The value of data analysis

Analytics dominate nearly every field of endeavor in this country except for one: the criminal justice system. Ours is one of the biggest systems in the world and yet it still lacks the means for adequate analysis.

As prosecutors in Wisconsin who often espouse competing political ideologies, this is one thing we agree on: the need to bring improved data collection and objective performance measurement to the world of criminal justice.

This belief grows out of firsthand experience. Trying to do our jobs without data is needlessly difficult at best, catastrophic at worst. Often we can’t answer basic questions about the demographics of who’s getting arrested, how many people are in jail because they are poor or how many are being jailed for low-level offenses, much less how these numbers break down by different races or income status.

So why hasn’t more been done? For one, numbers can be dangerous and misleading. When data are collected — if they are collected at all — they are generally recorded in different ways from county to county and state to state. Some data are missing; some are named differently. These practices stem from the idea that none of this information was ever meant to be researched or studied. So we acknowledge that collecting data and making them useful to us is a challenge.

But given how far data mining technology has come, given how successfully the education and health care systems have managed to use administrative data to assess their own performance, there’s little reason to deny the criminal justice system access to information that is critical to advancing the debates over policing and justice and incarceration that have taken on ever greater force in recent years. What’s more, data collection and analyses are a great way to root out poorly collected data and make the case for standardized data collection strategies.

Here’s the bottom line: our job as prosecutors is to improve public safety. So it is incumbent upon us to welcome legitimate opportunity to stand back and assess how well the criminal justice system is operating. Especially when data and analytics have already proven themselves useful.

In Milwaukee County, where John Chisholm lives, analytics provided by the Vera Institute have brought to light racial disparities in the way we’ve handled black and white defendants arrested for low-level drug offenses. The result has been initiatives designed to promote racial equality and reduce jail populations, both of which have been successful.

In Winnebago County, where Christian Gossett works, analytics provided by a new nonprofit, Measures for Justice, have exposed a racial disparity between the number of black and white defen- dants who are diverted from traditional prosecution. Does the discrepancy owe to racial bias? Do black defendants refuse deferral options because they mistrust the system? Does the particular location of Winnebago in the state play a role? Because we now know this disparity exists, we’re able to begin to identify its cause and eventually to rectify it. Data cannot answer all our questions, but they can help us to start asking them.

As elected officials, we can say with confidence that communities on both sides of the aisle not only support a more efficient and fair system but are asking for the kind of rigorous and comprehensive data collection and analyses that can make intelligent changes possible. This is why we are particularly proud that Wisconsin has taken the lead in helping to make data sets about our criminal justice system public by supporting organizations such as Measures for Justice. Via their data portal, Wisconsin data are now both easy to understand and encyclopedic, spanning the system from arrest to postconviction.

The good news is that our model can be replicated nationwide. There’s just no reason why county practitioners from every state can’t be looking at their data and making better decisions as a result.