|Why Surveillance Cameras Don't Reduce Crime
As a resident of the UK in South London, I am only to aware that the cameras
do not work.
The main problem is one of cost, security in any form costs a lot of money
it's supposedly costs UK businesses and home owners 4 Billion GBP (8,000,000,000
USD) This is not including the Government spend on the Police etc.
One effect that has been noticed is that initially crime moves away from the
cameras which is why they are often touted as a success. However within a short
period of time the crime returns. Usually the criminals have baseball caps and
hooded tops etc to hide their identity, and "do the job" very quickly.
The criminals obviously know that a camera like a burgler alarm only
represents a threat when you don't know how to alow for it in terms of time,
identification hiding etc.
The simple fact is that you need a human elerment on the ground (Police etc)
that can respond quickly.
Oh the big plus point for security cameras in the UK is that you can use them
for making money, you just sell the video footage of the criminals etc to who
ever wants to buy them.
The latest Idea is to cover the UK with special cammeras to pick up
unlicenced car drivers, so the secret is out they are now just there to make
money, like the Gatsos and other "traffic calming" technologies.
Sorry if I sound a bit cynical about CCTV and other cammeras, but I cannot
walk out of my front door without being observed by atleast two cammeras, and
still be in view if I walk the three or four miles into my local town center. Oh
and the violent crime etc on my road and the surrounding area has gone up for
the last three years.
Posted by: Clive Robinson at March 31, 2005 11:49 AM
Security cameras have the effect of making people (falsely) _feel_ safer and
as a result they reduce the use other security measures (such as not locking the
back door or not using the house-alarm) and thus making the robbers' job much
much easier. This is a classic example of an unintended consequence.
Posted by: Saar Drimer at March 31, 2005 01:06 PM
Bruce, do you agree or disagree? The link you provided might actually be
interpreted to say the opposite of your title. For example, the writer
"I don't know anybody, though, who thinks there shouldn't be a camera behind
the counter at the convenience store."
Hmmm. If they (you) think surveillance cameras do not reduce crime, then why
would they say all convenience store counters need one?
Or are you actually trying to announce something similar to "guns don't kill
people..."; something catchy like "cameras don't stop crimals, cops do, so we
need more cops"?
Posted by: Davi Ottenheimer at March 31, 2005 01:48 PM
Maybe the problem is that we don't have ENOUGH surveillance. David Brin
writes about what he calls a "transparent society" in which surveillance cameras
are everywhere. Then, criminals could be tracked from the scene of the crime
back to their lair. Wearing a hat or a hood would not help them because they
could be physically traced and arrested.
Posted by: Cypherpunk at March 31, 2005 02:15 PM
Roughly twenty years ago, radar/traffic cameras were tested in (I believe)
Friendswood, TX to catch speeders (it's a crime, too). The test failed as the
cameras couldn't differentiate between the judges, police and other members of
the mandarin court and the little people. I wonder if and when surveillance will
become so pervasive as to be opposed by the privileged.
Posted by: John Navratil at March 31, 2005 02:25 PM
Imagine my surprise at yet another article posted here in which yet another
attempt at security is denounced.
Bruce is always quick to point out what *isn't* secure, but we've still heard
nothing from him about what security practices he *would* recommend.
Apparently, we're all screwed. There's nothing we can do, so let's all simply
not do anything!
Oh, wait, that's not secure either.
Posted by: Eric K. at March 31, 2005 02:30 PM
Bruce, you know, the more I think about this, the more I wonder what you
might be thinking by annoucing to the world that surveillance does not reduce
Yesterday your log mentioned the futility of trying to avoid identity theft.
Fine, as Eric K. points out you seem to say we're all screwed. Maybe you are
depressed about the state of things.
But let me point out that C. Drake made a brilliant comment about a real-life
case where a guy tracks down and successfully gets his ID thieves convicted. You
should check it out. In particular, you should make note of the part where the
"In Portland, the police department is so strapped that unless it's a
person-to-person crime, it's pretty low priority."
