Ever since President Donald Trump issued a presidential executive order declaring a "crime emergency" in the District of Columbia Aug. 11, there’s been a debate not just about public safety, but the reliability of the data shared.
While D.C. officials point to their publicly available data as illustrating that crime is at a 30-year low, the White House has retorted those claims, asserting the city’s crime stats “likely significantly understate the level of crime in Washington, D.C.”
Additional scrutiny has come from the D.C. Police Union, who even before the federal surge of law enforcement officers, stated that, “MPD [Metropolitan Police Department] leadership is deliberately falsifying crime data, creating a false narrative of reduced crime while communities suffer.”
A congressional committee announced Aug. 25 that it is now investigating allegations of “widespread” crime data manipulation, requesting documents, information and transcribed interviews with employees about the topic.
A government can move toward transparency by publishing its data, yet still find itself in a fight to defend the accuracy of those very numbers.
The debate suggests that posting numbers alone is not enough.
As Nick Hart, president and CEO of the Data Foundation, a nonpartisan think tank, explained, building trust in data requires a philosophy of “radical transparency” that goes beyond a public data portal and into the data governance itself.
“Citizens need to understand not just what the data shows, but how it was collected, when it's updated and what it can and cannot tell us,” said Hart, in an email.
Despite being in the crosshairs of a data accuracy debate, Hart pointed to the fact that Open Data DC serves as “an excellent model,” for an open data platform by including crime data from the last several decades based on crime reports, as well as easy-to-use tools for citizens and researchers.
“This level of transparency and accessibility is exactly what other jurisdictions should aspire to achieve to see the frequency of different types of crimes and changes in trajectories at a localized level,” said Hart. “This detailed level of open data is actionable for communities that may want to discuss specific interventions for different types of crimes, and also for evaluating the success of different programs over time."
Despite the delivery of an open data portal, the city remains under scrutiny over the numbers, which reinforces the fact that trust can’t be confined to a web interface.
There are other practices that state and local governments can include with the numbers that could help establish additional transparency, including any changes to how data was collected or classified.
“Transparency about changes to data classifications is always helpful because disclosure of the changes helps users understand whether differences in crime statistics reflect actual changes in crime or changes in how incidents are categorized and recorded,” said Hart. “Without this context, users often can't properly interpret trends over time, across sources and jurisdictions.”
Additionally, data dictionaries or metadata, which defines and describes the characteristics of the data itself, is sometimes overlooked when building public data portals.
“Users need to understand what constitutes an 'assault' versus a 'battery' or how domestic violence incidents are classified,” said Hart.
The D.C. controversy reveals that while technology can provide the tools for transparency, it cannot, on its own, promise trust.
An open data portal that recognizes this will not only report the numbers, but the context behind them, as well as openly address disputes or problems with the data's integrity, ensuring that residents trust the full story behind the numbers they're viewing.
