sentence reduction, nor should apparent remorse in an interview be treated as a shortcut to release decisions.
• No sensational publication. Reports should publish patterns and prevention findings, not lurid transcripts, identifying details or material that could retraumatize families or turn offenders into public personalities.

These protections also improve validity. An offender who believes the interviewer controls parole will tailor his answers. A prisoner who is promised absolute secrecy when the law cannot guarantee it has not given informed consent. Ethical clarity produces better data as well as fairer treatment.

What the State could learn — and use

The study should not end as an academic report placed on a shelf. Each recurring pathway must be translated into a responsible agency, a preventive action and a measure of whether that action worked. The outputs could include:

• Police and prosecution tools. Evidence-led checklists for early digital preservation, corroboration, search priorities, suspect interviewing and identification of grooming or concealment patterns.
• Safer access rules. Institution-specific controls for homes, schools, madrassas, sports settings, transport, clinics and residential facilities where one adult can obtain unobserved access to a child.
• Disclosure systems. Age-appropriate, trusted channels through which children and caregivers can report boundary violations before conduct escalates, with a reliable response that does not punish the child.
• Prison management and treatment. Differentiated programmes based on assessed needs and offence pathways rather than one generic class for every convicted person. Evidence suggests appropriately designed cognitive-behavioural and risk-need-responsivity approaches can reduce, though never eliminate, reoffending risk for some offenders.
• Release and supervision policy. Structured assessment, lawful monitoring and individualized restrictions based on evidence, proportionality and judicial review — not indefinite punishment by label.
• A national prevention report. An anonymized annual publication identifying recurring access routes, failed safeguards, emerging digital methods and policy changes, with outcome measures and independent scrutiny.

The study succeeds only if an interview finding changes a prevention practice: a door is no longer closed without oversight, a complaint is recognized earlier, digital evidence is preserved faster, or an identifiable offence sequence is interrupted before another child is harmed.

Punishment still matters — but severity alone is not a prevention plan

The proposal does not compete with punishment. It begins after guilt has been lawfully established. Accountability, incapacitation and proportionate sentencing remain indispensable. The claim is simply that increasing an already severe maximum sentence cannot tell police how an offender selected a child, cannot tell an institution where access failed, and cannot tell a family which early boundary violation deserved intervention.

General deterrence research also cautions against equating deterrence with sentence severity alone. The perceived certainty of being detected and punished is often more influential than adding further severity to an already serious penalty. Offence-pathway research can support that certainty by improving recognition, reporting, investigation and evidence collection. It can also reduce opportunity. Those are preventive gains a statute book cannot produce by wording alone.

Nor should the persistence of new offences by new offenders be described as proof that nobody learns from punishment. Such a universal claim cannot be established. The accurate conclusion is that punishment of one offender does not automatically transmit knowledge about offence pathways to investigators, institutions or potential victims. If the system wants that knowledge, it must collect and test it deliberately.

Pakistan already has a legal doorway

The Anti-Rape (Investigation and Trial) Act, 2021 already provides an institutional foundation. It establishes a Special Committee and provides for a Sex Offenders Register. Official federal statements in 2025 also recorded work on risk assessment, management, rehabilitation and post-release controls for convicted offenders. The missing link is a national research programme that explains the behaviour behind the names placed in the system.

A practical route would be for the Special Committee to recommend and oversee the commission, while provincial Home and Prison Departments provide lawful access. The interviews themselves should be conducted by an independent university or research consortium selected through transparent criteria. A national protocol, common data dictionary and audited chain of custody would allow provincial findings to be compared without turning research data into an informal police file.

The constitutional basis is protective as well as punitive. Article 9 protects life and liberty; Article 25(3) permits special provisions for women and children; and Articles 35 and 37(d), as Principles of Policy, direct the State toward protection of the child and expeditious justice. Those Principles are not a substitute for enforceable legislation, but they reinforce a public duty to build institutions capable of reasonable prevention. Article 14's protection of dignity also requires the research to respect prisoners and to safeguard victims' identities. A constitutional State does not protect children by abandoning legality; it protects them through legality designed to learn.

What the study must never become

A national interview programme would become dangerous if presented as a machine for detecting future rapists. No questionnaire, brain scan, hormone test or diagnostic label can determine with certainty that a person will offend. The study must not create a public "type" against which innocent people are judged. It must not treat poverty, childhood adversity, pornography use, mental illness or any other background factor as destiny. Risk is probabilistic, and most people who experience a given risk factor never sexually abuse a child.

The study must also resist offender-centred storytelling. Understanding how an offender rationalized the act does not validate that rationalization, reduce responsibility or displace the child's experience. The offender's words are useful only insofar as they help identify a preventable sequence. Explanation is not exculpation.

Finally, researchers must publish limitations. Volunteers may differ from prisoners who refuse. Memory is imperfect. Case files vary in quality. Some pathways will be unique. Results from convicted offenders cannot describe undetected offenders with certainty. These limits argue for careful sampling, triangulation, replication and transparent uncertainty — not for returning to institutional ignorance.

Justice after judgment

Pakistan's criminal justice system is built to decide guilt and impose consequence. It must now develop a second capacity: to learn from the offence after guilt has been finally established. Every eligible conviction should be considered as a possible source of anonymized prevention knowledge, gathered voluntarily, ethically and scientifically.

The governing question should follow every sentence: what did the State learn that can protect the next child? If the answer is only the offender's name, the statutory section and the number of years imposed, the inquiry is incomplete. The file may be closed, but the public-safety problem remains open.

Every final conviction should yield two public goods: accountability for the crime already committed and knowledge capable of preventing another. Punishment tells the offender what society condemns. Structured interviews can tell the State where to intervene. Justice requires both.

This article is independent legal and policy analysis for research and general information. It does not concern the guilt of any unconvicted person and does not replace case-specific legal or clinical assessment.