Why does he mention this? It's because the victim is hot on the trail of the
thieves and is standing at the same register where his identity was stolen, and
he wants the police to respond. Instead, he's working the case himself.
"I hop in the car and drive down to Denny's and ask to speak to the manager.
[...] I pulled out my Visa. 'This card was used here this morning. Someone has
seen the thief. I know your registers store the day's credit card transactions.
Is there any way you can look up this number and tell me who served them?'"
Yeah, that's right, a CAMERA would help catch the bad guy. Ever step up to a
bank teller and wonder why there is a dedicated camera pointing right at your
mug? Perhaps it is because it performs a preventative as well as detective
control function for security.
Granted, the ACLU researcher you linked to has some interesting questions
that need to be answered, and maybe you were in a rush to post a title, but
today's log needs serious clarification.
Posted by: Davi Ottenheimer at March 31, 2005 03:10 PM
Surveillance cameras are about as good as locks... they are made to keep the
honest individual honest in a timely fashion... and really nothing more than
evidence that a crime has occurred at one point in time.
Posted by: Israel Torres at March 31, 2005 03:38 PM
I'm the author of the linked post, and though obviously I wasn't clear, I'm
not arguing against camera placement in convenience stores because those are
private cameras and the store owner can do what they want. I'm arguing against
cameras used by the government, by law enforcement, particulary in public
Anybody can toss out hypotheticals, but the longitudinal study from Britain
cited in the Grits post found that generalized camera surveillance empirically
did not reduce crime, although in narrowly defined circumstances (especially
parking facilities) they measured some benefit. Historically, when the
government installed cameras, the motivation was essentially a hunch that they
would help, but it was an intuitive leap, not a proven fact. Now, studies like
the one by the British Home Office are identifying precisely where cameras
prevent crime and where they don't. Using them elsewhere, I'd argue, is a waste
of police resources as well as unnecessarily invasive of privacy.
Posted by: Scott at March 31, 2005 05:17 PM
By the way, Bruce, thanks for the link! I just bought a copy of Secrets and
Lies I'm planning to read over the weekend, so it's a timely honor.
Posted by: Scott at March 31, 2005 05:21 PM
Perhaps some of those commenters who have merely skimmed the article may have
missed the link to the official United Kingdom Home Office criminological
research study headed by Professor Martin Gill:
"The impact of CCTV: fourteen case
This confirms previous studies, that CCTV in the United Kingdom (we have far
more experience of it than most other places in the world) does not decrease
crime or the fear of crime, for various reasons, which include the fact that
many CCTV schemes are not linked to expensive control rooms which are properly
manned 24 hours a day, 7 days a week.
CCTV does not have much of a deterrent value if the criminals do not actually
know that they are under observation. What is the point of a warning sign which
you cannot read in the dark, although the CCTV surveillance camera may have
night vision, or which is beyond your visual range, even though the camera has
powerful optical and digital zoom capability ?
Most CCTV systems have far more cameras than warning signs (some have no
visible warning signs at all), which is the wrong way around, if they are meant
to act as a deterrent.
Apart from the poor quality of many CCTV images from older systems,
especially ones which use analogue video recorders and which often re-use the
video tapes too often, and never pay to have the camera lenses cleaned
periodically, there is the whole problem of actually positively identifying
someone even from a good quality image.
Human beings are even worse at positively identifying faces from CCTV, than
they are in picking out suspects from a police line up. The failure rate is
somewhere between 40 and 50 percent.
Then there is the whole question of actually using CCTV footage as
evidence in court. Most systems simply have not invested in the
duplicate signed copies, sealed evidence bags, independent evidence custodians
etc. that are needed to provide a court of law with an unbroken chain of
evidence, showing that it has not been tampered with.
A date time stamp on a printout from a CCTV monitor can be faked on the most
basic of personal computers, and far more sophisticated video editing and
manipulation tools are easily available, especially for digital systems where
there is often no physical evidence of tampering, which can sometimes be
detected with analogue tapes.
Consequently CCTV footage is rarely, if ever, presented directly in court,
where it might be challenged frame by frame. Not every trial is a Rodney King or
OJ Simpson media event.
