The man behind this notion, Jack Maple, is a dandy who wears dark glasses, homburgs and two-toe shoes; yet he has become something of a legend in America’s police departments. For some years, starting in New York and moving on to high-crime spots such as New Orleans and Philadelphia, he and his business partner, John Linder have marketed a two-tier system for cutting crime. First, police departments have to sort themselves out. root out corruption, streamline their bureaucracy, and make more contact with the public. Second, they have to adopt a computer system called Comsat which helps them to analyze statistics of all major crimes. These are constantly keyed into the computer, which then displays where and when they have occurred on a color-coded map, enabling the police to monitor crime trends as they happen and to spot high-crime areas. In New York, Comsat’s statistical maps are analyzed each week at a meeting of the city’s police chief and precinct captains. Messrs Maple and Linder ("specialists in crime-reduction services") have no doubt that their system is a main contributor to the drop in crime. When they introduced it in New Orleans in January 1997, violent crime dropped by 22% in a year; when they merely started working informally with the police department in Newark, New Jersey, violent crime fell by 13%. Police departments are now lining up to pay as much as $ 50,000 a month for these two men to put them straight. Probably all these new policies and bits of technical wizardry, added together, have made a big difference to crime. But there remain anomalies that cannot be explained, such as the fact that crime in Washington D. C. , has fallen as fast as anywhere, although the police department has been corrupt and hopeless and, in large stretches of the city, neither police nor residents seem disposed to fight the criminals in their midst. The more important reason for the fall in crime rates, many say, is a much less sophisticated one. It is a fact that crime rates have dropped as the imprisonment rate soared. In 1997 the national incarceration rate, at 645 per 100 000 people was more than double the rate in 1985, and the number of inmates in city and county jails rose by 9.4%, almost double its annual average increase since 1990. Surely some criminologists argue, one set of figures is the cause of the other. It is precise because more people are being sent to prison, they claim that crime rates are falling. A 1993 study by the National Academy of Sciences actually concluded that the tripling of the prison population between 1975 and 1989 had lowered violent crime by 10%～15%. Yet cause and effect may not be so obviously linked. To begin with, the sale and possession of drugs are not counted by the FBI in its crime index, which is limited to violent crimes and crimes against property. Yet drug offences account for more than a third of the recent increase in the number of those jailed; since 1980, the incarceration rate for drug arrests has increased by 1000%. And although about three-quarters of those going to prison for drug offences have committed other crimes as well, there is not yet a crystal-clear connection between filling the jails with drug-pushers and a decline in the rate of violent crime. Again, though national figures are suggestive, local ones diverge: the places where crime has dropped most sharply (such as New York City) are not always the places where incarceration has risen fastest.