The city of Cambridge, MA is considering whether to switch on its network of surveillance cameras. Councillor Craig Kelley, who chairs the Public Safety Subcommittee [UPDATE: and whom, I should make clear, is skeptical about the merits of surveillance camera systems, scheduled seven public hearings on the newly proposed Security Camera Policy, but like most subcommittee hearings, they were relatively poorly attended]. The City Council voted unanimously on July 2 to ask the Mayor and the City Manager to arrange a better-publicized meeting to discuss the Policy.
That Her Honor the Mayor and the City Manager be and hereby is requested to arrange a community meeting with other stakeholders to discuss the proposed Security Camera Policy submitted by the Police Department for implementation.
The minutes of the July meeting are here.
This is the history.
Back in 2009, according to the ACLU of Massachusetts, the City Council voted against installing security cameras, but got pushback from the City Manager, who argued that Cambridge would jeopardize its ability to get future grants from the Department of Homeland Security if it did not actually install the cameras. A compromise was brokered where the cameras would be installed, but would not be turned on. It appears that the Cambridge Police Department feels that it’s a good moment, four years on, to raise the issue again. In the discussions over it, the ACLU is arguing for the following principles to be observed:
Involve the public in the decision making process. It is clearly wrong to put up cameras without notice. The public needs to understand the cost of cameras, their use, changes in their abilities, etc.
Establish a clear policy on data retention, access, sharing and how long it can be kept;
Establish privacy principles and make sure policy is implemented. Need robust oversight mechanisms; and
Strict safeguards should be put in place to prevent mission creep. The should be, for example, no warrantless biometric searches without appropriate safeguards.
This brings to mind debates in other communities grappling with whether to accept DHS grants. Local communities perceive DHS grants as effectively “free money for stuff”. In Concord, NH, right now, for example, the Concord PD applied for a DHS grant for a “Bearcat” armored vehicle, citing “daily challenges” from “Free Staters” and “Occupy New Hampshire” (which had not had a meeting in nearly a year). DHS said yes; and now, what police chief, when offered a free tank, would turn it down, even if it really wasn’t needed? The same goes for cameras.
There is a great deal of evidence out there on whether surveillance cameras reduce crime. The most recent overview of the academic literature, or “meta-analysis”, co-authored by Northeastern criminology professor Brandon Welsh, finds that there is no good-quality evidence that surveillance cameras reduce crime, with the exception that monitored cameras in parking lots tend to reduce, or at least shift, vehicular crimes. This is unexpected, given the enthusiasm of police departments across the country for surveillance camera projects. Such projects are often installed instead in places where they have not been demonstrated to reduce crime, such as parks and busy retail districts. Why?
The answer is, more or less, that it’s not all about actually reducing the crime rate. Surveillance cameras do appear to increase the probability of detecting people after they have committed crimes; it’s just that the deterrent effects of cameras seem often to be short-lived, so it doesn’t really affect the overall crime rate. So it’s a technology that makes the process of detection and prosecution easier. However, it may well be that people mainly care about whether cameras decrease their personal likelihood of experiencing crime. It’s widely believed that cameras do, but the research suggests that they don’t. So, if the aim is actually to reduce the crime rate, public dollars spent on cameras are wasted.
It’s hard for people to think in terms of overall crime rates. It’s natural in Cambridge for people’s memories to turn to the Boston Marathon bombings and the role cameras (mostly private) played in identifying the attackers. But here, again, the cameras that were up didn’t prevent the attacks, which is what would affect the crime rate.
I don’t have any policy objection to private parties installing and monitoring cameras on properties that they own. But it’s reasonable to ask whether the actual public benefit of shortening the time to detection and prosecution of the kinds of relatively minor crimes that are usually caught on camera, merit the investment of time in installation and especially in monitoring and maintenance of surveillance camera systems. In Waterbury, CT, for example, surveillance cameras are being installed around the Village Green, a park in the poor and black and Latino part of town, in an effort to suppress “lifestyle crimes” like public drinking and public urination. I question whether installing and monitoring the cameras there will do much more than route more poor people into the criminal justice system, at a heavy cost in terms of monitoring and maintaining the cameras and keeping people in jail.
From a legal perspective, the Supreme Court has found that individual members of the public do not have a “reasonable expectation of privacy” (REP) while in public places, such that they can bring suit under the Fourth Amendment against a government agency installing surveillance cameras that cover public places. They have even allowed the police to install hidden cameras on private property without a warrant in certain circumstances. However, the recent ruling in U. S. v. Jones raises some questions about the REP standard. That case, which involved a GPS tracker installed without a warrant on the car of a suspected drug dealer’s wife, produced very split opinions, but a bare majority found that Jones did have a Fourth Amendment interest in the privacy of his travels over a long period of time, even if all of his travels were technically in public places. It was the finding of the pattern of his travels, not the individual journeys, that triggered a warrant requirement.
What this means is that while there’s no Fourth Amendment issue when surveillance cameras are installed and turned on, there may be an issue if police aggregate the data from a network of cameras, and run search queries that enable them to discern the pattern of a suspect’s movements over a long period of time (28 days or more). The current proposal is to have the data be retained for 45 days.
The issue of the Security Camera Policy is gaining some traction in the City Council elections, which are being held on November 5, 2013. Nadeem Mazen, who is a candidate in that election, argues that if the cameras are turned on,
“I know that in a year, we will regret this decision, and we will attribute this to the slow creep of the surveillance culture over against building our communities. It’s through community-building that we can prevent attacks like the Boston Marathon bombing. Very often, violence, on a large or small scale, can be prevented by attention and pressure from the community. Quick fixes like surveillance cameras don’t have a track record of success; they take away liberty without granting us any actual safety.”
We urge our readers in Cambridge to attend the community meeting, which is expected to happen in September, and to make the point that surveillance cameras of the kind proposed do not actually reduce crime. When we have a date, I’ll post it in the comments.
UPDATE: There were over 50 people at the community meeting last night. ACLU-MA executive director Carol Rose ably set out the case against turning on the cameras. Only one speaker, a business owner in Harvard Square, spoke in favor of the cameras on the basis that “society is completely out of control.” It was surprising to see the depth of public opinion against the cameras, and I’m grateful to everybody who turned out.