Surrounded by monitors showing camera views of the city, Detective Marco Christlieb works at the Real Time Crime Center, inside the ӣƵ Metropolitan Police Headquarters in ӣƵ on Thursday, Dec. 21, 2023.
ST. LOUIS — Ness Sandoval wanted to know what happens when there’s a massive investment in a very poor neighborhood.
And the new headquarters in north ӣƵ for the National Geospatial-Intelligence Agency, the federal defense mapping experts, provided the perfect opportunity to find out.
Would the investment reduce crime? Would it boost economic values in the area?
The ӣƵ University sociology professor, a mapping expert himself, worked with graduate student Tara Smith on a of crime trends in and around the site, which they published in 2020.
The study depended on detailed crime data with geographic coordinates for precisely locating incidents. The data came from spreadsheets that the ӣƵ Police Department had long published publicly each month .
People are also reading…
The researchers used the data to map crime hot spots around the future NGA campus. They found that rather than reduce crime, the NGA project had merely displaced it — pushing it into nearby neighborhoods.
Afterward, Sandoval hoped to study the impacts of other big developments in the city, like the new CityPark soccer stadium or the “One Hundred” apartment tower on North Kingshighway near Forest Park.
But there was a problem.
In December 2020, the police department abruptly turned off its public spigot of detailed geographic crime data.
Violent crime was surging to historic highs in the city. Regional stakeholders wanted guidance on how to tackle the problem.
But the department never turned the data spigot back on, leaving community groups, researchers like Sandoval — and the public — without data key to understanding crime trends in the city. And though the department insists it is no longer able to publish detailed crime data for the public, it has provided it to insiders, the Post-Dispatch found.
“What’s frustrating is that the data exists,” Sandoval said, noting that ӣƵ police surely still use geographic crime data internally. “This data belongs to the public. It’s a public good.”
Police spokesperson Evita Caldwell told the Post-Dispatch by email on Wednesday that the department “takes transparency very seriously” and understands frustrations that the data is no longer available. But the loss was beyond the department’s control, she said.
The police department has long blamed a major technology change in 2020 for its inability to keep providing detailed geographic crime data to the public.
To switch to a new way of tracking crime, called the National Incident-Based Reporting System, ӣƵ used a to buy a new computer system. When the department brought the system online in December 2020, its vendor, Optimum Technology, had not developed an alternative way for the city to continue publishing detailed crime spreadsheets for the public.
After four months, the department began posting simple, on its website in PDF files. But experts say those basic reports are no substitute for granular, geographic data.
“The loss of detail in time, place, and specifics about the incident is really disappointing,” said Bobby Boxerman, a doctoral candidate in the criminology department at University of Missouri-ӣƵ, who used the old data in research projects.
Thousands of other law enforcement agencies have also undertaken the NIBRS transition, though, and many of ӣƵ’ peer departments found ways to keep providing detailed incident data to the public, despite the challenges.
ӣƵ County police, for example, stopped publishing new monthly data on the county’s for about eight months while analysts constructed new incident-level spreadsheets based on the NIBRS format. When county police resumed publishing in September 2021, they did so even more frequently, weekly, than they had before.
Sandoval called the ӣƵ police department’s website “embarrassing” when compared with cities like Houston, New York and Baltimore.
“The city has fallen far behind where other cities are today,” he said.
Insider access
Sandoval, journalists and others have tried since 2021 to get detailed, geographic crime data from the department by making public records requests under Missouri’s Sunshine Law. The department consistently denies them.
“There is no responsive record to this request,” wrote Erika Zaza, the department’s Sunshine Law administrator, to Sandoval on Nov. 9. “The department no longer produces the crime reports in that format since we converted to a new system.”
But even as the department continued to take this stance with the public, it was providing similar data to researchers working on a city-funded project.
The city’s health department hired Washington University’s Institute for Public Health to since 2021. The initiative, a pet project of the last two mayors, used residents known as “violence interrupters” to intervene in conflicts and, hopefully, reduce gun violence in three distinct areas of the city.
Because these target areas didn’t align with ӣƵ’ traditional neighborhood boundaries, using the simple neighborhood-level crime totals provided publicly by ӣƵ police without specific location data for each crime would have resulted in an inaccurate analysis, said Theodore Lentz, an assistant professor at the University of Wisconsin-Milwaukee, who helped analyze crime data for the study.
“If all I have are area-level totals, I’m limited to an analysis within those chosen boundaries,” said Lentz, who earned his doctorate in criminology from UMSL.
To evaluate the program properly, researchers needed detailed data with geographic coordinates for each incident, he said. And the ӣƵ Police Department gave them what they needed.
Each month, a police analyst would send the researchers a “geodatabase” file that they could open in ArcGIS, a commonly used mapping software. These files contained the same detailed data that ӣƵ stopped providing to the public after December 2020: incident number, type of offense, date, time, address, neighborhood, geographic coordinates, and more.
Neither Victoria Anwuri, the city’s health commissioner, nor Kimberly Vanden Berg, spokesperson for the health department, would answer questions from the Post-Dispatch about how arrangements were made between the city’s health and police departments and Washington University’s Institute for Public Health to share this detailed crime data with the researchers.
