By the numbers
As I watch the ongoing public controversy about street-checks (aka “carding”) run-on like Niagara Falls, I’m often stuck by the statistical data typically quoted by the mainstream media and other critics, who claim that the practice is inherently biased against certain identifiable groups.
By Tom Rataj
On the face of it these stats may look pretty disturbing and convincing to many people, but as with so many other things, looking a little closer and with a more knowledgeable eye, a much different picture begins to emerge. While I’m no statistician, I understand the truth behind the famous line comparing stats to bikinis – as important for what they show as for what they hide.
When the street-checks controversy was spawned by certain mainstream media outlets, it appears that most police services were completely surprised and flummoxed by how to deal with the resulting storm. Other than a few meek protestations and generalities about they’re important, most services just seemed to stick their heads in the proverbial sand, hoping that this too shall pass. Which it clearly hasn’t.
Critics demanded that police services prove the value of street-checks, by providing examples and data about their value in solving individual crimes. Providing that proof unfortunately is not so simple because one isolated street-check is often not the only key to solving one individual crime or a series of crimes. This simplistic view also fails to recognise all the other inherent value that the street-checks process offers, such as crime-prevention and deterrence.
In speaking with some of my personal acquaintances and neighbours, all from outside the policing community, I quickly learned that many of them had been convinced by the media coverage that street-checks were discriminatory, unnecessary and an infringement of civil liberties. They asked me questions about them and it became readily apparent that they all had a fundamental misunderstanding of the process, purpose and value of street checks.
The usual argument trotted-out by the media goes something like this: Gotham has X number of residents in total, and only X percent are a certain identifiable group, yet the members of that identifiable group represent a much higher percentage of the residents being street-checked by the police, therefore the police are acting in a discriminatory fashion against that group … and because of that all street-checks should be banned.
But hang-on just a second, what do those stats really mean, and do they actually have any value at all, all by themselves, in such a broad “this vs that” comparison? Without any context I really wonder how much value these raw stats have.
I wonder whether analysing the numbers along with a variety of contextual data would reveal a much different and more accurate and valuable picture of the situation. I would think that geographic and demographic information, crime types, rates and locations, date and time and other factors often plumbed by crime analysts would help provide that picture. Data about all the residents of Gotham would also need to be included, to provide an accurate understanding of what is happening.
In addition to the questionable stats, I suspect that part of the problem with street-checks, as with some other police activities, is that they have been compromised by the pursuit of numbers, driven by a “we want more” mantra from various levels of management with different motivations and agendas. Because of that quantity instead of quality pressure, officers go out and about, collecting numbers to appease the bosses.
The good looking numbers are often carefully collated and presented in quarterly, annually, year-over-year, and 5 and 10 year trends and other questionably useful standards, convincingly presented in colourful and splashy exploded pie-charts and 3-D bar- charts in meetings at various levels and of course the annual report.
While it’s important to measure numbers for many useful management purposes, there’s a huge risk in allowing that measurement and the numbers alone to become the sole focus and the driving force behind the work. The relentless push for more and bigger numbers can quickly erode the quality of those numbers and create a bunch of downstream problems such as what I suspect has happened with street-checks: a numbers-focussed exercise that morphed an important crime prevention, deterrence and intelligence gathering tool into a questionably useful statistics gathering exercise that’s come back to punish us.
I also wonder whether a bit more candor and a major reengineering right at the start of the controversy would have helped mitigate it and prevent the restrictive legislation that is surely to negatively impact community safety in the coming years.