Do charter schools work?

The Economist:

AMERICA’S universities are the best in the world, but the kindest verdict on its schools is “could do better”. It spends enough on them–around the rich-world average of 3.8% of GDP–but its pupils do poorly in tests of reading, writing and mathematics, and too many drop out before completing school. Teaching attracts few ambitious and able graduates; school leaders have little autonomy. The solution, to free-marketeers, seems obvious. Give taxpayers’ money not to a state-run monopoly, but to independent schools.
Since Minnesota started the experiment in 1991, most states have introduced independent, or charter, schools in some form. Evaluations have been broadly positive, but their enemies, including the politically powerful teachers’ unions, can fairly claim that more research is needed. Do charter schools’ pupils do better at tests because they have been coached intensively at the expense of a broad education? Do charters mean the most motivated students cluster in a few schools, to the detriment of the majority? Do they kick out–or coax out–the toughest to teach?