Can Big Data Help Dispense Justice?
A debate within the American judicial system is focusing on the growing use of data-driven predictions about future crime risks in shaping sentencing guidelines. According to reports this summer, at least 20 states have adopted so-called “evidence-based sentencing.”
Outgoing U.S. Attorney General Eric Holder raised concerns about the trend during the summer, calling on the U.S. Sentencing Commission, which establishes sentencing guidelines for the federal courts, to “study the use of data-driven analysis in front-end sentencing,” then issue policy recommendations.
As big data gains a foothold in the U.S. legal system, one concern is that analytics algorithms used to measure risk factors when determining prison terms can also reinforce biases in sentencing minorities and the indigent. Some argue that the big data approach to “smarter sentencing” may even be unconstitutional.
“It contravenes the principle that punishment should depend on what a defendant did, not on who he is or how much money he has,” argues University of Michigan law professor Sonja Starr.
Rather than applying data analytics to sentencing, Holder has instead argued that big data could be used to reduce sentences as one way to relieve prison overcrowding while achieving a measure of justice.
According to the Population Reference Bureau, the U.S. incarceration rate is the highest in the world at 500 prisoners per 100,000 residents. Other surveys put the rate at more than 750 per 100,000. Black males make up a disproportionate number of U.S. inmates.
“Data can also help design paths for federal inmates to lower these risk assessments, and earn their way towards a reduced sentence, based on participation in programs that research shows can dramatically improve the odds of successful reentry. Such evidence-based strategies show promise in allowing us to more effectively reduce recidivism,” Holder told criminal defense lawyers this summer.
Proponents of data-driven sentencing argue that in cases involving sex offenders, for example, data analysis was far more accurate in predicting future behavior, and hence the length of a sentence, than psychiatric opinions or parole board decisions.
Proponents partly agree with Holder that data analytics could help parole boards determine eligibility for reduced sentences as one way of relieving prison overcrowding.
That problem is growing in severity. In California, for example, a federal court gave the state two years to reduce the population of its overcrowded prisons. The order actually represents a reprieve since the original deadline to reduce the state’s inmate population to 37.5 percent above the prison system’s designed capacity was June 2013.
Hence, California and other states are looking for any tool they can find to develop new sentencing guidelines that do not exacerbate the prison-overcrowding crisis.
Meantime, it remains to be seen whether big data can help deliver a measure of justice when it comes to sentencing.