When you look at the history of homicides over the past 20 years, the results are staggering. To mention a few examples, homicides in New York City have dropped well below 500 per year, compared to more than 2,200 in 1990. In Chicago, which had more than 900 murders a year in the 1990s, that number has been cut in half. Minneapolis recorded 19 murders last year-one-fifth of its record of 97 killings in 1995. In Washington, D.C., there were 140 homicides in 2009, compared to 479 in 1991. And in Philadelphia, there were 87 fewer homicides in 2009 than just two years before-a 22-percent reduction.

We have seen a sea change in how the police define their mission. There was a time when the conventional thinking was that no matter what the police did, it made no difference, and police were not held accountable for increases in crime. That started to change in 1982 when George L. Kelling and James Q. Wilson wrote an article that appeared in The Atlantic, called "Broken Windows." Their idea, that broken windows and other signs of disorder in a neighborhood can contribute to crime, prompted police to start asking themselves what they could do to fix broken-window problems and prevent crime, as opposed to merely investigating crimes after they occur.

Today, thousands of local police departments have adopted this thinking, and they are constantly searching for ways to prevent crime.

How can police prevent crimes from happening? The strategies began with things like Compstat and problem-oriented policing. Community policing is another major change-today's police understand that they can't do it alone, so they make tremendous efforts to reach out to their communities and work together on crime reduction.

Broken Windows policing, Compstat, problem-solving, and community policing have become almost universal in American police departments. Today, the story is the extraordinary proliferation and refinement of ideas to prevent crimes from happening. Police departments across the country are developing new solutions to local problems, and spreading the word to others when they find something that works.

Just a few examples:

• In Los Angeles, most homicides are gang-related, and one murder used to trigger a bloodbath of retaliatory killings. The LAPD attacked that problem by making it their business to know everything about gang rivalries. Police Chief Charlie Beck has told his detectives that when there is a gang killing, "We not only want to know who committed this homicide, but what we can do to prevent the next one."

In San Francisco, the police are focusing on the "10-percenters"-the hard-core 10 percent of criminals who commit most of the crime in the city.

In Minneapolis, the police realized that juvenile offenders were committing a large share of the city's serious crimes, and they were simply falling through cracks in the city's juvenile justice system. So the police helped develop a comprehensive new approach that helps youths get back on track, but also stops the serious repeat offenders.

In Milwaukee, police knew that a particular gun store was selling huge numbers of guns that ended up being used to commit crimes, including the shootings of six police officers. So they staked out the store and watched for signs of "straw purchases" in which felons had other people buy guns for them. The result: 23 arrests in 15 weeks.

And Philadelphia is attacking the problem of domestic violence homicides, which have traditionally been considered very difficult to prevent. The city's new approach involves creating systems to ensure that repeated calls to the police from a certain address and other warning signs are tracked efficiently, so officers will be able to recognize the red flags of a potential domestic homicide before it is committed.

In city after city, police are working to figure out the "who, what, when, where, how, and why" of homicides and other violent crimes, and then they try to break up the patterns. High levels of crime are no longer seen as inevitable. Because police are able to use real-time crime information, patterns of crime that once took months or years to detect are now identified on a daily basis. The status quo is never acceptable, because unless homicides and other crimes go down to zero, there is always room for improvement. And when crime rates started moving in the wrong direction, as they did in 2005 and 2006, police redoubled their efforts and in some cases changed strategies, with a sense of urgency that was unprecedented, and they reversed the trend by 2007.

It is becoming abundantly clear that violent crime is dropping because local police, working collaboratively with their communities, have turned on its head the notion that crime is inevitable.

The Police Executive Research Forum (PERF) is a national membership organization of progressive police executives dedicated to improving policing and advancing professionalism through research and involvement in public policy debate. PERF is a member of the Board of Directors of the National Law Enforcement Officers Memorial Fund.