Arnold Ventures criminal justice data report aligns with SpotCrime's approach to open crime data

Arnold Ventures, a non-profit with focus on criminal justice reform, recently published the report “Because the Road to Reform is Paved by Data. CAMPAIGN FOR CRIMINAL JUSTICE DATA MODERNIZATION”.

The report notes that effective police reform relies on gathering data in order to effectively assess the criminal justice system and support reforms and restore public trust.

SpotCrime is no stranger to navigating the bureaucratic and technical divide between police agencies when it comes to collecting public crime data on a national level. All police database infrastructures are made up of different systems, making it hard to collect data on a national level.

The report makes 6 recommendations, that we’ve included below, and outlines a total of 36 action items for anyone willing to tackle this problem and make data collection across jurisdictional lines seamlessly easy. The report asks for the Office of Justice Programs (OJP) to act as the overarching power to enforce data standards and practices in order to strengthen police data in order to accurately determine how to move police reform forward.

We’ve been advocating for open, equal, and fair access to crime data for the past 10+ years. We noticed that some of the recommendations and action items SpotCrime has been advocating for are mentioned in the report! See our notes on each recommendation - and how SpotCrime has even tried to address some of the issues ourselves below:

AV Recommendation #1: Establish an accurate baseline of facts about the criminal justice system, and envision a 21st century system.

An action item being the creation of a National Commission on Criminal Justice Data Modernization. This would allow for federal oversight and assign a federal program to create and maintain standards and create and enforce policies and procedures on collection of data across systems and to the public.

SpotCrime created the SpotCrime Open Crime data Standard (SOCS) over 7 years ago in an attempt to help agencies standardize RMS and CAD data released to the public. Standards are definitely needed for any datasets coming from multiple agencies. Although it may seem like common sense to standardize data, because there is no nationally upheld standard, it has yet to occur. Right now every single police agency releases their data differently.

Additionally, we’ve noticed there is no urgency to restore data feeds that fail. Currently, the open crime data feeds for large agencies like King County (WA) Sheriff, Boston, and Chandler are down. King County has been down since 2019. Chandler for the past few months. None of the agencies have communicated a timeline as to when these open data feeds will be restored. Right now police agencies are only accountable to themselves. A commission would introduce a level of data accountability for police agencies ensuring feeds do not go offline, and if they do it is remedied quickly.

AV Recommendation #2: Radically increase accountability of the justice system through data transparency.

This recommendation includes an action item to publish a dashboard with ratings of police departments based on the amount of data they make public, and on the quality and timeliness of that data. Interestingly enough, in 2013 SpotCrime created and still maintains a crime data transparency ranking. We rank cities based on how transparent they are with crime (RMS/CAD) data. A 2 ranking means the crime data is open and available for all. A 1 ranking means the data is available somewhere, but it is not in an open format. A 0 ranking means the city is not transparent with crime data at all.

This recommendation also includes an action item to create model state criminal justice data transparency laws and foster adoption among states. SpotCrime has provided testimony on open data laws across the US. You can view our testimony on the Maryland Open Data Bill here. We’d like to note that any law regarding access to data needs some sort of ‘teeth’ measure to ensure lasting transparency. Although we testified on the MD Open Data bill that passed in 2014, a lot of police agencies in Maryland, most notably Baltimore County Police and Harford County Police, still do not make their crime data openly available.

AV Recommendation #3: Modernize the production and dissemination of criminal justice statistics. An action item being that the OJP should publish crime data more frequently and in a more timely manner, and should release all publications on a predictable schedule.

This responsibility should fall hard on the local level. Local police departments need to all be on a predictable and/or standardized schedule of releasing basic datasets like CAD/RMS (as well as arrest, use of force, officer involved shooting, etc) data. It should be made available in an easy to find place digitally and in a timely (daily or weekly) manner in machine readable format with an open license.

AV Recommendation #4: Improve the integrity of data used for decision-making, research, and policy.

An action item includes the OJP providing incentive funding to states that improve their criminal justice data quality via audits, technical assistance to local jurisdictions, advancing the use of standards, or the creation of data quality and transparency advisory boards.

We’ve found that some police agencies are using systems that are 15+ years old. The computational capabilities of the iPhone has increased dramatically since 2007. Why do we leave our government systems behind? Funding already incentivizes police agencies to submit UCR/NIBRS to the FBI, so it may work in this case. However, UCR/NIBRS is bulky, and slow. If there is no amount of oversight following data standards, there will be no progress. Additionally, the bigger fight might not hang on funding, rather a cultural shift in police agencies wanting to share this data with the public.

AV Recommendation #5: Make criminal justice data more actionable, by linking data for greater insight, and by building capacity to turn insight into action.

An action item suggested is that the OJP should fund state and local data and analytics capacity and develop a model and toolkit for broad replication. Showing other agencies what works and doesn’t work from experience is a great way to get other agencies on board. This has sort of already occurred with policing and mental health calls. Just this week, the feds announced funding for local mobile teams with mental health practitioners trained in de-escalating volatile situations. This is a replication of the Eugene (OR) CAHOOTS program in place for 30 years already, aimed at providing mental health experts to help police officers better respond to calls, saving the city not only thousands of dollars in the process but also lowering violent interactions with police. However, again, the adoption of any type of change or program would require a culture shift. Tying funding to this recommendation would help with adoption.

AV Recommendation #6: Improve the use of technology to equip criminal justice decision-makers with timely and accurate information. Specifically an action item being that the OJP should mandate that software or systems created with federal dollars be able to export machine-readable data and have a standard open API (application programming interface) for sharing data across systems.

This is probably one of the best recommendations. It is rare for police agencies to create their own internal databases. Typically, databases and technical work is contracted out to database vendors, who have no level of standard to follow when creating/maintaining databases, making it incredibly cumbersome when each police agency asks for something new to be added to the system, and incredibly expensive when police agencies attempt to gain access to their own data. This is the case with RMS/CAD data. Private vendors have been attempting to take over data in the public domain for years. SpotCrime has been quoted ridiculous amounts of money, including $24k in Oakland County to pull data that Oakland County is already surfacing in two areas (here and here). It is ridiculous that there is an attached cost to surface open data (that is already in existence) for public consumption. A federal mandate would require vendors to provide specific standardized reports for police agencies for free. It would hopefully pull together the current patchwork of vendor systems that don’t connect or want to talk to each other very well.

The report advocates for the OJP to take the lead on these recommendations. This is a good start, however we believe not only should the OJP embrace some of these recommendations, but state and local governments should lead the charge as well. The simplistic way to solve this problem of data is to require every police agency to have infrastructure to provide data of the previous day's activities (CAD and RMS) in an open spreadsheet. It is likely this infrastructure is already in place. This kind of data has been surfaced as a crime blotter for newspapers across the world for hundreds of years and this consistency of reporting and analyzing data is a hallmark of any professional organization. Start with the basic datasets, then move to other datasets like use of force, hate crimes, officer involved shooting, community engagement, etc.

Hopefully, something from this report will be adopted at the federal and local level within the next 5 years. Until then, SpotCrime will continue to advocate for open crime data. Don’t see crime mapped on SpotCrime in your city? It probably means your police department does not release data in an open format. Reach out to us and let us know!