Philadelphia's public charter schools have been under fire lately. First were revelations last spring that two of the city's 60 charter operators paid themselves outlandish salaries. More recently, the Philadelphia School District released a report claiming that charters cost the district $105 million annually.

The writing is on the wall: The Philadelphia School District has declared war on charter schools. That's a shame, because the evidence shows that charters are the only public schools without selective admissions policies that can teach hard-to-reach urban adolescents.

Even when spending as much as suburban schools, traditional urban public schools fail to educate students. They are hampered by crime, racial divisions and entrenched bureaucracies. As Frederick Hess has documented, big-city school superintendents serve "dog years": Their tenures tend to be poor, nasty, brutish and short.

While many city school superintendents hang on to their posts by handing out patronage to supporters, only three managed to improve learning: Alonzo Crim (Atlanta, 1973-88), Richard Wallace (Pittsburgh, 1980-92), and Rod Paige (Houston, 1994-2001). But even these three failed sixth grade: Their middle and high schools remained horrible.

Why are urban middle and high schools so hard to change? The difference is that urban elementary schools have agreed-upon goals: teaching kids how to read, write, and do basic math in a safe environment. With good leadership, these are things that bureaucracies can do.

By contrast, in large, impersonal middle and high schools, at-risk children focus on their peers instead of their teachers. The pressure to fit in by rejecting academics becomes most damaging at this point, as Stuart Buck writes in a forthcoming book.

But it isn't just the kids. Like college professors, middle- and high-school teachers have bigger salaries and egos than their elementary-school counterparts. They focus more on their prerogatives and less on the kids. And unclear academic goals at the middle- and high-school levels leave teachers free to push their own agendas.

The only solution is to completely remake urban middle and high schools. We must replace Philadelphia's traditional public schools with public charter schools.

Unlike huge, traditional urban high schools, Philadelphia's charter schools are small learning communities chosen by their students and faculties. They cost less than traditional public schools and, more important, work better for kids.

The academic data provided by No Child Left Behind show that Philadelphia's charter high schools achieve 95 percent of their academic goals, compared with 76 percent for district-run high schools. While 36 percent of Philadelphia's neighborhood high schools are designated "persistently dangerous," none of the charter schools is. And charter high schools boast a 94.5-percent graduation rate, compared with 49.6 percent for Philadelphia public schools.

The data suggest we can't afford not to open more charter schools - and close down traditional high schools.

Robert Maranto is a former Villanova professor. He can be contacted at rmaranto@uark.edu.