CCTV, Research Paper by Sebastian Roché, 2002

Also in... (Français)

CCTV does not stop crime

CCTV systems have been rapidly developed.  Why?  What effect do they have?  These two questions need to be answered separately.  There may be good reasons why there is bad practice, or on the other hand bad reasons why there is good practice.  Scientific analysis of data shows us that it is the first case: organisations (city councils, transport companies, landlords, schools, managements etc.) have good reasons to research further security, but resorting to CCTV will hardly or not at all contribute to this.  It is only really in inside car parks where this equipment is installed, but this leads to higher risks in outside car parks.

We do not really need to delve into the complicated reasons which explain the fascination for security-based technology.  If public or private organisations resort more and more to technology such as CCTV, it’s because “classic” human methods which used to monitor areas no longer work.  People no longer keep an eye out for each other: elderly women no longer keep a vigil out of their windows, and the village community model has also disappeared.  Cities are largely anonymous, as is violence: surveys show that only 10% of victims know their attacker.  And other crimes are even more anonymous.  Insecurity has increased in the last forty years and the means to limit crime through the courts are weak.  Politicians and other interested parties look for dissuasive solutions, including CCTV.  Mayors and transport managers need to reassure their voters and customers.  But what can they do?  They may be tempted to give in to the different CCTV advocates, be they social surveys or the installers themselves: at least they offer a way to act (or in any case they appear to, or rather the politicians try to convince themselves they do) in order to help those who need greater security.
So there are good reasons.  But are CCTV cameras the right solution?
The debate on CCTV cameras over whether they intrude on public freedom or whether they act as a deterrent for criminals is irrelevant if you do not know how to measure the effects of the technology.  What good is it debating the dangers of a lack of citizens’ freedoms if CCTV does not protect citizens?  If we accept losing a bit of liberty, it’s so as we can be better protected. Certainly, the predictions of George Orwell in 1984 are definitely wrong: in our society, there is no “thought police”, in fact the opposite: freedom of expression is greater in Europe and in the United-States than anywhere else in the world.  It may be that CCTV doesn’t take any of our freedoms away, but if it doesn’t help to improve the citizens’ protection, the citizens still do not win: it is they who pay taxes or pay through the products they buy (bus tickets or for parking). The citizens’ keen interest in a measure which is going to prove useless will not last.
We need serious evaluations to understand what works.  In the majority of cases, CCTV systems are installed based on personal experience or advice (“Someone told me that they thought that this system worked”), but also from a vague belief (“this should have an effect since ‘we can see better”’), or a small study (“in such a road in a given city there was a reduction in crime following the installation of a CCTV system”).  Only resorting to independent evaluations focused on established methods will be able to establish whether installation would be advisable.  This is important.  The money spent on CCTV cannot be spent elsewhere, and if it doesn’t work, this money has been wasted, and the community will be deprived of the reduction in crime that there were expecting.  The public is not going to be happy for ever with good intentions or good will (if CCTV is installed, it’s because the administration or local elected representative are really trying to tackle the problem).  The greater the security, the lower the crime rate: this is what the community is looking for.
There are lots of “pitfalls” in these apparently convincing demonstrations, the most common being the lack of any references.  Let’s look at an example.  Today, crime is falling in certain communes.  The council newspaper in one of these explains that with this fall, crime has reached a level which is “the lowest seen in 10 years”. As this break through happened after the installation of a CCTV system, temptation would be to conclude that the system is very effective, and that “I believe what I see”.  But the problem with this evaluation is that the neighbouring commune, which does not have CCTV cameras in its streets, has seen the same decrease.  Thus in order to see if CCTV brings a possible benefit, we must be able to compare the changes in the places studied (streets, car parks, schools etc…) and other comparable places (streets, car parks, schools etc…) and in which this technology has not been installed.  This is known as the “control condition”.
What do we definitely know if we only look at a summary of the most indisputable studies? [1] Firstly, CCTV can theoretically have three different effects: marginally lower crime rates, higher crime rates, or neither of these.  In practice, we see all three occurring.
