Organisation: The Globe and Mail (Canada)

Publication Date: 03/21/2017

Size of team/newsroom:large


Unfounded is an in-depth, 20-month Globe and Mail investigation into how Canadian police services handle sexual-assault cases. Our goal was to help Canadians understand the systemic issues with how sexual-assault cases are handled in this country, and in so doing, drive change in government policy. In reporting this story, Globe reporter Robyn Doolittle came across a small study from an Ottawa law professor into something called police service “unfounded” rates. When police complete a criminal investigation, they give it a code to signify the outcome. One of those codes is “unfounded,” which is used when the investigating officer believes an allegation is baseless and that no crime occurred. When Doolittle tried to find out how often this code is used, she found that Statistics Canada had stopped collecting those numbers in 2002. The only way to obtain the data would be to use freedom of information legislation to obtain numbers from each police service individually. The Globe decided that in order to get an accurate national picture, we would need to collect data from the roughly 175 police forces in Canada. Only through a comprehensive analysis would we be able to provide all Canadians with the data for their own communities, so that we could start a national conversation and inspire case reviews and policy change. Our probe revealed that from 2010-2014, 1 in 5 people who reported sexual assault to police had their case dismissed as “unfounded.” Once a case is classified “unfounded” it is no longer considered a valid allegation. It is not reflected in local or national statistics and is scrubbed from public record – as if it didn’t happen. The Globe’s data revealed that sex-assault complaints are nearly twice as likely to be designated unfounded as physical assault allegations. The Globe built a visual data story in order to walk readers through our main findings. At any point while exploring the national picture, they can use our look-up tool to search their own police jurisdictions to see the unfounded rates for sexual assault and physical assault, as well as the charge rates for sexual assault. To try to understand why so many cases are falling apart, the Globe investigated 54 sexual-assault allegations. The results showed that from coast to coast, police officers are ignorant of consent law, neglecting sex assault investigations, and relying on rape myths. In total, 36 agreed to share their stories publicly. We included these stories in a separate interactive to show the human side of the data. In its reporting, the Globe conducted hundreds of interviews with people who regularly interact with sex-assault victims, including police officers, Crown attorneys, legal scholars, trauma experts, advocates, activists, criminologists and health care officials. The Globe filed more than 300 freedom of information requests to obtain the data from police services, as well as copies of police investigative files.

What makes this project innovative? What was its impact?

So often when people discuss how the “system” is failing sexual-assault victims, the discussion tends to focus on one-off anecdotes. The Globe’s series was the first to use data to look comprehensively at the problem and initiate change. The response to the Globe’s series was swift. The federal government announced $100 million to combat gender-based violence, citing the Globe investigation. The ministers in charge of Public Safety, Justice and Status of Women vowed action in the form of better oversight, training and policies. To date, 54 police services – about a third of the country’s forces – have committed to auditing thousands of sex-assault cases going back to 2010. The series also prompted a national debate over whether police services should be forced to adopt a revolutionary oversight model that’s been running in Philadelphia for the past 17 years, in which outside advocacy groups are invited to review case files once a year. “Sexual violence, sexual assault, is still far too prevalent, not just on campuses but in workplaces and in communities across the country… As we’ve seen from the excellent bit of very deep investigative reporting The Globe and Mail has just put out, it is still not taken seriously enough by our society,” Prime Minister Justin Trudeau said days after the series began. Statistics Canada stopped collecting unfounded numbers in 2002 over concerns that police were misusing the category or simply not documenting unfounded cases at all. But even when the federal agency gathered this information, it was not available on a jurisdiction by jurisdiction basis. The Globe’s interactive tool was the first of its kind to make this information widely public. Our analytics showed high time spent, as readers across the country explored the data and looked up their own communities. The project was top-performing on our site for days after the launch, and the term “unfounded” was trending country-wide due to the social media response.

Technologies used for this project:

First, we sent freedom of information requests to all the police forces in Canada. There were so many that we used Mail Merge from MS Office to populate the request letters and envelopes with the respective force’s contact details and address. Then, we built a custom data intake and verification tool and workflow using KeystoneJS (Node and MongoDB). This proved to be invaluable in the coming months as we input the data provided to us in various formats. For shared access to the raw responses, we used Dropbox. We had to get a sense of what the data was telling us before it was all in and verified, so we used R to process a CSV export from the intake tool, run analysis, and generate both a data file for the interactive as well as an Excel workbook for reporting. The R process allowed us to continually check overall results throughout the reporting process and quickly update the data files as we received late responses to our request or found errors through the verification process. The R process itself was verified independently by a data-scientist from outside of editorial at The Globe using SQL and Excel to reproduce our calculations and make sure our conclusions were accurate. Our visualizations were done using a combination of D3, our in-house charting application Chart Tool, QGIS and Illustrator. Additional interactive maps and visualizations were also created for analysis and internal use. To build the interactive that told the stories of 36 people who had reported sexual assault, and in order to present the core stories of the series in a uniform way and link them together, we created templates using Node, Gulp, ArchieML and other technologies. They allowed for the quick editing and publishing of multiple related stories as the series unfolded. We were also able to create modules that would pull data from our interactives into the long reads to further tie the project together and illustrate the stories.
Follow this project

Comments (0)

You have to be connected to contribute

You have to be connected to follow

Leave this project and no longer be informed about this project

By joining this project, you will be informed by email when an update or a new contribution is posted on the website.

Thank you for your active participation !

The GEN Community Team