Some of New York City's public schools now house "sweep rooms" for students cutting class, "holding cells" in the basement, and dozens of police and security guards. Unnoticed by many, these new additions are part of an unprecedented expansion of the criminal justice system into the educational domain.

In doing this, Mayor Michael Bloomberg has received inspiration from his predecessor. He has adopted the same "broken windows" theory that shaped former Mayor Rudolph Giuliani's approach to fighting crime in New York City neighborhoods. The theory holds that minor signs of disorder, such as a single broken window, send a signal that no one cares, creating an opening for more -- and more serious -- offenses. So, to prevent crime you fix the broken window.

Thus, the Impact Schools Initiative, which takes its name from another police department program, Operation Impact, that put more police officers into high crime areas. Impact Schools is a joint effort by the police department, the Department of Education, and the mayor's office. Launched in January 2004, it targets middle and high schools with high levels of reported crime, and employs aggressive policing strategies to reduce violence and disorder.

But in light of new data profiling the Impact Schools, one must ask, "Which came first -- broken windows or broken schools?"

Impact Schools have more economically disadvantaged students and more black students than other city schools. They also tend to be more overcrowded, and less well funded than schools throughout the city, according to new research (in pdf format) by the Drum Major Institute for Public Policy. These problems existed before the launch of the mayor's Impact Schools initiative; only time will tell if conditions are changing. Looking back, we have to ask ourselves whether these schools had already been crying out for special attention -- not in the form of cops, security cameras, and "sweep rooms", but teachers, books, and computer rooms.

The facts speak for themselves. According to Drum Major Institute, the current 22 Impact Schools are operating at 111 percent capacity -– making them five percent more crowded than the average New York City public high school and seven percent more overcrowded than schools the Department of Education describes as similar. The worst of the Impact Schools, Walton High School and Christopher Columbus High, both in the Bronx, were at more than 180 percent capacity in 2004.

The city spent an average of $1,482 less for each student at an Impact School than it did at the average city high school in 2003. Furthermore, while the city increased spending on direct services in high schools by $1,217 per student between 2002 and 2003, average spending on students at the 22 Impact Schools rose by only an average of $609.

More than six in ten students at the 22 Impact Schools serving the lowest income families are poor enough for a free lunch (a benefit available to families whose incomes are up to 130 percent of the poverty line). This compares with about five in ten at the average city high school. At the six poorest Impact Schools, more than eight out of ten students are eligible for free lunch. At Walton High School -- also among the most crowded of the Impact Schools -- that number reaches nine in ten.

Additionally, the Impact Schools have a higher percentage of students of color and significantly fewer white students than the average city high school. More than half of students at the Impact Schools are black, compared with 35 percent in the average city high school. And there is also a slightly greater proportion of Hispanic students in the Impact Schools. At the same time, Impact Schools have a substantially lower share of white students -- 4.6 percent, as opposed to 14.2 percent.

Finally, ninth and tenth graders at Impact Schools were far more likely than their counterparts elsewhere in the city to be over age for their grade. By the time they reach high school, these older students, many of whom have had a difficult time in school, require additional resources to improve their academic performance. Getting extra help is made all the more difficult when a school is packed to the rafters and stretched to the limit, as is the case in the 22 Impact Schools.

Mayor Bloomberg seems eager to point to numbers showing crime falling at the Impact Schools this year over last.

But a recent analysis by the National Crime Victimization Survey has shown that putting police officers in schools does not stem violence. In fact, the opposite seems true: Schools using security measures such as law enforcement agents and surveillance cameras were associated with increased reports of school violence and disorder, while schools relying on more participatory efforts to educate youth on school rules and appropriate conduct were associated with less reported school crime.

The city has chosen to respond to the headline-grabbing visible symptoms of school disorder rather than treating the root causes, a complex social and education problem. If the city took a more thoughtful look at the 22 Impact Schools it would find they have much more in common than high rates of crime. Addressing those problems, and bringing the chronically resource-poor and overburdened Impact Schools up to the level of the rest of city schools, could go a long way toward making those schools safer, better places of learning.

The question for New York City now is not whether we fix the broken windows, but how?

Rhee, a fellow of the Drum Major Institute for Public Policy, is executive director of the Prison Moratorium Project