Inter-Agency Coordination in the Criminal Justice System

Textbooks describe the criminal justice process as a single sequence, from a reported offence through investigation, prosecution, adjudication, and sentencing to corrections. In practice each of those stages is run by a different organisation with its own statute, its own chain of command, and its own budget line.

In India, the police fall under state home departments and are governed largely by the Bharatiya Nagarik Suraksha Sanhita 2023, which replaced the Code of Criminal Procedure 1973 with effect from 1 July 2024.

Prosecution is handled by a state's directorate of prosecution, courts are administered by the judiciary through the state high court and the district judiciary, and prisons sit under a separate prisons department. In the United States, police departments are typically municipal or county agencies, prosecutors are elected district attorneys or appointed federal attorneys, courts are run by a separate judicial branch, and corrections departments answer to state governors.

In England and Wales, policing, the Crown Prosecution Service, His Majesty's Courts and Tribunals Service, and His Majesty's Prison and Probation Service are four distinct organisations under different reporting lines: policing sits with the Home Office, the Crown Prosecution Service is superintended by the Attorney General rather than by either of the other two, and the courts service and the prison and probation service both sit within the Ministry of Justice.

This separation exists for good reasons. Keeping the police who investigate a case institutionally distinct from the judiciary that tries it, and both distinct from the corrections service that carries out a sentence, is a basic safeguard against the concentration of coercive power in a single body.

The cost of that safeguard is that no single agency owns the case end to end, and no single manager is accountable for what happens to it as a whole. Each agency is what organisational theorists call a functional silo, a unit that optimises its own workflow and its own reporting metrics without a mandate to track outcomes once the case leaves its hands.

The silo effect is reinforced by performance measurement. Police forces are commonly assessed on cases registered, arrests made, or charge sheets filed within a statutory period. Prosecutors are assessed on conviction rates or successful plea outcomes. Courts are assessed on disposal rates and pendency reduction.

Corrections agencies are assessed on custody safety, escape rates, or reintegration outcomes. None of these measures rewards the agency for the ease with which the next agency in the chain can pick up the file. A police force under pressure to raise its charge-sheet filing rate has an incentive to close its own file quickly, even if the file is thin in the exact respects a prosecutor will need at trial months later.

Three handoffs recur across jurisdictions as the points where coordination is most likely to break down. The first is investigation to prosecution. A police investigation produces a case diary, witness statements, and forensic reports built to the evidentiary standard the investigating officer understands, but a prosecutor reviewing the file for trial readiness often needs additional clarification, corroborating documents, or a chain-of-custody record that the investigating team did not anticipate.

Where prosecutors have no early role in shaping the investigation, as has historically been the case in much of India's police-led model, defects surface only after the charge sheet is filed, sometimes long after witnesses have become unavailable or evidence has degraded.

The second handoff is prosecution to court. Docketing a case for trial requires the court registry to receive the charge sheet, list witnesses, and schedule hearings around the availability of the police, the forensic laboratory, and the defence, none of which the court controls directly. A single unavailable witness or an unreceived forensic report can force an adjournment, and adjournments compound because each rescheduled date competes for a slot in an already crowded cause list.

The third handoff is court to corrections. A remand order or bail order, or a final sentence issued by a court, has to be communicated to the jail holding the accused, and any delay in that communication can mean a person spending extra days or weeks in custody despite a court order that should have released them, or a convicted person left in an ordinary lockup rather than being moved to serve a sentence in the facility the sentence contemplates.

A fourth, less visible handoff runs backward: from corrections and courts to the police records that determine whether a person is treated as a repeat offender, and from the court's final disposal back into national crime statistics. If a conviction or acquittal is not promptly reflected in the police's own criminal-history database, a subsequent investigation of the same person can proceed on an incomplete record, and national crime figures compiled from police returns can misstate conviction outcomes for years afterward.

The backward handoff matters most at the point where a court, prosecution, or corrections outcome feeds a decision that another agency has to make later. A bail hearing on a new charge is decided in part on the accused's recorded criminal history, and if a conviction from a different police jurisdiction has not yet been posted to that history, the court is deciding on an incomplete record.

A sentencing court applying an enhanced penalty for a repeat offence relies on the same records, so an unposted prior conviction can understate the sentence a statute otherwise requires, while a wrongly retained entry for a matter that ended in acquittal can overstate a person's risk in a later, unrelated investigation.

National crime statistics are built on the same backward flow and inherit its gaps. India's National Crime Records Bureau compiles its annual Crime in India report from returns submitted by state police, and a final court disposal that has not yet been reported back to the police station that registered the case will not be reflected in that year's conviction figures, even though the case has in fact concluded.

The Federal Bureau of Investigation's Uniform Crime Reporting and National Incident-Based Reporting System programmes in the United States face a comparable limitation: participating agencies report arrests and offences fairly consistently, but case-disposition data flowing back from courts to the arresting agency is far less complete, which is why national conviction and clearance figures are treated as directional rather than exact by researchers who work with them.

