Moneyball in criminal justice
Predicting crime; is it science fiction or reality? And what about our privacy?
I have been watching a series called “The Wire” and it made me think about how safe I actually feel while walking the streets in the Netherlands. It impacted the way I look at the city of Rotterdam even if you compare it with other places such as Baltimore, Maryland where the series is about. I decided to start doing some investigation on the topic and decided to write a blog about it.
15 years ago predicting crime seemed like an unrealistic, very abstract concept. Nowadays fighting crime by smart statistics and rigorous analysis is in the lift. More and more police departments across the world are trialing or implementing the techniques. Predictions go as far as determining where and when the most likely crime will be happening and estimating to what degree there is possibility of relapse with a certain offender (Milgram, 2013)(Accenture, 2012). A common mistake people make is confusing predictive analytics with Minority Reporting. However, Minority Reporting is about the question who while predictive analytics seeks to find the answer to the questions when and where.
In a TED talk by Anne Milgram that I watched, the previous attorney general of the state of New Jersey, she explained how she came to change the way she saw criminal justice. One of her main realizations was the fact that there was no knowledge about who were in the criminal justice systems, no data availability, no data sharing between departments, and no expertise on what was important to look at. In addition, she explains how decisions-making were made based on instinct and experience. Her conclusion was, looking at the entire justice system under her jurisdiction, that they were not doing a good job. This is what triggered her to implement rigorous data analyses and smart statistics to fight crime. The result was that murders reduced with 41% and crimes decreased with 26%.
Nevertheless, numerous questions could be raised. For example: How good are the data used? Is crime data vulnerable to manipulation and error? How do we know the underlying algorithm are not erroneous? Is our right to privacy protected? (Huffington Post, 2015)
To me all these questions and concerns raised are nullified when you look at examples provided by people like Anne Milgram. These numbers tell an important story. In my eyes these numbers matter greatly; public safety is of utmost importance since without it we cannot be educated, we cannot be healthy, we cannot exercise and we cannot be happy.
However, fact is that with big data, privacy issues remain on the surface. Personally, I believe privacy issues are less important than our safety. Unfortunately, there are struggles that come with ‘small data’ being replaced by ‘big data’.
Ask yourself a question: What if the “reasonable suspicion” doctrine of seizing and interrogating a suspect is being challenged by an interconnected web of data and knowledge combined (Ferguson, 2014)?
I believe you cannot fight crime with little yellow post-it notes. What about you?
Accenture, (2015). London Police Force Uses Analytics to Fight Gang Crime – Accenture. [online] Available at: https://www.accenture.com/id-en/success-london-police-force-using-analytics-fight-gang-crime.aspx [Accessed 7 Oct. 2015].
Berg, N. (2014). Predicting crime, LAPD-style. [online] the Guardian. Available at: http://www.theguardian.com/cities/2014/jun/25/predicting-crime-lapd-los-angeles-police-data-analysis-algorithm-minority-report [Accessed 7 Oct. 2015].
ComputerWeekly, (2015). Met Police trials analytics to fight gang crime. [online] Available at: http://www.computerweekly.com/news/2240233449/Met-Police-trial-analytics-to-fight-gang-crime [Accessed 7 Oct. 2015].
Datafloq, (2015). Datafloq – The One-Stop Shop for Big Data. [online] Available at: https://datafloq.com/read/los-angeles-police-department-predicts-fights-crim/279 [Accessed 7 Oct. 2015].
Ferguson, A. (2014). Big Data and Predictive Reasonable Suspicion. SSRN Electronic Journal.
The Huffington Post, (2015). Predicting Predictive Policing in NYC. [online] Available at: http://www.huffingtonpost.com/andrew-guthrie-ferguson/predicting-predictive-pol_b_7757200.html [Accessed 7 Oct. 2015].