The same is true for all the "add on" technologies which go well beyond
simple video CCTV, which are being tested and deployed e.g. Automatic Number
Plate Recognition, facial recogniton, "gait" (how you walk) analysis,
"Suspicious behavior" anlaysis (how long you linger near a "protected" object),
whether you appear to be engaged in "stabbing" or "kicking" motions etc. all
conducted in the visible light or infrared spectrums, with or without photon
multiplier image intensification.
Then there are the controversial and voyeuristic "see under your clothes" or
"see your child naked" imaging technologies like Passive Millimetre Wave,
Teraherz, Ultra Wideband and Low Intensity Backscatter X-Rays imaging etc.
These technologies cannot be dis-invented, and they have their place in the
armoury of anti-crime and security tools, but they are simply not a
technological magic fix for societal problems, especially where they are somehow
promised as being "cheaper" then employing police or security guards "on the
Posted by: Watching Them, Watching Us at March 31, 2005 05:28 PM
The essay quoted is quite correct insofar as it goes. I also note a story
that is of some peripheral interest:
However, while monitoring cameras so as to detect crime might not be
productive, using the tapes as an aid to solving a crime appears to work quite
Posted by: Peter at March 31, 2005 05:32 PM
"Aussie man reports crime in Devon
The Octagon kiosk is covered by
A man in Australia tipped off police in Devon after seeing a
suspected burglary on a webcam based in Exmouth."
"It transpired the pair were a man and a woman having an argument, not
conducting a burglary"
A false alarm, even one reported from the other side of the world, is still a
Posted by: Watching Them, Watching Us at March 31, 2005 06:04 PM
Thanks for clarifying. I understand your position: "I'm arguing against
cameras used by the government, by law enforcement, particulary in public areas.
Anybody can toss out hypotheticals, but the longitudinal study from Britain
cited in the Grits post found that generalized camera surveillance empirically
did not reduce crime."
If you take Bruce's log entry "Why Surveillance Cameras Don't Reduce Crime"
and read your post, and then read the study, you have to wonder why the log
entry wasn't titled "How to Reduce Crime with Surveillance Cameras" or "How to
Effectively Deploy and Manage Surveillance Cameras". Alas, Bruce simply stated
that Cameras do not reduce crime....
What you have happened upon is an opporuntity to make Surveillance technology
more effective in reducing crime. All I ask is that while in hot pursuit of
Civil Liberties you try to avoid throwing out the baby with the bath water.
Surveillance technology is just now reaching a period of innovation and
adoption that will make it a relevant and useful tool in prevention as well as
detection of crime. We would be remiss to blame the failure of the deployment
and administration of cameras on the devices themselves, on Bobby's or even
Posted by: Davi Ottenheimer at March 31, 2005 07:02 PM
I'm not sure that I'd say that security cameras, even the sort of
science-fiction see-all network scenario proposed earlier, are going to deter
crime, simply because they don't target the reasons why people commit crimes in
the first place. Nor would I think it feasible to use tapes or files from such
devices in court, because proving something "beyond reasonable doubt" is such a
big ask when you have to convince a jury. Muddy the waters enough with
timestamps and grainy images, and any jury will get confused.
But I see them as just another tool in detective work. Use them in
conjunction with other techniques to find suspects or reduce the pool to a
number that can be dealt with by other means. A good cop can usually work out in
an interview if a suspect is worth investigating further to gain hard evidence
that can be used in court, such as through a search, or through evidence from
third parties, or records from banks and other institutions with reasonable
security on their information.
What concerns me about these things is whether they are cost-effective. If
they aren't delivering results, then it makes no sense in spending the
community's money on employing people to bore themselves rigid watching a dozen
monitors at once, or worse, to amuse themselves by prying into the lives of
community members, such as zooming a camera down the top of a pretty girl, or
following the progress of a romance in a side street. Add in infra-red or x-ray
capability and you are just asking for misuse.
