The Effects of School Closure Threats on Student Performance: Evidence from a Natural Experiment

We exploit a natural experiment to test whether school closure threats can increase staff effort and improve performance. The Hong Kong government overestimated post-1997-Handover mainland Chinese immigration and local births, creating excess capacity in many school districts. In 2003, with student enrollment falling to new lows (increasing cost per student), the government announced that it would close schools that were unable to recruit enough students. Since schools compete for students within their own district, the accidental excess capacity created closure threats that varied by district. Difference-in-differences analyses show that after initiation of this policy, student scores in heavily overbuilt districts were lower than scores in other districts and lowest in districts with the fewest students per class. Although closure threats were counterproductive for improving performance, the school closures eventually improved overall school quality, as typically, the lowest performing schools in each district closed.