Job Security for Poor Performers

By Andrew G. Biggs 

Job security for public employees comes at a cost to government and the public. Firing the poorest-performing teachers may be the biggest free lunch available in education policy. In my work on public employee pay, one factor I’ve paid particular attention to is job security, the fact – and it is a fact – that public employees are significantly less likely to be fired or laid off than similar private sector workers. This effective insurance policy against unemployment has a value to workers. But sometimes it’s assumed that job security is a freebie that has no cost for government or the public. A new study from Thomas Dee of Stanford and James Wyckoff of the University of Virginia shows why that’s not true. Dee and Wyckoff look at Washington, D.C. public schools’ IMPACT program, which assessed teachers’ performance. High-performing teachers were given financial rewards, while low-performing teachers were threatened with dismissal. Dee and Wyckoff isolate the effects of the IMPACT program on teachers’ behavior by comparing teachers who almost received financial rewards with those who crossed the threshold and received the reward; similarly, they compared teachers whose performance ratings were almost poor enough to warrant a dismissal threat with teachers who received the dismissal warning. Dee and Wyckoff found that: •    Financial rewards given to high-performing teachers further improved their performance. One reason might be that D.C. teachers who maintained high performance ratings receive permanent pay increases in addition to their bonuses. •    Dismissal warnings for low-performing teachers had two effects: a 50 percent higher voluntary quit rate among those threatened with dismissal, but substantial performance improvements among those who remained on the job. In other words, strong job security for public employees keeps more poor performers on the job and reduces incentives for poor performers to improve. The threat of dismissal would induce poor teachers to either quit on their own or improve. This lends further strength to my blunt-instrument proposal to improve public education: fire the worst 5 percent of teachers and shift their students to other classrooms. Having a bad teacher makes a significant difference to quality of education, while class size matters very little: eliminating the poorest teachers would raise their students’ average lifetime earnings by around $250,000, while reducing school’s teacher costs by 5 percent. But Dee and Wyckoff’s results say the benefits would be even greater: the threat of dismissal would induce poor teachers to either quit on their own or improve; plus, teachers in, say, the 6th through 10th percentiles might improve in order to avoid a dismissal warning. Firing the poorest-performing teachers may be the biggest free lunch available in education policy. Given that very little else seems to have worked, maybe education policy makers should give it a try. Andrew G. Biggs is a resident scholar at the American Enterprise Institute.