Caldwell, the police spokesperson, told the newspaper she was not aware of the data provided to other city departments, and could not comment.
Rich crime data
Prior to 2021, the ӣƵ police department was considered by some a leader in providing open data, faithfully posting a spreadsheet each month with details of every reported incident.
“We had this really rich crime data for like 10 years,” said Branson Fox, who led a study on the in ӣƵ in 2018 and 2019, when he was a statistical data analyst at Washington University.
Boxerman, the UMSL doctoral candidate, agrees. He used ӣƵ’ detailed data for his own research into fatal versus non-fatal gun violence, he said, and many of his mentors and colleagues used it in their projects.
“The geographic component really allowed researchers like me to see how crime is distributed,” he said, as well as how it affects different places and communities. And having the dates and times each incident occurred makes it possible to break down trends over time, he said.
In Fox’s case, using ӣƵ’ detailed data enabled him and his fellow researchers to find that the risk for homicide and aggravated assaults in 2018 and 2019 was significantly higher near vacant buildings and vacant lots.
The study raised questions with public policy implications that Fox wanted to explore in a new study. Is the city increasing the risk for crime when it demolishes vacant buildings?
“We didn’t get to do that work,” he said. COVID hit. Things slowed down.
But the real killer, he said, was that ӣƵ’ detailed geographic crime data was no longer available.
‘We don’t even have a timeline’
In April 2021, after the ӣƵ Police Department hadn’t posted any crime data at all on its public website for four months, the department posted a on its “Crime Statistics” webpage.
“We understand the importance of this information,” the notice read. “We anticipate the information being available in late Spring.”
Privately, researchers and community stakeholders who depended on the detailed data were frustrated. One such group, a violence-reduction partnership led by members of the Institute for Public Health at Washington University, reached out in April 2021 to Emily Blackburn, the manager of the police department’s crime analysis unit then. The group asked Blackburn for “inside intel” on the department’s failure to post any new crime files for months, according to emails obtained by the Post-Dispatch.
“We are still awaiting that functionality from the Vendor, so we don’t even have a timeline on it,” Blackburn replied, citing a conversation with Jerome Baumgartner, the department’s director of planning and research. “This was promised to be ‘ready’ by implementation, but as IT projects sometimes go it was not.”
By summer, ӣƵ began posting simple, neighborhood-level crime totals publicly, but not the detailed, incident-level spreadsheets with geographic coordinates.
The department also on its webpage, blaming “unforeseen technology delays and incompatibility in comparing crime” as the reason it had not yet resumed publishing the detailed crime data, and promising it was “in the process of revamping our Public Portal to make it easier for citizens to access crime data.”
Today, nearly three years later, the remains atop the webpage.
Boxerman called it unacceptable.
“I understand the challenges of working with this kind of data, and that police departments face a tremendous amount of pressure, and that data organization probably isn’t a top priority,” he said.
“That being said, the data have been offline for almost three years.”
The public portal
Documents show that the department was indeed working on a new public portal in 2021.
Optimum Technology, the vendor developing the public portal, released an early version for the department to test in June 2021, according to internal status reports from the police department’s app development team, obtained through public records requests. One month later, the department had reported “defects” in the portal for Optimum to fix. By September 2021, reports show Optimum had resolved some issues, and was still fixing others, while the department itself was buying a security certificate and setting up a web address for the portal.
That same month, Optimum billed ӣƵ $132,350 for work developing the public portal, plus two other components of the records management system. Nearly half of that, $60,800, was for the portal, documents show. The full bill was paid in December 2021.
But after January 2022, the internal status reports show no further progress on the portal, even though representatives of the police department told the Post-Dispatch in May 2022 and November 2022 that its development was still going. By December 2022, the portal ceased to be mentioned in the status reports at all. Later that month, Robert Tracy was announced as the city’s next police chief.
Asked about the status of the portal, Caldwell, the police spokesperson, said Wednesday she understood it remains a work in progress. She said the IT department had found Optimum’s early version of the portal to be “less informative and less user friendly” than the Missouri Highway Patrol’s statewide data portal.
And while development of the portal slowed, internal status reports show that the police department was still accommodating the data requests of insiders.
In July 2022, the office of Mayor Tishaura O. Jones asked the police department to provide a crime data feed “as soon as possible” to make it easier to produce a crime data dashboard for her. The project was completed by mid-November 2022.
And Chief Tracy requested early this year that IT produce Compstat reports for him in a customized format. Work on the project continued through at least mid-May, but it wasn’t clear from the documents if it was ever finished.
Caldwell said the department would begin publishing those Compstat reports in the first quarter of 2024.
Meanwhile, regional leaders continue pushing to make the city a geospatial leader. In May, the city hosted the GEOINT Symposium for military and intelligence officials, and private companies. SLU announced in November that it would add 20 geospatial-focused faculty members. And the NGA’s new $1.7 billion campus is on track to open in 2025.
So it’s perplexing, experts said, that the city’s police department hasn’t prioritized finding a way to publish detailed geographic crime data for the public.
“They were the gold standard,” Sandoval said. “But now, ӣƵ is the anomaly.”
