In Washington, D.C. in 2054, a Department of PreCrime determines who is going to commit a crime before it happens. The government uses three mutants, known as "precogs," who have precise visions of future events. Police are sent in advance to arrest the not-quite-criminals and, voila, the crime rate drops to zero. That is the backdrop of the movie, "Minority Report," based on a story from the late Orange County sci-fi writer Philip Dick.
Dystopian stories take real-life trends and extrapolate them far into the future, as a way to explore the moral conundrums of current policies. Flash forward 17 years from the movie's release (or back 35 years from the future!), and we find the Los Angeles Police Department wanting to impose its own version of what is known as "predictive policing." Instead of mutants, LAPD uses computers and human analysts.
The department pinpointed high-crime LASER zones—Los Angeles Strategic Extraction and Restoration—and tried to determine where to deploy a greater police presence. That sounded OK, but the computer system also created a profile of actual people who might have a propensity to commit crimes based on data about gang membership and arrest records. That's startling.
The inspector general found that "44 percent of chronic offenders had either zero or one arrest for violent crimes" and "about half had no arrest for gun-related crimes," according to a Los Angeles Times report, which noted that LAPD ultimately suspended the tool. Apparently, these technologies work better in the movies.
It reminds me of the state attorney general's APPS program (Armed Prohibited Persons System), which sends agents to the homes of people who are no longer are deemed eligible to own firearms. It sounds like a great way to remove guns from "bad guys," until one realizes that the complex computer-generated lists are woefully inaccurate, according to some reports. Our government cannot get its current databases right, so how could we expect it to predict the future?
The fundamental problem with these gee-whiz policing policies is not solely the technology, but "the lack of transparency and public accountability in deploying crime-targeting tools that could so easily be misused to oppress rather than protect neighborhoods," as the Times opined. I'd take it further. California's law-enforcement agencies are such bastions of secrecy that it makes it hard for the public to trust them as they head off in some innovative directions.
New technologies that pinpoint crime problems and create gang profiles could potentially be useful. They might be better than policies such as gang injunctions that basically case a wide net over entire neighborhoods and restrict the rights of the innocent along with the guilty. But can we trust them?
As a Wired article from last year explained, most departments nationwide use a wide range of Automatic License Plate readers on buildings, signs and patrol cars, thus collecting massive amounts of surveillance data about citizens' every public move. That has helped solve crimes, but police agencies are amazingly secretive about how many cameras are out there, how they use the data and whether any particular tracking might violate a person's constitutional rights.
Wired noted that "Officers misusing law enforcement databases for their own purposes is a perennial problem at the LAPD and elsewhere." That touches on the "trust" problem that is at the heart of any of these controversial systems. Police departments employ thousands of people. Most officers behave honorably, but it's hard to feel confident that police and sheriff's departments, and district attorneys for that matter, do much about those who misuse their vast powers. As always, "sunshine" is the key to accountability, and law enforcement tends to cloak its activities in secrecy.
For instance, this newspaper group is now part of a broad media coalition that will work collaboratively to publish stories about police misconduct by filing a large number of public-information requests under a new law that makes it easier to access such records. Police unions, and some government agencies, have fought the legal release of these important public records under that law. The public would have less reason to be concerned about new surveillance technologies if police agencies were less resistant about conforming to open-government laws.
Consider that California's police-union-friendly Attorney General Xavier Becerra is, as this newspaper explained recently, "threatening legal action against reporters with UC Berkeley's Investigative Reporting Program after they properly obtained spreadsheets with names of police officers, former officers and applicants for policing jobs who have been found guilty of misdeeds including child molestation, bribery and drug trafficking."
One doesn't need a predictive-policing program—or employ a group of mutant savants—to realize that police officers who have been convicted of serious crimes ought not to be trusted with a badge. If California's police agencies want broader public support as they develop futuristic policing techniques, they ought to do a better job earning our trust with the basics.
This column was first published in the Orange County Register.
Comments