International Journal on Criminology Volume 4, Number 2, Winter 2016 | Page 81
International Journal on Criminology
But let us get back to the heart of the matter. Can predictive policing really be
taken seriously?
First off, let us try to solve the earthquake prediction affair. Here is what
geophysicist Bill Ellsworth, a researcher at the US Geological Survey, had to say
in a recent interview: “No one knows how to predict earthquakes … If earthquakes
are predictable, we do not yet know how to do it. And there is a good chance that
they are not predictable … We do not even really know the equations that govern the
way that earthquakes work…” As the interviewer then observes: “A scientist like Bill
Ellsworth, even with all the data at his fingertips and all the computer modeling money
can buy (our italics), admits that predicting earthquakes to some degree … escapes
him.”
In similar fashion, the criminal milieu clearly forms part of the real world,
where complex human interactions cannot always be understood even by the most
sophisticated models, and where statistical analysis tools tend to produce meaningless
results. In addition, we cannot rely on laboratory experiments for a causal analysis.
Man, Machines, and Crime
Let us start with an observation, one that is commonplace for any criminologist
but that eludes IT professionals and big data devotees. Human activity is not a natural
resource that can simply be extracted from the cloud and monetized. It is not like coal
or oil that can be mined at will. Because—and this has been the reality for a thousand
years!—human-beings, regardless of whether they are criminals or not, never let
themselves be watched without reacting. Our reptile-like brains and our genes are
stamped with the imprint of millions of years spent avoiding predators in the fight for
survival. And these predators were armed with lethal claws when we, mankind, had
no such threatening appendages: no horns, no hooves, no armor. Our sole weapon was
our huge brain: we adapted, we hid, we used trickery, we cheated. In short, mankind
reacts.
Mankind is constantly trying to resist being observed and watched by
computers. We are active. So, exploiting big human data is not some mundane mining
activity like extracting coal or oil; it is a game of chess or, if you like, a boxing match.
When it comes to illegal activities, here is an eloquent example of this phenomenon.
How do drivers react in France when the authorities increase the number of radars
lining the roads and motorways? They adapt by cheating. The practice of cloning
vehicle registration plates (which get caught by the radars) has gone through the roof:
in 2010 this type of violation jumped by 98%, and in 2011 it was up by 73%. From
2010 to 2012 the number of clonings rocketed from 5,079 to 17,479.
80