This is a story of three neighboring communities with higher than average crime rates: East Orange, Newark, and Irvington. All three communities are located in New Jersey.
It started in 2007 when East Orange declared that their embrace of CCTV and gunshot detectors lowered crime rates and murders from 2003 to 2006. During that very same period, the city of Newark murder rates trended up dramatically each year until Newark installed CCTV cameras and gunshot detectors in 2007.
So guess what happened? In early 2009, Newark declared their new usage of CompStat, CCTV cameras, and gunshot detectors a success because 2008 murder rates were lower by 30%. On May 5th, 2009, Mayor Cory A. Booker and Police Director Garry F. McCarthy declared that 2009's first quarter had the lowest murder count since 1959.
As a surveillance researcher, I know that there is no published evidence that cameras deter violent crime. This is especially true for Newark because they have 280,000 inhabitants and only 109 installed public cameras. This is way too small of a number to have any potential impact. Case in point: Businessweek documented a gun battle and shooting within clear sight of the Newark cameras, so the cameras-as-deterrants argument is simply not valid.
What is valid and very interesting is the fact that gunshot detectors cover a huge area (miles!) and are highly accurate in terms of location triangulation. When East Orange installed gunshot detectors, murders trended up in neighboring Newark. When Newark installs gunshot detectors, guess what? Murders are now trending up in Irvington.
According to the NY Times: Murder rates also sank in the neighboring towns of East Orange and Elizabeth, though they rose slightly in Irvington; officials there speculated that the increase could be linked to drug dealers moving across the city line because of increased police pressure in Newark. Sgt. Keith Sandberg of the Irvington police stated: “Is there a displacement issue? Possibly, Newark made a push, and when you push, people are either going to push back or move.”
While all three communities have been experiencing an overall downward trend in crime since 2003, it appears there may be topical evidence for murder displacement between these towns when gunshot detectors are installed. Gunshot detectors have a very large coverage area which increases their impact which ultimately enhances their effectiveness.
While Newark is declaring all three campaigns successfull, I suspect the biggest impact came from their CompStat initiative and subsequent change to police tactics. I think Newark should also acknowledge and give credit to all crime reduction campaigns including their repeated gun buy back programs.
What lies ahead for Irvington? It's too soon to tell, but this will make an interesting study if murders increase in 2009. Irvington better hurry up and get gunshot detectors before more of its neighbors do!
I agree with you that the success stories given on behalf of cameras often conflate the effects of technology with those of enhanced police presence/techniques. Ultimately, this is going to remain a problem until a careful analysis is done. To date, the most thorough research, e.g, San Francisco and LA have shown limited impact. I believe the conclusions you have reached about these NJ communities also apply to Chicago (the poster child for surveillance). However, I have only a little data to back this point up with.