First of all, the lack of effect.  We know that CCTV can help lower crime rates in different ways (by making criminals think that it’s too risky, by sending police to where a crime has been spotted by the cameras, and by drawing the attention of citizens to a threat which makes him take more precautions for example).  In certain cases there are no effects whatsoever, whatever crime we look at (thefts, violence, vandalism etc.).
Secondly, the positive effect.  When looking at places which have CCTV (compared to areas which are inspected), evaluation shows that a reduction in crime, when there is one, is minimal, and in certain cases is so low that it is barely noticeable.  The studies found that crime was reduced by the order of two percent, in particular in residential areas or in city centres.  This reduction is completely negligible, especially taking into account the amount invested.  The fact that the users (national and local police forces for example) find the system practical is another question: they have the impression that it guarantees effectiveness as they “look at the screen and they intervene”.  Evaluation is there to remind them that this is not the opinion of the people who analyse the effectiveness of the measure.
The only strong positive effects are from CCTV systems in car parks, and even this does not concern physical thefts and attacks.  We will come back to this point.
We cannot forget the negative effects, i.e. the opposite to what was expected: a rise in crime.  In the city centres, out of five evaluations, three showed a negligible improvement of 2%, and two others had an “undesirable” effect.  On public transport, two evaluations also showed a small improvement, one with no effect and one with an opposite effect to the desired one.  In London, in Oxford Circus underground station, after 32 months in operation, violent crimes have increased by almost half (47%), thus twice that of the “inspected” (no CCTV) station on Tottenham Court Road!  When all these evaluations are collected together, if we calculate an average, we will not see a significant impact on crime levels.
We are not condemning CCTV in principle, but judging it on the facts.  It can improve situations, under very restrictive conditions, and for specific types of crimes.  This is the case for car parks, as we have already said.  At times, installing CCTV in a car park reduces up to 40% thefts from cars.  But not thefts of vehicles.  We do not know exactly why, nor whether this is achieved by the CCTV cameras or if other factors contribute.  In fact, in the evaluations which saw an improvement in security in car parks, the manager had taken other measures such as improving the lighting, or police patrols had more often passed through the area.  If the lighting and the patrols are enough to explain the improvement, these should be favoured, as they cost less.  After all, the police can easily be deployed when and where crime happens. Not cameras.  Finally, the evaluation reports concerning the improvement of lighting show that this has a better increase on crime with a decrease of twenty percent in the number of crimes [2]. More powerful public lighting could reduce crime more than CCTV; that is one of the lessons from these evaluations!
Furthermore, the studies note the effects of transferrals, in time and in space. When cameras do not operate all the time, crimes are committed when the system is shut down.  When an area is monitored, there are effects which may increase crime in the neighbouring area.  Inside car parks in Hartlepool, Bradford and Coventry have been equipped with CCTV and the number of crimes has dropped.  But, if we look at the changes in the number of crimes in all car parks, inside and out, crime continues to rise.  More precise studies are necessary if we want to know if overall (prevented crimes minus transferred crimes) citizens have lost out.
The debate today is far from the issue of car parks: it’s about extending the miraculous solution of technology to public spaces (schools, shopping centres, streets in city centres).  Today, Great Britain is a country which has greatly developed this option.  In the United States, this has not happened, which has not prevented the country from experiencing a sharp drop in the crime rate.  In France, there are around ten times less cameras in public areas than in Great Britain, which doesn’t it from having a lower level of violent attacks than Great Britain [3].
These results, and above all those from independent scientific evaluations, give elected representatives and managers a lot to think about and consider before a lot of money is invested: CCTV systems are not magic wands – quite the opposite.  We can work out how much CCTV costs, and we can also say that it does not lead to significant improvements.  Do we want to communicate to the citizens by using technology as a clear symbol of determination, or do we want to reduce the risk of being a victim and increase the number of criminals who are caught?

© Sebastian Roché
Political scientist, Researcher with the CNRS,
General Secretary for the European Criminology Society
Last work published: Zero Tolerance? Paris, Odile Jacob.
[1] Brandon C. Welsh and David P Farrington (2002). Crime prevention – effects of closed circuit television: a systematic review, Home Office Research Study 252.
[2] David P Farrington and Brandon C. Welsh (2002). Effects of improved street lighting on crime: a systematic approach, Home Office Research Study 251.
[3] According to victim surveys:  There were 6 violent attacks/threats per 100 residents in France compared to 12.4 in Great Britain in the year 2000.