Predictive Algorithms
What was the original purpose of the use of predictive algorithms?
In today’s world, everything from bank transactions to the education sector and even in politics and elections, everything is run on big data. For this reason, automated-decision making systems have gained a lot of favor in influencing decisions. This has led to the introduction of solutionism, where tech companies come up with solutions for all social problems, including crime (Morozov, 2013). With the number of individuals held within the American Judicial System on the rise annually, there was a need to come up with a solution to high rates of incarcerations. This created the niche for predictive algorithms. With over thirty years of applied statistical methods in the US Judicial system in probation and parole, the predictive algorithms seemed the best approach to making the system more efficient. Predictive algorithms use variables such as the defendant’s criminal history and their socio-demographic features to predict their behavior while on probation().
What are the consequences (potential or existing) of this practice? (What moral or constitutional issues are present?)
The application of predictive algorithms in the US Judicial system has raised significant concern, especially concerning fairness, taking into consideration its use in sentencing. Following concerns about disparities in sentencing, former Attorney General Eric H. Holder Jr advocated for similarity in sentencing between offenders with similar criminal backgrounds and related offenses (Eubanks, 2018). However, the application of these algorithms in sentencing threatens to increase the disparities. Additionally, the use of gender and socio-economic characteristics of an individual in sentencing is considered unconstitutional. According to Christin, Rosenblat & Boyd (2015), the Supreme Court disqualified the use of statistical tendencies as a proxy for characterizing an individual in criminal justice and sentencing. Furthermore, though none of the predictive algorithms use race as a variable, other variables used such as zip codes may act as proxies for race. This makes it possible for the use of these algorithms to target specific communities.
Do you believe predictive algorithms are still viable tools for criminal justice practitioners?
Though it is argued that the algorithms are used to provide mere indicative predictions of the defendant, significant research indicates that judges and prosecutors are more likely to follow the projections rather than make independent decisions (Noble, SU (2018). Further research suggests that some of the judges even change their decisions to match the predictions given by the algorithms. With such tendencies, the number of incarcerations is bound to increase, rather than reduce as intended. In addition to that, these algorithms justify the punishment based on four reasons; rehabilitation, deterrence, retribution, and incapacitation (Christin, Rosenblat & Boyd, 2015). Choosing one of these and overruling the others is bound to create inconsistencies in the Judicial system. Therefore, though designed to reduce the rate of incarceration, the use of predictive algorithms may end up the opposite.
References
Christin, A. Rosenblat, A. & Boyd, D. (2015). Courts and Predictive Algorithms. Data & Civil Rights: A New Era Of Policing And Justice. datacivilrights.org
Eubanks, V (2018) Automating Inequality: How High-Tech Tools Profile, Police, and Punish the Poor. New York: St Martin’s Press.
Morozov, E (2013) To Save Everything, Click Here: Technology, Solutionism, and the Urge to Fix Problems that Don’t Exist. London: Allen Lane.
Noble, SU (2018) Algorithms of Oppression: How Search Engines Reinforce Racism, 1st edn. New York: NYU Press.