Can Algorithms Predict Crime?

The origin of criminal behavior has been much debated over the years as some believe there to be a biological basis that makes people more prone to committing and others deem one’s environment to be responsible for causing crime (Gruman, et al., 2017).  Is there a certain gene that when expressed causes people to commit crime or is it due to a hormonal imbalance in the brain that pushes one towards a life of crime?  Alternatively, do people blame one’s upbringing, their friends and their families as the reason why someone commits a crime?  If society is to blame for crime, can we somehow analyze societal trends in order to predict crime?  The city of Bristol has attempted to answer this question by implementing an algorithm that tells how likely each citizen is in being a victim or a perpetrator of a crime (Booth, 2019).

Approximately 170,000 citizens of Bristol are listed in a database that assesses the possibility of each person being involved in a child abuse or kidnapping crime (Booth, 2019).  In addition, this algorithm determines how likely a person will behave and take part in antisocial behavior.  This may seem surprising; however, the algorithm utilizes information such as past criminal behavior, gang membership, domestic violence and school truancy in order to predict criminal activity.  The reason this new method has become so helpful to society is because it uncovers information about people quicker than many psychologists and therapists could have found if they met with each person.  Police and local government can pinpoint certain areas at high risk for people to commit crime and either increase police presence or bring in more counselors to talk to people in the area about their antisocial behavior (Booth, 2019).  In addition, if someone appears to be at high risk for criminal behavior and are already seeing a counselor, then the counselor can be notified and deal with the person appropriately.  So far, the algorithm has proved to be a success as they studied five people who were the victims of sexual exploitation.  Three of the five individuals were in the top 100 list of people in the area who were at high risk of becoming victims and the other two were also high up on the list of potential victims (Booth, 2019).  Although this algorithm seems promising for other towns to use, it does not mean that it is 100 % effective.  Even though 3 of the people were on the most at- risk list, it would be difficult to reach out to every person on the list and talk to them about avoiding situations where they can be exploited.  In addition, the algorithm may report a person to be at high risk for committing crime, but it may be wrong and could hurt people in the process.  Also, this does not mean that all other forms of protection including the police and the legal system are less important because humans are better able to understand a real- world situation and talk to at risk people rather than a computer algorithm that uses certain patterns to predict crime.  The algorithm utilized by Bristol is very promising as it can look at trends in behavior such as truancy and past antisocial behavior in order to prevent crimes as well as help those at risk receive the proper attention for them to get better.

This new technology uses the social psychological and the sociological theories of criminal behavior in order to find those in society who pose a risk to other people.  Sociology uses ideas such as anomie, socioeconomic status, neighborhoods and education and attempts to understand how they relate to criminal behavior (Gruman, et al., 2017).  Researchers have found that those who come from a lower socioeconomic status and who report feeling more alone and isolated in society have higher rates of criminal behavior.  In addition, social psychology theories such as the general personality and social psychological model (Gruman, et al., 2017) of criminal behavior implicate that those who have not completed school or have a low degree are risk factors that can increase the likelihood of criminal activity.  Also, those who have been associated in the past with people who have criminal records are also more likely to commit a crime.  These ideas paint an accurate depiction of proven trends that is found amongst those convicted of a crime.  If a town wants to make their community safer and protect the innocent then using an algorithm that finds those who fit the typical characteristics of a criminal can allow police and counselors to find them and rehabilitate them so they can function normally in society.

-Madison Laezzo

Resources:

Booth, R. (2019). How Bristol assesses citizens’ risk of harm – using an algorithm. Retrieved from https://www.theguardian.com/uk-news/2019/oct/15/bristol-algorithm-assess-citizens-risk-harm-guide-frontline-staff.

Gruman, J. A., Schneider, F. W., & Coutts, L. (2017). Applied social psychology: understanding and addressing social and practical problems. SAGE.

2 comments

  1. Madison,

    This is so fascinating! While I’m a bit taken aback that an algorithm is being used to identify high-risk individuals (for both criminal behavior and victimization), it’s not necessarily a surprising step considering the wealth of research that exists on the topic. Although a central compilation of data that highlights certain risk factors is extremely useful, the risk lies in taking it as fact, and labeling the individuals that the algorithms have deemed high risk. I would also consider utilizing such an algorithm as a means for providing interventions in high-risk neighborhoods, providing increased police presence, allocating money to provide more prosocial opportunities, increasing education opportunities, and creating an atmosphere that help make antisocial behavior a less likely outcome. I believe that this is the difference between being reactive and proactive.

    Thank you for sharing this, I can’t wait to look into it further.

    Jess

  2. I have never heard of a method like this to analyze crime. I found this article to be very interesting. One of the benefits to this method is that this system can paint a picture of risk for an individual in seconds, something that would take social workers years of knowledge to build (Booth, 2019). This is helpful because social workers receive automatic emails when new data arrives that creates a concerning pattern. The email urges the support team to review their plans. This seems like successful tool to be used by communities that want to make their towns safer and protect the innocent, as you mentioned.

    Reference:
    Booth, R. (2019). How Bristol assesses citizens’ risk of harm – using an algorithm. Retrieved from https://www.theguardian.com/uk-news/2019/oct/15/bristol-algorithm-assess-citizens-risk-harm-guide-frontline-staff.

Leave a Reply


Skip to toolbar