It may take years before data submitted to Statistics Canada on crimes that have been dismissed as unfounded can be considered truly reliable, even though the first batch of numbers is expected next summer, the federal agency warns.
Yvan Clermont, the director of the Canadian Centre for Justice Statistics, said Statscan plans to phase in changes to its crime-reporting protocols at the request of police-service representatives, who say officers need more classification options to document criminal cases. Mr. Clermont said the roll-out of those changes is still being determined. From there, it will take time for each police service – Canada has about 170 – to adapt.
On Wednesday, Statistics Canada announced that as of 2018, it would resume collecting and publishing statistics on unfounded crimes. It has been 14 years since the agency decided to stop releasing this information despite an outcry from researchers who warned that without the data, it would be impossible to track changes in numbers that many believe reveal how often police believe complainants.
The change of heart comes in response to a recent Globe and Mail investigation that revealed one of every five sexual-assault allegations reported to Canadian law enforcement goes undocumented because officers have dismissed them as unfounded – a police term that is supposed to mean an investigation has shown no violation of the law occurred or was attempted.
In responding to the series, numerous police chiefs stated that their service's unfounded rate appeared artificially high because of errors in the classifications used to signify the outcome of closed cases, known as closure codes. A common complaint among chiefs was that cases reported by a third party, such as a parent or friend, would sometimes incorrectly be classified as unfounded after the alleged victim declined to co-operate with investigators. In other cases, unsubstantiated allegations were being improperly coded as unfounded.
As a result, when the Police Information and Statistics Committee (POLIS) – a subcommittee of the Canadian Association of Chiefs of Police (CACP) – met with Statistics Canada officials earlier this month to discuss the unfounded issue, they asked for more classification coding options. POLIS representatives asked the federal agency to create two new categories, one for third-party reports and another called "founded – not solved," which would capture allegations for which evidence is limited.
"The recommendations provided by POLIS allow police to report such incidents in a more victim-centred manner – one that correctly conveys our belief in the victim regardless of whether or not the incident can be substantiated through the investigative process," CACP director Mario Harel said in a statement on Wednesday.
Mr. Clermont said the closure code definition for unfounded will not change, so he is hopeful that most police services will be able to report those numbers by next summer as planned. However, as the two new options come into effect, it could affect the number of unfounded cases.
"I think the important point here is this is going to be a very transparent process, which is going to highly improve the quality of that information and the comparability of jurisdictions. [It will] offer very interesting opportunities for analysis going forward," he said.
Concerns about the quality of the data originally prompted Statistics Canada to stop collecting unfounded numbers. Police services have been reporting crime statistics to the federal agency since 1962 through the Uniform Crime Reporting Survey. But over time, forces began taking different approaches. The federal agency eventually found that some police services were not reporting unfounded statistics at all. Others were using the code incorrectly.
Holly Johnson, a University of Ottawa criminologist who is one of the few researchers to study unfounded rates, said it will be interesting to see how crime statistics shift now that police services are addressing the unfounded issue. Ms. Johnson said the concern is whether police forces that are blaming coding errors will simply shift the cases elsewhere rather than try to examine whether the cases are truly being handled and investigated properly. A good measure will be if the charge rate also goes up.
"You hope they're looking critically at unfounding, not just the way they're coding, but the way they're reacting to them. One would hope that [the service] is going: 'Okay, we were premature in dismissing this one. Let's go back,'" she said. "If unfounded goes down, what goes up? Is it the charging rate that goes up? We'd hope that's the case. If it's another category, such as 'founded not solved,' you wonder if there's been change or if it's just a public relations exercise."
On Wednesday, as part of the announcement that Statistics Canada would resume publishing unfounded statistics, the Liberal government revealed it would release a 2006 study of unfounded sexual-assault cases by researchers Linda Light and Gisela Ruebsaat. The report, entitled Preliminary Study of Police Classification of Sexual Assault Cases as Unfounded, is one of the most comprehensive and commonly cited studies of the issue in the country – but was never published. When the researchers were in the final stages of releasing the study, which was funded in part by the federal Department of Justice and the British Columbia ministries of community services and public safety, they were told it was being shelved.