The chattering classes, particularly those who are products of the Sixties and Seventies, instinctively oppose any extension of police activity. Yet the events of the past few years make many of them uneasy. Last autumn, following on from the saturation coverage of 11 September, the media reported a wave of crime stories. You could not go to a dinner party without hearing of another burglary or of a mobile phone robbery committed against the son or daughter of the household. An estate agent in Battersea was murdered in a carjacking; a young Asian woman was shot in the head during a mobile phone robbery. Major cities, it seemed, were being overwhelmed by crime.
In the spring, during the run-up to the local elections, the political parties' focus groups found crime was the issue of the moment, greater even than education, health or transport. But there is an important difference between crime and these other issues. You can buy private education or tuition, you can buy your own car, you can take out private health insurance. In other words, you can step outside state provision. That is far more difficult and a much greater step, when it comes to policing.
Overall, crime in Britain is falling and has been for about five years, whether you take the official statistics, based on police reports, or the British Crime Survey, based on the public's actual experience of crime. But the type of crime that is most feared - violent, personal, confrontational crime - is growing. Moreover, the number of calls to the police for emergency assistance - not only about crime but about the incivilities of life, such as the drunks, the gangs of youths, the road rage, the vice, the drug dealing - is growing at an exponential rate. In 1996-97, the control room at Scotland Yard received 1.5 million emergency calls; last year, that number rose to 2.5 million. That explains why you now so often see the police whizzing around in striped cars.
It also helps to explain why, as people so often complain, the bobby on the beat has become a threatened species. But that is not the only reason. New forms of crime and new legislation have combined to create a range of police tasks that simply did not exist 20 years ago. As well as coping with an increasingly complex criminal justice system, the police have to deal with child protection, counter-terrorism, drug trafficking, people smuggling and so on. The amount of time available for visible patrol has thus diminished, and the result has been a decline in public reassurance as well as a decline in police capacity to deny criminals the kind of habitat where they are most effective.
Yet both police leaders and government were complicit in the near-extinction of beat work. From the mid-Seventies, police leaders were increasingly conscious of the withdrawal of other agents of social cohesion - typified by caretakers and park keepers on the one hand, churches and trade unions on the other - that had underpinned an approach to policing which had worked for most of the 20th century. They were also faced with evidence that random patrol failed to deter anyone, and with the need to devote more resources to investigating increasingly numerous and complex crimes. So they turned to a new, community-based, problem-solving approach, trying to get upstream of the factors that cause crime: the built environment, school failure, drugs, alcohol or inequality itself. Although worthwhile, this approach also proved ineffective because it coincided with the disintegration of authority in western society. The problems causing criminal behaviour became ever more complex and, as it were, farther upstream and harder to beat.
At the same time, governments began to apply to the police the systems of measurement and monitoring they had applied to education and health. All these measures dealt with the response side of police work: speed of action on 999 calls, for example, or the success rate in solving burglaries. What gets measured gets done. What was not successfully measured was public perception, both of personal safety and of the quality of police service.
The Nineties saw a period of long decline in British policing, particularly in London. Although there were notable successes, the decade was characterised by a decline in public confidence, by a significant decline in police numbers and then by the excoriating report on the racist murder of Stephen Lawrence. A recently published report by Mike Hough and Marian FitzGerald, two well-known criminologists, suggests a police force in London that is seen both by its public and by its own staff as increasingly remote and less visible and accessible. So when the street-crime-met-carjacking-met-homicide furore began at the end of last year, it is no wonder that commentators turned to New York.
Bill Bratton was commissioner of the New York Police Department from 1994 to 1996. As the alleged inventor of "zero tolerance" policing and as the undoubted architect of the extraordinary reduction in crime in New York, Bratton is one of the most famous police officers in history. During his period in office, murders fell from nearly 2,500 a year to fewer than 1,000, shootings from about 6,000 to fewer than 3,000, robberies from 85,000 to barely 50,000. In his last year in office, New York had 200,000 fewer crime victims than in his first year.
Could these startling results be translated across the Atlantic? There were five factors at work in Bratton's success. First, he had support across the board. It is worth remembering how bad New York had got: violence had made the city's subways unusable, open drugs markets were everywhere, and the murder rate was higher than in any comparable centre of finance and tourism, anywhere in the world. (London, for instance, has about 200 homicides a year.) People were afraid, very afraid. As a result, Bratton had complete political support from the mayor, Rudolph Giuliani, who had powers far in excess of what is available in London. Whatever Bratton wanted, Giuliani could and did give him. This included numbers. Bratton inherited about 27,000 police officers: by the time he left there were 42,000 - partly, it should be said, because Giuliani's predecessor, David Dinkins, had pushed through the tax rises to pay for them. Equally important, normally liberal New Yorkers and their media were prepared for policing tactics that would not be acceptable in London or anywhere else in the UK. Zero tolerance is a misnomer: it actually means targeted intolerance, and New York's finest did not always tread lightly.
Second, prosecutors in the US have to face election. If the mayor and the police chief were for rounding up the bad guys, then so were the district attorneys. With the different checks and balances in the British system, it is impossible to imagine lawyers of any sort, let alone judges, seeing a war on crime as part of their job in the way that their US counterparts did.
Third, the management of the NYPD was wide open for improvement: it had lines of accountability running in all sorts of directions and such a fear of corruption - the legacy of its previous history - that little priority was given to fighting crime. This appalling heritage allowed Bratton to turn the force upside down. Most notably, he introduced CompStat, a computerised crime statistics system, and quasi-theatrical weekly meetings at which he or his equally fiercesome first deputy commissioner ruthlessly examined precinct commanders in front of their peers about crime rates.
