Does open crime data provide some indication to the quality of policing?

The recent controversial encounters with the police and citizens in Ferguson, New York, Cleveland, and now North Charleston raise the question - does the way in which a police agency makes their crime data available determine the the quality of policing? 

With the exception of Ferguson PD thanks to recent DOJ review, we can’t make claims about the quality of policing for North Charleston, Cleveland, or NYPD. 

However, we do have enough knowledge to assess the quality of each agency’s stance on open crime data and make some inferences on how this affects public trust. 

Based on the correspondence we’ve received from police agencies over the years, we think certain agencies may be telegraphing their commitment to community policing by how they release and respond to requests for public data. 

In Ferguson, SpotCrime was getting open data access for a period of years, however the data stopped two years prior to the shooting of Michael Brown. We do not have a full explanation of why the data stopped, but we understand that the data stopped being delivered to the St. Louis County Police Department database. Ferguson stopped reporting data openly a few years before Michael Brown was shot. 

For the NYPD, SpotCrime has been requesting public crime data since our inception in 2007. It is always difficult to get a response from an often Byzantine designed organization as the NYPD. The NYPD is such a large organization, making a FOIA request to the NYPD is a Kafkaesque experience.

At one point we were asked to provide a video cassette to prove we are news media organization. This was around 2009, the same year that NYPD spent $1 million on typewriters. Finally in 2014, the NYPD started releasing data to a public mapping system. The data is deliberately difficult to access and has been purposely neutered - data is months old and does not identify the specific date of the incident. If there were multiple crimes in the same location, it is impossible to know if they happened on the same day or weeks apart. 

All other US agencies that make crime data available provide the date and often exact time of the incident. So, it is easy for us to make a statement that the quality of data falls short of the expectation of open data. We've never been able to get a full explanation from New York as to why they don't provide a specific date. As far as we know, the NYPD is the only agency in the US to identify crime this way. This, among other reasons, is why New York gets a ranking of 1 on our transparency ranking. 

For North Charleston, SpotCrime reached out and asked for data in 2010 and 2012. In both cases, our request went unanswered. North Charleston currently employs a proprietary public facing crime mapping system called Raidsonline. The first thing to notice when you visit this vendor website is that you are immediately asked to give up some rights before you look at the crime data via a pop up Terms of Use. The system specifically restricts the public and the press from using the data. Additionally, the website is designed to prevent any data from being copied down using a copy and paste function. It is not machine readable. For these reasons, the data is not open and does not fall under the definition of open data. 

Often we hear that police departments prefer closed proprietary systems in order to prevent the public data from going out in the wild. 

An exclusive single source proprietary system like Raidsonline is a useful tool to control the data and limit independent analysis allowing police agencies to often pull or change data without anyone noticing. In our view, using this type of system minimizes the reach of data because the audience is limited and the ability to share information is deliberately stunted. Agencies that use a proprietary system claim they are being open with the data in tandem with allowing a vendor to threaten anyone with legal action who copies down the data. 

This counterfeit transparency inherently contradicts the purpose of open data and the intent to create public trust. 

Of the four agencies, Cleveland police provides the most robust crime data feed to SpotCrime. Unfortunately, the data is not open to everyone and there is a fee. Similar to North Charleston, Cleveland had elected to use only a proprietary system, but after a year of letter writing, the department agreed to provide a feed. The catch for the public is that the crime mapping vendor gets paid, gets to sell the data to industry essentially extracting double value from their control over the data, and the local press, national press, and the public are charged a fee for the same data given to the vendor. 

These four cases are all anecdotal. However, we can say definitively that each department does a poor job of making their crime data open and publicly available. 

We understand the fears of openly providing crime data. There's a potential for the data to be misinterpreted and/or exaggerated. This is the inherent problem of wanting public trust - you have to trust the public, and accept the risks of the data being used improperly. Based on historical evidence, the occurrence of misinterpreting the data is infrequent. And our hope is that any time crime data is exaggerated, the claims can easily be debunked by the data set made fully open. 

Open data will not solve all our problems. It is no panacea, but it's a great first step for law enforcement. It's an easier first step than body cams and by far much less expensive.* 

Our Suggested Solution: 

Based on our experience, we estimate that over 80 percent of police departments in the US could begin making data fully public within the year. This is a much faster penetration rate than a body cam program could produce. 

As a proponent of open crime data and open data in general, SpotCrime encourages every police department to make data available in an open format, not in a proprietary format. 

If your police department introduces a fancy new public facing crime map, as a citizen or member of the press - please ask for equal and fair access to the data. And ask the police department to explain the restrictions on the vendor website. What do they mean? Why are they there? What is the public benefit of the restrictions? Can the vendor profit from the data while restricting access to the public?

Colin Drane 

Founder 

SpotCrime.com 

*SpotCrime is not against body cams for police officers, but we understand the long term complexity and costs of maintaining the equipment, storing the data, and training officers. If an agency can’t manage their crime data and make it public, how will they effectively manage their digital video data.