To my mind, the answer in using surveillance cameras as a crime-fighting tool
is to make police departments pay for them out of their normal budgets. If they
get results, then they will be paid for. If not, then the money and manpower
will be diverted to more productive endeavours.
Posted by: Peter at March 31, 2005 07:12 PM
We can complain about the CCTV technology, or we can gather up a list of
requirements and map it to the emerging technology. The fact is, the failures of
existing and past surveillance technology is being addressed and improved fairly
rapidly, if not exponentially.
Quite frankly, I find more people ask for surveillance than not these days
when we talk about how to feel safe when walking to their car, standing in
elevators, etc. (as the study rightly identifies). This should take us directly
towards the reason why I am happy to see that the ACLU is interested in the
issue: when people realize security is FOR them and not to be used AGAINST them,
they generally can't wait to sign up. The problem is therefore mainly about
restoring trust in public spaces, so I look forward to hearing the ACLU's
recommendations on how that can be accomplished today. Sorry, no hypotheticals
Posted by: Davi Ottenheimer at March 31, 2005 07:16 PM
The latest Idea is to cover the UK with special cammeras to
pick up unlicenced car drivers, so the secret is out they are now just there to
make money, like the Gatsos and other "traffic calming" technologies.
I think you'll find the issue with unlicensed drivers, is that that also
means their insurance may not (will not?) pay out in the case of an
Before you see this as 'revenue' raising, consider the issue of someone you
know/love being hit by a car driven by an unlicensed driver. Any expensive
medical bills, future treatment and payouts you may expect to receive for care
may not be forthcoming.
Finally, as to speed cameras, if you don't want to pay, don't speed. It's
actually that simple.
Posted by: Anonymous at March 31, 2005 08:39 PM
"Apparently, we're all screwed. There's nothing we can do, so let's
all simply not do anything!"
The Register has an interesting perspective on the problem you identify:
"But why, asks a reader, has the MSM [mainstream media] ignored [Spam King
Richter's bankruptcy]? Probably because after a decade of libertarian
propaganda, a kind of weary fatalism has set in."
Email is an excellent example of a the challenge of technical evolution and
mainstream adoption with regard to a "throw up your hands and quit" approach.
Spreading gloom and doom does not negate the fact that regulation is a big piece
of the puzzle (along with technical and security innovations).
So Bruce, let's see some innovative and pro-regulation entries,
Posted by: Davi Ottenheimer at April 1, 2005 01:20 AM
My personal experience with CCTV...
I parked my motorcycle in possibly the most secure location in the whole of
Hitchin... In clear view of not one, not two, but three CCTV cameras; within an
arms length of the Corn Exchange door staff and a busy area of town.
On my return, approximately 30 minutes after closing (by which point the door
staff from the pub had gone and it had qot a bit quieter) my bike had gone.
[It was probably also the only time I didn't use an additional lock]
I reported the incident to the police, who did nothing - not even check the
CCTV tapes! (I know this because I have a contact in the council, who operate
the cameras [and therefore tapes]).
...if the police are not going to use the resources available to them
(probably because they are out hassling motorists) why bother?
Posted by: Ben Smyth at April 1, 2005 05:38 AM
OT [very OT] On a separate note, you may be interested in Steve Bellovin's
security panacea -
Posted by: Ben Smyth at April 1, 2005 05:39 AM
@Watching Them, Watching Us
I do believe that under british law warning
signs are REQUIRED to be posted in order for the evidence to be admissable in
court, not having those signs makes them even more pointless and a waste of
money, I believe they work in certain instances, in certain limited ranges, but
plasting them over Brixton high road does bugger all to lower crime , the angle
of some of them requires the criminal to actually have to look up in the
direction of a camera to get a good facial, i mean whos going to do that?
Increasing the police force is one option, but that doesnt give people much
assurance either, as alot of us are loosing faith in the force
Posted by: Phoenix at April 1, 2005 06:43 AM
Very often the example found here remind me on that famous machine that
goes 'ping' from the Monty Python film "The Meaning Of Life".
It seems that a lot of technological "solutions" are machines that go 'ping'.
