Crimes of future past

Crimes of future past

Recently the chief of the Dutch police made an announcement that the Dutch police will be investing in more technology in order to predict crime. But is technology able to do that?

This week the Dutch police made an announcement stating that they will implement new technologies to predict crime. The Dutch police have made similar statements in the past and have actually developed new technologies with the intention of combatting crime more efficiently and effectively. The development towards a more technology-assisted police force is not just visible in the Netherlands. Police organisations all over the world are moving towards so-called intelligence-led policing. As the name of the concept implies, intelligence-led policing aims to use data and data analysis to improve police performance. Data on crimes committed in the past and information about neighbourhoods and such would be able to point out where and when crimes will be committed in the future. This logic was also used by the Dutch chief of police: “the force has so much data available, but the systems are not properly connected. We need people who can do that and who can analyze the data. Who can, as it were, predict when and in what form crime will occur.”

Pre-crime and promises

The ability to predict where and when crimes will take place is of course the dream of every police organisation. It would enable them to prevent crime from happening, resulting in a much safer society. Although predicting crime may seem like a far-fetched idea to many, the idea of pre-crime is not all that odd to police organisations. Often references are made to the book or film Minority Report, where a mix of technology and human oracles are able to see the future and therefore where crimes will be committed and by whom. Although the current state of the art does not include the use of oracles (as far as I know), the new technologies do make similar promises. Hitachi developed a system that according to them will be able to predict crime and so has Microsoft. The Fresno police force even developed a system that can assign threat scores to citizens.

Although these technologies make great promises about predicting and suppressing crime, they do come with downsides. Technology is not without its faults, after all, the most obvious one being wrong “predictions.” While these technologies are presented as crime predictors, this is most likely an overstatement of their capabilities. They are not actually able to predict the future, they are only able to give an indication of probabilities. This small but important difference is lost in the language used to describe these technologies. This can at best lead to unrealistic expectations or at worst label innocent individuals as criminals. Another factor that is often left out of the discussion is the acceptance of new technologies by police officers. Research has shown that changing police practice and culture is not an easy task. Both street-level and management-level officers do not always like to change the way they work. Technologies that partly take away officers’ freedom to make decisions and replace their experience with digital decision-making can encounter resistance during implementation, which can greatly diminish the impact of technology in practice.

Faith in technology

Regardless of all the potential downsides and difficulties, policy-makers and politicians still have a lot of faith in technology (see also my previous blog on this). In their view, the new digital technologies would be able to solve problems that cannot be fixed by humans. This phenomenon is sometimes called techno fix: the trust placed in technology to fix complex societal problems. With the ‘crime prediction technologies’ often presented as all-seeing oracles by the developers, it can be hard to blame policy-makers for these views. However, with all the statements being made about the capabilities, the amount of government funding used for development and the general increase in the use of technologies in law enforcement, it would be nice to see some empirical evidence concerning their prowess and practical benefits instead of theoretical claims. This, however, seems to be lacking for the moment.