Wednesday, June 27, 2007

The Pittsburgh Panopticon

My Pitt colleague John Burkoff and I are both quoted in Rich Lord's P-G piece this morning on the City of Pittsburgh's proposal to install a broad web of surveillance cameras in public places throughout the City of Pittsburgh.

Here's a link to the full Request for Information that triggered the story.

The proposal initially sounds innocuous enough: The City, together with the Port Authority, wants to take reasonable anti-terrorism measures to guard the region's rivers and highways and bridges. That's "Phase One," with implementation planned for this Fall.

Phase Two, with no fixed date attached, is described this way:

Phase Two will deploy cameras in City business districts to promote a safe corridor around the business perimeters and encourage city neighborhood business development. The objective of this phase is to deploy cameras in the business districts and tie them into the existing network developed in Phase One. The Phase Two project is not scheduled at this time.

And here is Phase Three:

Finally, Phase Three will deploy cameras in six-square block area increments in high-risk neighborhoods. As noted in Phase Two camera systems deployed in this phase will also be tied into the existing systems for the final development of a citywide network which will potentially change the way we provide public safety and deploy public safety personnel. The actual locations for Phase Two and Three have yet to be determined, however, the locations may be confirmed before the procurement of any network. Please note that the initial phases are intended as a pilot for the expanded citywide camera network which may be implemented pending the success of the pilot.

I added the bold font to that last sentence because it's the kicker, and it's what I meant when I told Rich that the anti-terrorism rationale is a pretext -- for surveillance to promote general purpose public safety and neighborhood business development. Hopefully the public will have something to say about the plan. Hopefully the public will respond not to the anti-terrorism rationale (in political terms, who could be against that?), but to the proposal's Benthamite implications.

Benthamite implications? It's one thing (though it's not necessarily a good thing) to be watched when you know that you're being watched. It's something else entirely -- and rarely a good thing -- to be watched all the time, when you don't necessarily know it. Jeremy Bentham, known generally for sponsoring a flavor of utilitarian philosophy, also designed the Panopticon. Wikipedia describes the building:
The Panopticon is a type of prison building designed by English philosopher Jeremy Bentham in the late eighteenth century. The concept of the design is to allow an observer to observe (-opticon) all (pan-) prisoners without the prisoners being able to tell if they are being observed or not, thus conveying a "sentiment of an invisible omniscience." In his own words, Bentham described the Panopticon as "a new mode of obtaining power of mind over mind, in a quantity hitherto without example."

Finally, the Panopticon, like any surveillance regime, raises Plato's question: Quis custodiet ipsos custodes? Who watches the watchers? Back to what I told Rich: The Pittsburgh plan is completely silent on what I call "the human back end." So Pittsburgh arranges to collect all of this surveillance data. What then? Who sees the data? What's done with the data? When? And why? For the dystopian version, watch Enemy of the State. It's not just local law enforcement watching. It's the bad side of the National Security Agency.

But we don't need to invoke dystopia; more likely than comprehensive and conspiratorial government abuse is comprehensive municipal incapability. Do the City and the Port even have the resources to make the program effective for its intended purpose? Or would Pittsburgh be collecting a lot of data without the ability to make meaningful use of it? My understanding of the British and German experiences with these CCTV systems is that the systems are very, very expensive -- and yield very, very modest -- or at least often unscrutinized -- benefits. Pittsburgh can do this better? Would this money -- more than $800k of local money -- be better spent on more cost-effective public safety initiatives?

11 comments:

Schultz said...

I love it, security cameras all over downtown Pittsburgh. All that action downtown will put whoever is behind the cameras to sleep. This seems like another one of those "well, federal funds cover most of the cost, so let's just spend the $800k."

Like you said, with all of the problems this city has that $800k could be put to better use - public schools, transportation, economic development, etc, etc.

Jim Russell said...

Pittsburgh will have more data than it can possibly use. The genius of the panopticon is not that it allows the watchers to watch all the prisoners 24-7. What the panopticon does is make the prisoners unsure if they are being watched or not. So, they modify their behavior accordingly.... just in case.

Michel Foucault once wrote that confession is like a panopticon. Priests can't be everywhere for everyone, modifying behavior. Confession is a "technology" that enourages people to self-modify, self-regulate.

Anonymous said...

