Across the country, a new movement is taking root, backed by the Obama administration, that is trying bold and controversial new methods – a kind of shock therapy – to fix the nation's worst schools. These are the bottom 5 percent, the roughly 5,000 public schools that chronically underperform and that, in many cases, society has given up on.
They are the schools that produce the worst test scores, suffer the worst dropout rates, yield some of the worst violence in the hallways.
Turning them around could help save successive generations of kids who quit and often end up jobless, mired in poverty, or worse. It could also dramatically improve the nation's educational performance.
Experts say that fixing even a fraction of these schools would lift the nation's test scores and education rankings, since the bottom-tier schools so depress overall performance.
Yet achieving this won't be easy. These schools have resisted the best intentions and ideas of educational gurus for decades. Now comes a new effort, led by President Obama's reform-minded secretary of Education, Arne Duncan, who vows to turn around 1,000 of the schools over the next five years.
He's putting $4 billion of federal money into the quest, and the methods his department is backing aren't incremental. They range from revamping a school from scratch, with virtually all new teachers, to installing a new principal who will carry out major reforms.
It's an ambitious goal – and a risky one. The record on dramatically improving the worst-performing schools in the country that, like Phillips, seem mired in failure is dismal: One recent study put the success rate at about 1 percent.
It's also controversial. Teacher unions and those on the left worry about replacing most of a school's staff. Many on the right see it as a waste of taxpayer money for something that won't work anyway.
But advocates argue that there's no choice without giving up on the futures of a large chunk of the nation's students. And they claim that done right, turnarounds have a far better chance of succeeding than the record would indicate.