A Practical Guide to Analysing Police Procedures

A guide to analysing police procedures begins with a hard rule: do not judge an investigation by its final outcome alone. A conviction, an acquittal, a cold case or a public controversy may tell us where the matter ended. They do not, by themselves, establish whether police followed the evidence, preserved options for later inquiry, or tested the theories that mattered.

For anyone examining a contested Australian case, the proper question is more exacting: what did investigators know at each point, what should they reasonably have done with that information, and what record shows that it was done? That approach moves discussion beyond suspicion and into verifiable procedure.

Start with the investigation as it existed at the time

Hindsight is useful, but it can also distort. An investigator in the first hours of a disappearance does not have the benefit of later witness accounts, forensic results, media reports or court evidence. A fair review separates what was available then from what emerged later.

Build a chronological record before drawing conclusions. Include the first report to police, the initial attending officers, the decision to treat the matter as a welfare concern, missing-person investigation or suspected crime, and every significant change in direction. Record times where possible, not just dates. In many cases, the difference between 7.00 pm and midnight explains why an opportunity was available or lost.

A timeline should also identify gaps. If a witness was not spoken to for weeks, a vehicle was not examined promptly, or a location was not searched until after public attention increased, that does not automatically prove negligence. It does create a question that requires an answer: what information was held at the time, who made the decision, and what was the operational reason?

Identify the procedural benchmark

Police procedure is not one fixed document that answers every operational question. It is a combination of legislation, agency policy, training, common investigative practice, forensic capability and the practical conditions at the scene. The benchmark will differ between a sudden death, sexual assault, missing child report, armed robbery and historic homicide review.

The most reliable analysis compares actions against the requirements that applied at the relevant time. Procedures change. Digital evidence practices in the 1990s were plainly different from those expected now. Equally, basic obligations were well understood long before modern technology: secure a scene, record observations, identify witnesses, preserve exhibits, document decisions and disclose relevant material.

This is where public commentary can go wrong. It is easy to condemn an investigation because a modern technique was unavailable or not routinely used decades ago. It is equally wrong to excuse a basic failure by calling it a product of its time. A sound review distinguishes between those two situations.

Examine the first response and scene management

The first response often shapes the entire case. Initial officers may have only minutes to assess danger, provide assistance, separate witnesses, protect a location and call for specialist support. Their notebooks, body-worn video where available, radio logs, photographs and incident reports are central records.

Ask whether the scene was properly identified. A crime scene is not always confined to a room, a vehicle or a patch of ground. It can include access routes, nearby bins, a footpath, a creek bank, a phone box, clothing, a discarded item or the digital spaces connected to the event. Investigators should be able to explain why boundaries were set where they were and whether they were expanded when new information emerged.

Continuity matters. Every exhibit should have a documented path from recovery through packaging, storage, examination and court production. A break in continuity does not necessarily make an item worthless, but it can affect reliability and admissibility. Look for labels, property records, forensic submission forms, seals, dates and the names of officers handling the item.

Scene contamination needs careful language. Members of the public, family, emergency services and police may enter a location before it is recognised as significant. The relevant issue is not whether anyone entered. It is whether entry was recorded, risks were managed, and later interpretation accounted for what could have been altered or introduced.

Test the investigative hypotheses

A competent investigation does not start by proving a preferred suspect guilty. It starts with workable hypotheses and tests them against the evidence. In a death investigation, those possibilities may include accident, suicide, misadventure, third-party involvement, or an unknown sequence of events. As evidence develops, some explanations become less likely. The record should show why.

A warning sign is tunnel vision: when information consistent with one theory is pursued energetically while inconvenient material is minimised, delayed or left untested. Tunnel vision can arise from pressure, early assumptions, unreliable intelligence, a persuasive witness or the understandable desire to solve a serious crime quickly. Its presence is not proved merely because police formed a view. It is shown by the way contrary evidence was handled.

Look for alternative suspect enquiries, checks on alibis, elimination processes, unexplained forensic results and witness accounts that did not fit the prosecution or police theory. Then ask whether they were investigated proportionately. Not every lead warrants the same resources. A vague rumour is different from a named person with opportunity, relevant knowledge and corroborating circumstances.

Assess witnesses by process, not personality

Witness evidence is often treated as a contest of who appears believable. Procedure is more useful than instinct. When was the account first obtained? Was it recorded contemporaneously? Did the witness provide a formal statement, participate in a recorded interview, or give information only through another person? Were there later changes, and can they be explained?

Investigators must manage the risk of contamination. Witnesses who discuss events together, follow intensive media coverage, see photographs, or are repeatedly approached may unintentionally blend memory with later information. This does not mean their evidence is false. It means the development of the account matters.

Identification evidence requires particular care. A witness may be confident and mistaken. The fairness of the identification process, the witness's prior familiarity with the person, viewing conditions, elapsed time and any exposure to images all deserve scrutiny. A photographic identification conducted after the witness has already seen a suspect's image in the media has obvious limitations.

Follow disclosure and the paper trail

In a major case, the investigation is documented across statements, running sheets, notebooks, occurrence records, exhibits, intelligence reports, forensic material, call data, photographs, warrants and correspondence. The absence of a document can be as significant as its contents, particularly where an action is said to have occurred but no contemporaneous record supports it.

Disclosure is not a technical afterthought. Material capable of assisting the defence, undermining the prosecution case or opening a reasonable line of inquiry must be properly identified and dealt with under the applicable legal process. Reviews should examine whether relevant material was retained, indexed, provided to prosecutors and disclosed in a usable form.

This is especially important in older cases. Records may have been archived, copied poorly, lost, misfiled or held by different agencies. An incomplete file does not establish misconduct. But it does limit confidence in claims that every line of inquiry was pursued or every relevant item was considered.

Separate error from misconduct

Police investigations are conducted by people under pressure, often with imperfect information. Errors occur. A missed question, delayed statement or poorly worded report may reflect workload, inexperience, a genuine mistake or a more serious failure. The evidence must determine the conclusion.

Misconduct involves more than an outcome nobody likes. It may include deliberate suppression, fabrication, improper influence, knowingly false records, unlawful conduct or an intentional refusal to pursue material evidence. Those are grave findings and should never be made casually. Strong language is not a substitute for documents, testimony and chronology.

At the same time, accountability does not require proof of corruption. An investigation can be seriously deficient because of poor supervision, inadequate resourcing, faulty assumptions or a failure to revisit early decisions. For victims, families and accused people alike, the effect can be profound.

Write findings that can be tested

The final discipline is how conclusions are expressed. State the source, the action or omission, the relevant benchmark and the consequence. For example: a witness was identified in the first report, no statement appears until a later date, the explanation for delay is not recorded, and the delay may have affected memory or the ability to test the account. That is stronger than simply claiming police ignored a witness.

Use careful terms. “Not established” is not the same as “did not happen”. “No record located” is not the same as “no record exists”. “Inconsistent with procedure” is not automatically “unlawful”. Precision protects the integrity of the review and gives others a fair opportunity to test it.

Graeme Crowley Investigates approaches contested matters on that basis: follow the chronology, locate the original material, test competing explanations and identify what the record can genuinely support. A police file should never be treated as beyond question, but neither should it be dismissed for the sake of a more dramatic story.

The useful closing question is always practical: if this case were reopened tomorrow, what specific document, witness, exhibit, scene, decision or unanswered inconsistency should investigators examine first? That is where analysis can move from criticism towards a meaningful search for justice.