This is why the reforms discussed later in this topic target the backward flow as deliberately as the forward one. A shared case identifier that lets a prosecutor pull a police file is only half the design; the same identifier has to let a police record system pull the court's final disposal automatically, or the backward handoff reproduces the identical coordination failure in reverse, just further from public view because no pending hearing forces anyone to notice the gap.

The most direct response to coordination failure has been to build shared digital infrastructure so that a case record created at one agency can be read, rather than re-entered, by the next. India's Crime and Criminal Tracking Network and Systems, commonly called CCTNS, digitised police station records nationwide so that a first information report, an arrest, or a charge sheet is captured once in a common format rather than in whatever paper or local system a station used before.

Courts have run a parallel digitisation effort under the e-Courts project of the Supreme Court's e-Committee, which computerised case filing, cause lists, and case status across district and high courts, and which now underpins online access to case status for litigants and lawyers.

Digitising each agency separately does not by itself solve coordination failure if the systems cannot talk to each other, which is the interoperability problem. A police database and a court database can each be internally efficient while using different case identifiers, different classification codes, and different data formats, so that a case cannot be tracked automatically as it crosses from one system into the other.

Building interoperability requires agreeing on a common case identifier and common data fields across agencies that were never designed to share them, a technically modest step that is organisationally difficult because it requires each agency to modify a system it already controls to suit a use case outside its own mandate.

The same problem appears outside India. In the United States, the Federal Bureau of Investigation's Criminal Justice Information Services division runs national systems such as the National Crime Information Center for shared police records, but state and local courts, prosecutors, and corrections agencies largely run separate case-management systems, so interoperability is negotiated state by state rather than through a single national platform.

In England and Wales, HM Courts and Tribunals Service has moved criminal case management onto a shared digital platform intended to let police, the Crown Prosecution Service, and the courts view a case file through a single system rather than exchanging paper or emailed documents, directly targeting the same handoff problem described above.

India's most direct attempt to solve interoperability is the Interoperable Criminal Justice System, developed under the e-Courts project. It links the police database, court case-management data, the prosecution's own tracking system, prisons, forensic science laboratories, and the fingerprint bureau, so that a single case identifier can, in principle, be used to trace a matter from a first information report through to its final disposal and any subsequent custody.

A comparable ambition sits behind the Crime Multi Agency Centre, developed under the National Crime Records Bureau to let police and other law-enforcement agencies share information on crime and criminals with each other in real time, rather than each holding an isolated slice of the same case.

Technology alone does not fix a coordination problem that is partly organisational. Several jurisdictions pair digital linkage with structural reforms that put a person, rather than only a database, at the joint between agencies. Liaison prosecutors embedded with a police investigation team from an early stage can flag evidentiary gaps before a charge sheet is filed rather than after, shortening the investigation-to-prosecution handoff described earlier.

Victim-witness coordinators, used in various forms in the United States and increasingly recognised as good practice in India's district-level victim assistance schemes, are responsible for keeping a witness informed of hearing dates and case status even as the file itself moves between police, prosecution, and court, which reduces the loss of witnesses that drives a large share of coordination-related attrition.

Case conferencing, a scheduled meeting between the investigating officer, the prosecutor, and sometimes a court official before a matter is listed for trial, is used in parts of the United Kingdom to agree on a realistic trial date and confirm that the required evidence and witnesses will actually be available on that date, rather than discovering a gap only when the case is called in open court.

None of these reforms replaces a shared database, but each addresses a different failure mode: liaison prosecutors fix defective handoffs before they happen, victim-witness coordinators keep the human participants a case depends on from disappearing, and case conferencing catches scheduling mismatches before they consume a court's limited trial time.

Poor coordination has consequences beyond the individual case that is lost. Delay compounds across a system that already runs with limited court and prosecutorial capacity, so every case that stalls at a handoff occupies a slot on a cause list or a police officer's caseload that could otherwise be used to move another matter forward.

Duplicated work is common: a document lost between agencies has to be recreated, a witness statement recorded once by police is sometimes recorded again by the prosecution because the original cannot be located in the format required for trial, and a bail or remand order can be re-litigated because the order communicated to the jail differs from the order the court actually passed.

The least visible cost is to victim and witness confidence in the process. A victim who reports an offence, cooperates with an investigation, and then finds that the prosecutor's office has no record of a document the police confirm they submitted experiences the coordination failure as institutional indifference, even where no single official acted in bad faith. Repeated experiences of this kind depress future reporting, which in turn affects the dark-figure problem in official crime statistics discussed elsewhere in this corpus.

Because coordination failure has no single agreed metric, researchers and court administrators typically rely on proxies. Pendency, the stock of cases awaiting disposal at a given date, captures accumulated backlog but not where in the process a case is stuck. The disposal rate, cases decided as a share of cases filed in a period, shows whether a system is keeping pace with new filings but not whether disposals are timely or whether cases dropped out through attrition rather than adjudication.

A widening gap between the rate at which police file charge sheets and the rate at which courts dispose of the matching cases is one of the more direct diagnostic signals of a coordination bottleneck specifically at the prosecution-to-court handoff, distinct from a general capacity shortage that would show up as slow disposal across the board.