Merit pay for teachers

Harvard economist Ed Glaeser has an NYT Economix blog post today on the controversial subject of merit pay for teachers.  Glaeser cites a number of studies that show that teachers do respond to incentives, which is hardly surprising.  The trick, Glaeser notes, is managing the incentive program.  In some states and school districts, the testing bar is set so low that virtually all teachers qualify for the incentive pay.  Readers of Steve Levitt's Freakonomics recall the study he did of Chicago teachers who helped their students cheat on exams so that the teachers would get their bonuses.  An important lesson from the economics of incentives is that people will try to game the system.

Glaeser suggests that instead of micromanaging the performance of individual teachers in relationship to test scores, incentive schemes in public schools should focus on school-wide outcomes and principals be held accountable.  I am not sure this is very realistic either for two reasons: many principals are not prepared to take on this responsibility (neither are most bosses in the private sector) and in most school districts unions will not let them have the flexibility to make changes in staffing.