One of the solutions often proposed for schools in which students perform poorly is closing down the school. It's one of the four options required for schools to receive School Improvement Grants in the current administration and has been an option for consistently low-achieving schools under No Child Left Behind. The Senate HELP Committee's proposal for reauthorizing ESEA maintains school closure among seven options for persistently low-achieving schools.
"Shut it down" sounds like a logical, if extreme, option when all else has failed, but a study by John Engberg from RAND and his colleagues presented some disturbing data about school closure. They found that students in schools that are closed due to poor performance actually do substantially worse on reading and math tests in the new school to which they are sent for at least a year, and then recover and end up doing about as well as they were doing at their original school. In other words, after all the expense, acrimony, and heartache involved in closing a school, the students involved do not benefit.
This does not mean that schools should never be closed. Schools often have to be closed due to declining populations or economic factors. Sometimes a school has such a dysfunctional environment or bad reputation that it needs to closed, and every once in a while closing a school might impress other low-performing schools, in the sense that Voltaire suggested that it is sometimes good to execute a (losing) admiral "to encourage the others."
Yet the Engberg et al. findings caution those who want to use school closure broadly. A school building does not cause low achievement. Bringing new leaders and new staff, and new programs with strong evidence of effectiveness, seem more likely to benefit struggling schools.