Many operators of closed circuit cameras (CCTV) in the report also appear to be bigots, who often watch people for no good reason and do little to protect women. About 600 hours of monitoring by 148 cameras in three cities resulted in the arrests of only about 20 people, mostly for minor offences.
The unpublished study, obtained by The Independent, raises doubts over whether the Government and local authorities are wasting money by spending millions of pounds on installing camera systems.
The Home Office is understood to be concerned about abuse of CCTV systems and Alun Michael, the Home Office minister, is examining ways of providing greater protection for people's civil liberties.
However, the police and ministers are convinced that cameras are effective at catching criminals and preventing offences, a factor difficult to quantify although critics also point out that crime is often simply displaced to another location. The report looks at the day-to-day operation of schemes and not their effect on crime rates.
Researchers spent nearly 600 hours watching CCTV operators monitor cameras in three sites between May 1995 and April 1996. One was in a shopping and business district, with a lot of pubs and clubs, in a major city. The second was on a market square of a large affluent county town and the third was on a busy high street in a poor inner city borough. All locations were kept a secret and were staffed by private security guards.
Nearly 900 individuals or groups were targeted. These resulted in about 20 arrests from 45 deployments of police officers. The majority were for offences involving violence. One-quarter were drink related. One was for a suspected mugging and two involved the brandishing of knives. Others were for theft and vandalism.
The report "Categories of control: the social construction of suspicion and intervention in CCTV systems", by Dr Clive Norris, at Hull University's centre for criminology, and Dr Gary Armstrong, found that operators targeted people they considered "deviant".
Nine out of ten targets were men and four out of ten were teenagers. Black people weretargeted and were up to two and half times more likely to be watched than white people, taking into account their numbers in the population. Nearly four out of ten were surveyed for "no obvious reason".
People were often picked out by "operator's negative attitudes towards male youth in general and black male youth in particular", says the report, funded by the Economic and Social Research Council.
Few women were observed and 10 per cent of the cases were for voyeuristic reasons. The report concludes that the cameras do not make women safer on the streets. There was also some evidence to suggest that "good samaritans" who intervene on camera can later be wrongly implicated in crime. The authors conclude that unless CCTV is better controlled it will become a "tool of injustice" that will further marginalise disadvantaged groups and result in discriminatory policing.
The Home Office has awarded more than pounds 37m to CCTV schemes in the past three years. A spokeswoman said: "CCTV is not a panacea and is likely to achieve the best result if used as part of a wider strategy."
l Report available from Dr Clive Norris, Centre of Criminology, University of Hull, HU6 7RX; pounds 4.Reuse content