They don't necesserely solve anything, but there's a lot of 'political'
motivation to buy them.
Posted by: AC at April 1, 2005 09:11 AM
I am not a car driver I preffer to ride my push bike. Also I do not know if
you are from the UK or not, but the traffic laws etc are somewhat odd in the
First off in the UK there are allready quite effective messures in place for
collecting unpaid "road fund" which the tax is often known as and is around
Basically if you are the "registered owner" of a car and you have not paid up
you get a court summons after about 1.5 months. There are also other arangments
for imprisonment etc.
Also you pay your "Road Fund" to the UK Government (central not
regional/local) who then spend it on what they see fit (Generally not roads
accident prevention or Health or other related activities).
I looked back to some of the recent quotes about these new cammerers given by
the press, and it would appear they are for directly applying "extra fines"
which would have the revenue split between the cammer operators ( the likes of
Capita etc) and the authority (either police or regeional/local government).
These Local authorities then spend the money on what they like. On a side note
Capita who run the congestion charging system in London have been subject to
many many (upheld) complaints about the way they extort money out of motorists
who have paid etc.
SO you get fined by central government if you do not pay your road fund, and
get additional fines by the local government if you drive your car in their
I have also seen reports on studies that show that "non physical" traffic
pasivation systems not only do not work, they actually increase the likley hood
of accidents. This is because drivers get to know where the cammeras are and use
excessive speed between them then brake suddenly to reduce speed where they are.
The results of this lunatic behaviour are fairly obvious and have been reported
in the press. Also that some people simply regard the fines as a normal business
expense so they do not act as a deterant.
Additionally the tab for road accidents is usually picked up by the health
service in terms of immediate (and often longterm) care.
SO I still stand by my original comment that the cammerers (being a non
physical traffic control system) are just a money raising system.
If anybody has evidence to the contry I would be very happy to see it.
Posted by: clive robinson at April 1, 2005 09:40 AM
Camera survelliance needs money to work.
Hollywood has tons of
Posted by: Diego Zenizo at April 1, 2005 09:49 AM
@Ottenhemer, you wrote:
"What you have happened upon is an opporuntity to make Surveillance
technology more effective in reducing crime."
The British study found that cameras didn't reduce crime in 13 of 14 areas
tracked over time. We can debate whether the glass is 13/14ths empty or 1/14th
full, but on the whole, I think it's hard to argue for continuing to waste
resources on generalized camera surveillance in public areas.
Also, just to have said it, the opinions expressed on my blog are my own, not
any ACLU policy unless expressly stated. The post represented my personal
musings trying to reconcile the results of the British survey with Dennis'
arguments. I proposed an untested, maybe untestable hypothesis to explain the
study results, but wouldn't want anyone to draw implications from that about
ACLU's official stance. Best,
Posted by: Scott at April 1, 2005 04:19 PM
Why would we believe that surveillance cameras would deter criminals
You have to understand the way criminals think. Most people who
commit a crime do not believe they will ever be caught however lacking in logic
this sounds. They also tend to be more impulsive and lack the ability to really
think things through.
Convience store robers tend to be after some quick cash
to meet the needs of an addiction so they are even less smart about their
Do cameras increase arrest rates?
For the careless criminal they would
seem to make good evidence.
The problem with survelance cameras is that they are a tertiary response to
the problem, they come to late in the chain of events. The best they can be used
for is to try to catch the criminal not prevent or intervene.
Posted by: Laura Jean at April 1, 2005 05:39 PM
"I reported the incident to the police, who did nothing - not even check the
CCTV tapes! (I know this because I have a contact in the council, who operate
the cameras [and therefore tapes])."
Aw, c'mon! Just how often do police do much about stolen vehicles? They take
the details, but if there's nothing unusual - like there's a kid or a briefcase
full of diamonds inside - that's about it.
I realise you'd like the cops to drop everything and get back your motorbike
quicksmart, but fair suck, mate!
Now, if you had been murdered in full view of the cameras, you could expect
some serious action.