Fourth, Bratton believed the "broken windows" theory of James Wilson and George Kelling, who suggested that there is a causal link between tolerance of damage to the environment and the incidence of crime. They argued that if a community tolerated graffiti, vandalism, open drug dealing and aggressive begging, criminals would eventually believe that they were exempt from interference by authority. In his previous job as head of the New York transit police, Bratton had applied the same thesis, targeting fare dodgers on the grounds that people who wished to rob and murder on the subways were unlikely to pay for their tickets.
The final piece of Bratton's approach - and the one most often missed by commentators - is the self-belief that he engendered in his force. On his very first day in office, he made clear that the NYPD could and would reduce crime. He relentlessly pursued that idea, tearing up the previously received wisdom that only a change in social conditions would have an effect on crime.
Like everyone else, Londoners were stunned by the events of 11 September. On 11 and 12 September, crime fell markedly. By the evening of 11 September, however, the Metropolitan Police began to transfer hundreds of officers - at its height 1,500 - from outer London to the city centre. On 13 September, street crime began to rise in the outer boroughs and go down in the centre. By Christmas, the rise in street crime everywhere except central Westminster had become almost vertical, with nearly 7,000 robberies and snatch thefts in January this year, an increase of almost 54 per cent on the same month in 2001.
Yet that trend has been reversed, with fewer than 5,000 such crimes in June 2002, 14 per cent below the level of the year before. That pattern of decline in street crime was repeated across most major cities in Britain. It was, in microcosm, the New York story. Following the Prime Minister's focus on street crime, big city forces had redeployed hundreds and thousands of officers. The numbers matter. But these numbers are small in comparison with what is needed to restore public confidence.
This is where we must return to Bratton and to Wilson and Kelling. Further research into the "broken windows" thesis has shown a specific correlation between incivilities and disorder and street robbery in particular: successful street robbers seek out and hide behind a disorderly habitat, and that habitat is best combatted by an increase in police patrols.
If the answer to driving down crime is more police, then the answer must be to recruit them. The Met is growing by about 1,000 a year, which, with wastage, means about 2,500 new staff, probably about as many as it can absorb. But unlike New York, London and other big cities in Britain have large, powerful boroughs with tax-raising powers. These boroughs do not trust the police very much because, they almost invariably complain, the money they give to them is not reflected in visibly increased policing strength in their own area. Moreover, people do not care that the Met polices a city with the same population as the NYPD but with twice the geographic area, undertaking not only the functions of the NYPD but also of the FBI and the US Secret Service, with 14,000 fewer officers and a budget £1bn smaller. What they care about is that London feels less safe and they want something done about it.
The law contains an interesting lacuna: police forces are heavily constrained by numerous pieces of legislation, but there is no law preventing people setting up their own police force if they wish. A number of boroughs are thinking of doing just that: to patrol their streets, as well as their parks and housing estates.
Thus, just as the police service is finally waking up to the value of patrol, it is possible that we may see hundreds of different law-enforcement agencies springing up across Britain. We could return to the period before the first professional police in 1839 - it could be described as the Balkanisation of policing.
There is therefore no time to grow police numbers sufficiently and thus the police - and the Met in particular - have cajoled, pleaded and persuaded anyone who will listen, but particularly the legislators, that what is now needed is a new breed of police employee. The Police Reform Act, which received royal assent in July, has provided just this: a new form of patrolling officer called a police community support officer.
The first examples will arrive in central London in a few days. Their powers will be limited: to hand out fixed-penalty tickets for minor disorder. But because they cannot make arrests, do not need extensive training, and will not give evidence in court, they can just be deployed as a police presence wherever the local community decides. They will provide a permanent, visible patrol service, preventing the low-level disorder that leads to more serious crime. There will be six pilot sites around the country and, in the Met, plans are developing to franchise this service, allowing local authorities to buy it for deployment within their boundaries.
Thousands have applied for this new job. Many are people who, although physically and mentally well suited, say they do not want to be police officers because it is too long a commitment, too dangerous and too responsible. They want to patrol local communities and see that as a career in itself.
The police may thus have stumbled on something quite extraordinary. One of the ironies of police work is that, although everyone praises the bobby on the beat, most young people prefer something more glamorous. Just as in medicine, where everyone knows that public health provides the great advances but the Nobel prizes go to interventionist surgery, so the great reputations in policing are those of career detectives. Here, society may at last have found people who are genuinely committed to the police equivalent of public health.
And how we need them. We are not so far from a situation where community safety becomes a commodity to be bought and sold, with myriad different enforcement agencies holding different powers and with local authorities, residents' associations and even private individuals buying patrol services for their local streets and houses. In Miami, for instance, large parts of the city take the form of gated communities not accessible to the public police except in an emergency. The ghettos of affluence in South America and South Africa show what can happen to communities where the rich can buy security and the public police are reduced to responding to incidents in sink estates.
In 1994, the criminologists David Bailey and Clifford Shearing wrote: "Modern democratic countries like the United States, Britain and Canada have reached a watershed in the evolution of their systems of crime control and law enforcement. Future generations will look back on our era when one system of policing ended and another took its place."
That new system may well be something like what arrives on the streets of London this month. Although the parallels with nursing auxiliaries and teaching assistants are obvious, the advent of these new patrolling officers is a revolution in British policing. We may perhaps be comforted that Bill Bratton himself has written to Police Review backing their introduction.
Ian Blair is deputy commissioner of the Metropolitan Police