The Allure of Big Data Policing

As I mentioned in my initial post, the goal of my book project on “The Rise of Big Data Policing” is to examine how technology is changing the “who,” “where,” “when,” and “how” we police – especially in large urban cities.  As I write in the introduction:

"New technologies threaten to impact all aspects of policing, and studying the resulting distortions provides a framework to evaluate all future surveillance technologies. A race is on to transform policing. New developments in consumer data collection have merged with law enforcement’s desire to embrace “smart policing” principles in an effort to increase efficiency amid decreasing budgets. Data-driven technology offers a double win—do more with less resources, and do so in a seemingly objective and neutral manner."

In the book, I make the argument that in addition to the strong lure of new technology and cost efficiencies, there is an openness to new “data-driven technologies” as a result of the recent upheaval arising from a heightened awareness about police violence in America. 

Over the last few years – and again last week – the death of African Americans at the hands of police officers has generated protest, anger, and dissent. In addition, policing systems like the NYPD’s stop and frisk program created fear, resentment, and frustration about how citizens should be treated by law enforcement.  My argument is that out of this tragedy and frustration, the idea of policing strategies being guided by data-driven objectivity has become quite alluring.  Replacing human discretion with algorithmic precision – at least in theory – seems like a move toward progress.  Following data and not hunches seems (again in theory) like an improvement.  More after the break.

Again, from the introduction:

"This book arises out of the intersection of two cultural shifts in policing. First, predictive analytics, social network theory, and data-mining technology have all developed to a point of sophistication such that big data policing is no longer a futuristic idea. Although police have long collected information about suspects, now this data can be stored in usable and sharable databases, allowing for greater surveillance potential. Whereas in an earlier era a police officer might see a suspicious man on the street and have no context about his past or future danger, soon digitized facial-recognition technologies will identify him, crime data will detail his criminal history, algorithms will rate his risk level, and a host of citywide surveillance images will provide context in the form of video surveillance for his actions over the past few hours. Big data will illuminate the darkness of suspicion. But it also will expand the lens of who can be watched.

The second cultural shift in policing involves the need to respond to outrage arising from police killings of unarmed African Americans in Ferguson, Missouri; Staten Island, New York; Baltimore, Maryland; Cleveland, Ohio; Charleston, South Carolina; Baton Rouge, Louisiana; Falcon Heights, Minnesota; and other cities. This sustained national protest against police—and the birth of the Movement for Black Lives—brought to the surface decades of frustration about racially discriminatory law enforcement practices. Cities exploded in rage over unaccountable police actions. In response, data-driven policing began to be sold as one answer to racially discriminatory policing, offering a seemingly race-neutral, “objective” justification for police targeting of poor communities. Despite the charge that police data remains tainted by systemic bias, police administrators can justify continued aggressive police practices using data-driven metrics. Predictive policing systems offer a way seemingly to turn the page on past abuses, while still legitimizing existing practices.

For that reason, my aim in this book is to look at the dangers of black data arising at this moment in history. Only by understanding why the current big data policing systems were created and how traditional policing practices fit within those systems can society evaluate the promise of this new approach to data-driven law enforcement."

As recent evidence of this pattern, just this month the City of Chicago – home to the birthplace of person-based predictive policing, a Department of Justice Civil Rights investigation into the Chicago Police Department that found widespread racial discrimination, and a horrific murder rate – decided to go all in on big data policing.  The City of Chicago has decided to combine a place-based predictive policing technology, a person-based predictive policing technology, and a real-time “Strategic Decision Support Center” to target the people and places driving violence.  Not surprisingly, the City, like others before it, got good press for its high-tech answer to the increasing violence. 

And, to me this one of the secrets of why big data policing is so alluring: it offers “an answer.”  Chicago needs an answer about how they are going to stop the shootings.  As of April 26, 2017, there were almost 1000 shooting victims in Chicago this year.  Over 4000 people were shot last year.  Politicians, chiefs of police, any sane person needs some answer about how they are going to stop the killing.  It doesn’t have to be a good answer.  It doesn’t have to work. But, you need to have some response.  Big data policing and all the fancy technology provides an adequate (and potentially satisfying) response.

It is also “an answer” that seems to be removed from the hot button tensions of race and the racial tension arising from all too human policing techniques.  Having more information, smarter information, more real-time information appears neutral and fair, and a lot better than just sending in more officers (which might increase tension).  This true even though the data comes from these real police officers and impacts the daily decisions of these real human beings.

To me, the big reason why we will continue to see the adoption of new data-driven technologies, even in advance of any data-driven studies to show the technologies work, is because of this need for an answer.  Ask yourself, if you were a mayor or chief of police wouldn’t you want to be able to point to something – like a new predictive program or system – to answer that age-old question of “What are you going to do to end crime?”  A black-box futuristic answer is a lot easier than trying to address generations of economic and social neglect, gang violence, and a large-scale underfunding of the educational system.  It is also a lot easier than more intrusive policing methods that require more officers on the streets. 

So, Chicago and other cities are going keep finding “an answer” whether or not the big data policing solution actually improves things in the long term. 

In my last post of this series tomorrow, I will address whether predictive policing requires a “policing solution.”  Many of the current risk assessment and predictive techniques are good at identifying risk.  The question remains whether police need to be the institution that remedies that risk.