Posted by: Peter at April 2, 2005 01:38 AM
Point taken, but there was a spate of bike thefts going on at the time -
hence you'd expect some organised outfit.
Posted by: Ben Smyth at April 2, 2005 05:19 AM
I'd expect "some serious action" to a murder, cameras or not.
I believe the point is that the cameras were put in place to prevent lesser
crimes (like bike theft), not on the of chance that they catch a murder.
The camera system failed to prevent the theft and was not used to recover
Either the system failed, or people were lied to about the reason it was put
Posted by: Thomas Sprinkmeier at April 2, 2005 06:58 AM
Stealing a bike and murder are two extremes of the crime spectrum. I
seriously doubt that these particular cameras were touted as being able to
eliminate minor crime. Crimes that very few police departments would do much
about anyway, not because they don't care, but because they have more serious
crimes to investigate.
I suggest that if a police department has time to chase down security camera
tapes of a minor crime like this, then the community probably doesn't need much
in the way of security anyway.
But if the crime had been more serious, say robbery or assault or rape, then
action would likely have been taken to examine the tapes, not just for possible
identification of the offenders but to determine what had happened.
It's not the cameras that determine if action should be taken, nor even the
people controlling the cameras. It's the people who are going to have to
investigate the crime, make an arrest and prosecute the offender who determine
if action will be taken.
Posted by: Peter at April 2, 2005 01:27 PM
As I see it, the point of this article (which I agree with) is that cameras
by themselves don't deter crime. There has to be somebody watching the camera,
or at least scanning through them often enough that a crime is likely to be
seen, and the watcher has to have the ability and willingness to send help to
the scene immediately. (A worthwhile enhancement would be a mike that detects
specific sounds such as a scream and switches the scanning person's view to its
If this level of service is not going to be provided, then I'd put up signs
warning people that they're responsible for their own safety and advising them
to carry guns (if allowed there).
Whether cameras are unacceptable invasions of privacy is another question. I
feel that even when used in public places, where people's coming and going are
public information, if there are so many cameras (and good enough software for
viewing what they've seen) that you can follow a person's movements from place
to place, then doing so is in effect "stalking". Doing so should be prohibited
for non-police (outside of their own private property) and should require a
warrant for police.
Posted by: John David Galt at April 2, 2005 02:14 PM
You disable the things with a shotgun or the tools you have. Wear a mask and
the camera is useless. The camera is $600.00 and the mask is $1.00 at the thrift
store. The security people are sitting there looking at footage a week after the
crime is over and done with. More dumb world domination schemes, so put cameras
Posted by: Jim at April 2, 2005 10:43 PM
"The camera is $600.00 and the mask is $1.00 at the thrift store. The
security people are sitting there looking at footage a week after the crime is
over and done with. "
What's the problem? The camera isn't going to solve the crime. But seeing how
a criminal acts, even a week after it happens, provides a lot of information.
Enough for modus operandi to be noted and perhaps linked to somebody already
A security camera is just another tool for crimefighting. Not a magic
Posted by: Peter at April 3, 2005 01:58 PM
I did live in the uk until recently, so I understand the rules.
"SO you get fined by central government if you do not pay your road fund, and
get additional fines by the local government if you drive your car in their
I'm not sure what you are complaining about... OK, if you believe the press
(and why wouldn't you) these are purely revenue raising cameras for people who
haven't paid their road tax, what's the issue ?
Fair enough if you've paid and they say you haven't ... complain.
Fair enough if your car is stolen (kept past the renewal date) and you get a
fine ... complain.
If i hire a dvd from blockbusters and don't return it, they will fine me. If
i hire a book from the library and don't return it, they will keep adding fines
onto the account until i return it and pay off the fines.
So why should life be magically different for motorists ?
Posted by: Anonymous at April 3, 2005 08:32 PM
As I indicated I am not a motorist (for various reasons), however I have two
The first is the cost of such a system compaired to it's extreamly low return
rate (Look into the London Congestion charge for an example of this). The only
people who realy benifit are the private companies that run the systems.
