This week, SIIA published an issue brief assessing the use of data analytics in the criminal justice system. Not surprisingly, data analytics has helped to reduce crime and improve the criminal justice system, particularly through its application in predictive policing and criminal risk assessment. The report also explores critical questions and concerns raised about the effectiveness and unintended outcomes of the various tools in use today.
This report is timely, as it coincides with a critical decision handed down by the Wisconsin Supreme Court about the use of evidence-based risk assessment tools at sentencing. The Court supported the use of predictive tools, such as the COMPAS tool at the center of the trial, but it ruled that “risk scores may not be considered as the determinative factor in deciding whether the offender can be supervised safely and effectively in the community.” [emphasis added]
Essentially, the court’s decision confirmed the value of data analytics, and it reached a conclusion similar to that in our report: the potential benefits are great enough to continue utilizing data analytics in criminal justice, but law enforcement and criminal justice officials must be responsible in their deployment of these technologies.
Importantly, the Court also made a very insightful observation about the continuing evolution and application of these technologies:
“It is important to consider that tools such as COMPAS continue to change and evolve. The concerns we address today may very well be alleviated in the future. It is incumbent upon the criminal justice system to recognize that in the coming months and years, additional research data will become available. Different and better tools may be developed. As data changes, our use of evidence-based tools will have to change as well. The justice system must keep up with the research and continuously assess the use of these tools.”
The Court’s conclusion wisely encapsulates the perspective we must apply broadly to the use of data analytics, that we can’t look at analytics from a singular perspective as either pure or ineffective. Rather, like all technology, analytical capabilities are applied across diverse situations and evolving rapidly. In many cases, there are not only tremendous opportunities for positive outcomes, but also a very real a risk of negative, unintended outcomes if not applied carefully. Thanks to this sensible Court decision, data analytics is likely to continue playing a key role in the improvement of the criminal justice system, but where continued scrutiny is necessary to review and evaluate these risk assessment tools for their effectiveness, efficiency and to avoid unintended disparate impacts.
More broadly, some have suggested that there is an overall accountability problem with respect to the use of data and analytics. This line of thinking asserts that we need more “due process” regarding the use of algorithms where citizens have the right to know more about scoring for employment, housing and education, and to be able to correct faulty conclusions.
Fortunately, the Fair Credit Reporting Act (FCRA) already provides a comprehensive set of rights and responsibilities for the use of data in wide variety of contexts. The FCRA applies not only to measures of credit worthiness, but also more broadly to eligibility determinations such as employment, housing and government benefits—these are all areas where Congress concluded additional protections were necessary to offset the risk of harm that could adversely affect the life chances of people.
Since enactment in 1970, the FCRA has provided a robust accountability framework, consisting of a series of rights and obligations to protect consumers, including requirements for disclosure and notice of adverse action, consent for sharing information with employers, access and correction requirements, and use limitations. This framework has also proven quite resilient over time, adapting effectively to significant changes in technology over the last decade.
Given this recent Court decision and the robust protection provided about the FCRA, it doesn’t appear that we have an accountability problem. Rather, we have a strong accountability framework in place that promises to help guide the effective and fair application analytical capabilities for years to come.