Thursday, May 13, 2021

“Fixing” Criminal Justice Data and My Alternative Approach

 

Last month, the Arnold Ventures group released a report titled “We Need Criminal Justice Data That Doesn’t Exist.  Here’s How the Biden Administration Can FixIt.”

This is a subject near and dear to me as I have been working on much of this for a long time.  I agree with many parts of the report.  However, I was disappointed to see that the full report missed many standards and frameworks that have been previously built?  More below...



---

The main points of the paper as they summarized them are:

  • Recommendation #1: Establish an accurate baseline of facts about the criminal justice system, and envision a 21st-century system
  • Recommendation #2: Radically increase accountability of the justice system through data transparency
  • Recommendation #3: Modernize the production and dissemination of criminal justice statistics
  • Recommendation #4: Improve the integrity of data used for decision-making, research, and policy
  • Recommendation #5: Make criminal justice data more actionable, by linking data for greater insight, and by building capacity to turn insight into action
  • Recommendation #6: Harness modern technology to equip decision-makers with more timely and accurate information

But let us identify some of the things missing from the report.

First, SEARCH, the non-profit National Consortium for Justice Information and Statistics ( https://www.search.org/about-search/company-background/ ) has been working as they write “50 years of experience supporting the information sharing, interoperability, communications, information technology, high-tech crime investigative and criminal records systems needs of State, local and tribal justice and public safety agencies and practitioners nationwide.” 

Second, our friends at the IJIS Institute (https://www.ijis.org/page/AboutUs ) that “is a nonprofit alliance working to promote and enable technology in the public sector and expand the use of information to maximize safety, efficiency, and productivity” was founded in 2001 assist in the creation and collaboration of justice data.

Third, NIEM (https://www.niem.gov/about-niem ) – “NIEM’s origins lie in the Global Justice Information Sharing Initiative. This grassroots effort by a handful of organizations supporting state and local government set in motion the creation of a seamless, interoperable model for data exchange across government agencies. The pre-release of the Global Justice XML Data Model (GJXDM) was announced in April 2003.”

“Parallel to the GJXDM effort was the creation of the U.S. Department of Homeland Security (DHS) and the mention of metadata in the President’s strategy for the homeland security community to begin working towards standardization.”

Oasis-Open LegalXML Electronic Court Filing – ( https://bit.ly/3hp2De0 ) “The OASIS Electronic Court Filing TC will develop specifications for the use of XML to create legal documents and to transmit legal documents from an attorney, party or self-represented litigant to a court, from a court to an attorney, party or self-represented litigant or to another court, and from an attorney or other user to another attorney or other user of legal documents.”  We are up to version 5.0 for this standard that is widely used.

One standard that was mentioned was the FBI:UCR (uniform crime report) a program that includes the National Incident-Based Reporting System (NIBRS) https://www.fbi.gov/services/cjis/ucr/data-documentation  Please note this is an incident data collection approach that in most instances does not provide a complete picture in our experience.

And the United Nations has a similar coding system for criminal justice statistical capture that we have been working with in recent years. - https://www.unodc.org/unodc/en/data-and-analysis/statistics/iccs.html 

These omissions in The Arnold Ventures report make it difficult in my view to conclude that even basic justice data cannot be obtained.  But I can see that this is may also, be true from various viewpoints.

However, I think there are two additional structural barriers that also need to be considered:

First, the criminal justice system is made up of tens of thousands of government and even private entities.  Police, Sheriffs, Prosecutors, Public Defense, Courts, Probation, Social Service, Mental Health, and Substance Abuse programs that may be public or private.  Each has its own budgets, technology foundations, and technology.  And understand, their IT systems are acquired at different times using different technology focused on supporting each entity’s core missions.  Statistics are a byproduct.

Second, the data sought is in the case file documents.  Since the IT systems are developed for organizational and process support, the documents contain the data that the researchers need.  Therefore, I fear that the general approach advocated by the paper is doomed to failure as it will not provide the data they seek.

We see a lot of work being done in Legal Technology on eDiscovery.  My suggestion is to focus on getting access to the documents from the criminal justice entities, converting them into useful electronic format, and then exploring eDiscovery and text analytics.

 


2 comments:

  1. Two points to add to Jim's cogent remarks:
    1. Asking people to collect data they do not need for their work, let alone for someone apparently trying to evaluate them, is textbook invitation for incomplete, inconsistent, and misleading data. If you want data for a specific purpose, you need to go collect it.
    2. The primary role of the criminal justice system is not to provide transparency. Sure, transparency is a good thing, but not their main focus. Instead of focusing just on transparency, demonstrate how collecting more, and more useful, data will help them better manage their programs and make the world a better place. Go produce some 'quick wins' to demonstrate what useful data can do.

    I agree with Jim's alternative, if you want data, go figure out how to get it from the documents and systems that exist. Data extraction is a growing field, help it grow.

    ReplyDelete
  2. Thanks Alan. I completely agree with your additional points.

    ReplyDelete