The second is mission creep, after the expensive system has been installed
and found to be nolonger effective (criminals work by natural selection which is
why technical only security systems initialy work then fail), they have to find
a new use (such as road tolls) to justify extracting vast quantaties of money
out of people for carrying on their everyday activities (see my earlier post
with regards to micro chips in dustbins).
The result of this sort of stupidity is that a percentage of the population
will always find ways around these systems and usually at the expense of
everybody else. The goverment solution is to get more and more high tech
solutions which are all going to fail eventually as people learn to adapt to
them... The only people to benifit are the private companies who supply the
If you want security the solution you put in place needs to be as flexable
and adaptable as the criminals (ie human beings). Then support them correctly
with adaptive technology to give them a slight edge. Not the other way around,
it is doomed by the proceses of adaption and natural selection by the criminals
who will always end up with the edge over the police.
An example of how to make a technology like cammeras flexable is to put them
in rented unmarked cars/vans/etc and move them around so finding them is very
difficult, therefore the ability to adapt by the criminals is vastly
It is after all a game of Cat and mouse, and the cats should always behave
like cats not mice.
Posted by: clive robinson at April 4, 2005 04:30 AM
I did read the original paper, as I find that newspaper reports of
research are rarely accurate. On the basis of this reading, I have to say that
the report which characterises this as failing in all but one case out of 14 is
For a start, although there were 14 projects, they did ~75 studies within
these, depending how you count them. Prof. Gill found that in many cases it was
not possible to arrive at a definite conclusion due to the many difficulties
that can be associated with such studies. This was not necessarily because no
effect was seen nor necessarily because the effect was small; for example in one
case vehicle crime was reduced 41% from 1641 to 972, but this apparently huge
result was statistically insignificant because the background rate in that
region fluctuated by very wide margins. If a study can't distinguish between a
genuine 41% reduction and normal variations then the only way to obtain useful
information in that study would be to maintain observations for a much longer
period than their funding allowed. In another case a fall in vehicle crime of
42% _was_ statistically significant, but still could not be counted as changes
in parking regulations had altered the manner of parking vehicles in the area,
so the results where not directly comparable. Many other examples were
confounded by the fact that cameras were installed at the same time as other
crime-reduction projects. Altogether only 16 of the 75 studies returned usable
data, indicating overwhelmingly that this study was far too short to be useful.
Of those 16, 6 studies in 3 projects showed a definite positive effect, which is
a good deal better than 1 in 14.
Among the remaining 10 we have the difficulty of studies in which a reported
crime rate increased, but it was the opinion of the authors that this was due to
improved detection or reporting of an existing crime rate (mainly shoplifting
and public disorder near pubs), usually due to other projects installed at the
same time. Again, more research is required to really understand these
There were also two studies in which there was a definite increase in crime,
but in one of those cases the increase consisted largely of attempts to
vandalise the cameras, at the same time as the burglary rate plummeted; while in
the other, the camera setup was a new experimental system which was found to be
It is also notable that there seems to be considerable variation according to
the care in design and planning of a system, ranging from the Borough experiment
which was completely useless, to the carefully planned Project Hawkeye which
resulted in a statistically significant 73% reduction in its targeted crime.
Overall, it seems to me that the principle results of this study are:
that it was too short and more work is required;
* that in the cases where a
definite conclusion can be reached, CCTV produced a useful effect in a little
over one third of studies;
* to the extent cameras are effective, they are
more effective in high crime areas; and
* there is considerable difference
between carefully planned and carelessly planned or poorly maintained
Posted by: Roger at April 4, 2005 11:30 AM
... not to mention that who knows who is on the other side of the camera
lens? Sure it may be intended for a security team, but there are plenty of
examples where attackers have complete access to a surveillance system without
being discovered in a timely fashion. The more complex a system, the more places
there are to hide. (e.g. wireless systems, internet cameras)
Posted by: Israel Torres at April 4, 2005 03:19 PM
"Overall, it seems to me that the principle results of this study are:
that it was too short and more work is required;"
Isn't that the result of every study?
Posted by: Thomas Sprinkmeier at April 4, 2005 06:51 PM