Perhaps they intend to use them like many indoor security cameras are used: No one is watching, but if something happens they can check the video. It wouldn't directly reduce crime, but it might make it easier to catch the criminals. That in turn might make people think twice about before commiting a crime. Maybe.

Now, is it worth the money? Who knows. I suppose it's cheaper than putting more cops on the street? But whether it will be more or less effective remains to be seen.

Jefferson Provost said...

The problems of having "more data than [they] can possibly use" will be resolved within the next 10 years. Automated identification and tracking of faces and vehicles, and detection of
suspicious movement patterns are all already possible in the laboratory. They are probably being deployed in wealthy surveillance environments like Las Vegas casinos.

Time progresses and Moore's law continues to apply. It won't be long before the systems are cheap and robust enough to be used by a resourceful city. Once the cameras and network are deployed, upgrading the system to use automated processing/tracking.

I think that this kind of surveillance is inevitable, and it will make the streets safer. Mike's question about watching the watchers is the right one. If think it makes more sense to fight for something like David Brin's Transparent Society than to try to prevent surveillance made ever cheaper and easier by technology.

Mike Madison said...

The fact that casinos already use a lot of this stuff is a clue to some of the political and social problems that persist even as the technology gets less imperfect:
--The relatively wealthy and powerful are watching the relatively poor and powerless;
--They're not watching to protect you. They're watching to protect themselves;
--And they have turned over data to the FBI on a non-specific basis.

Jim Russell said...

The technology cuts both ways. Watching the watchers will also become more practical and effective.

Not to get too utopian, but open and transparent knowledge ecosystems such as the blogosphere are more efficient at innovation than closed systems such as government.

To make my point another way, the internet can already handle the deluge of data produced by ubiquitous surveillance. Put those streaming cameras online and let the public see what the government sees.

Mike Madison said...

Don't worry, Jim, I'll always call out utopians. Who watches the watchers watching the watchers? It's one thing to have extra computing cycles looking for extra-terrestrial intelligence. It's something else to have my neighbors watching law enforcement watch me.

Of course, this is Atlas, and we know that it's turtles all the way down. This might the sort of thing that, like Wikipedia, works in practice but not in theory.

Anonymous said...

So we can get 24/7 security cameras to spy on the public but we still can't find out how much employees of the state earn?

Without some sort of Sunshine Law to allow us to know what our government is really up to, how can we trust them not to abuse a surveillance system as well?

Given the history of public employees abusing things like DMV databases, why shouldn't we think this would be any different?

Jefferson Provost said...

Yesterday afternoon on WDVE, Sean McDowell mentioned the camera plans. (No mention of the multi-phase pilot program, though.) He was getting opinions from callers. The opinions that he played on the air were universally negative.

Not surprising, considering how he presented the news, but several calls had weird twist: They said that surveillance cameras on (privately owned) private property were perfectly acceptable, but (publicly owned) cameras in public spaces were an unreasonable intrusion.

So... the only place these callers expect privacy outside their own homes is in public!

Mike Madison said...

The callers' perception isn't so odd, I think. We've gotten accustomed to the phrase "expectation of privacy" because the Supreme Court gave it to us, but as a conceptual framework, it's terribly misleading. We *do* have legitimate personal or private interests when we're out in "public," in the sense that we conform our behavior in various ways to (i) what we know the law to be (ii) our surroundings and (iii) the behavior and expectations of the people around us, i.e., social norms. We *don't* as a rule conform our behavior to the expectations of people who might (or might not) be watching from some eye-in-the-sky. When cameras go up, there's a meaningful risk that this may change. (Ever do something in public that was legal -- but that you would have thought twice about if you suspected that the police -- or your spouse -- or your boss -- or someone you didn't know at all -- was watching?) That's Bentham/Foucault, mentioned earlier by Jim "Globalburgh" Russell. To some extent this is changing already via camera phones, mini-cams, and blogs, as Tom Friedman wrote in the Times the other day ("The Whole World is Watching," hidden, unfortunately, behind stupid Times Select). The results of that sort of thing are usually pretty transparent. At the least, public authorities should be held to the same transparency standard.

Jim Russell said...

I think the key to this issue is the point with which Mike ends his post. Is the plan cost-effective? I imagine the answer to that question is rather complicated.

Might I suggest a better use of the $800,000? Let's install cameras in all the government offices in Pittsburgh and stream a live feed to the public.

The public could learn a thing or two from Bentham and Foucault.