A Case for Merit Pay: It’s Easier to Identify Good Teachers than to Train Them
Florida Governor Charlie Crist vetoed a merit pay plan passed by the state legislature. Rumor has it that he did so in order to gain teacher union support for his independent campaign for a U. S. Senate seat. But is there a substantive case to be made for keeping the current system of recruiting and compensating teachers instead of adopting innovations like merit pay?
Decades ago progressive reformers persuaded states and districts to put school boards under tight controls when they hired and paid teachers–on the theory that they would otherwise appoint their friends and neighbors to the job and pay them according to their political connections. Teachers could be hired only if they had a state-certified teaching license earned at a school of education. After a few years, teachers were given tenure and could be removed only by means of a complex, quasi-legal process. Salaries were based solely on their academic credentials and teaching experience, not according to the whim of any supervisor.
Now we are learning that almost every one of these decisions was wrong-headed. A just-released paper prepared by Matthew M. Chingos and myself for a Harvard Program on Education Policy and Governance conference on merit pay (which will take place on June 3-4) shows that Florida teachers who majored in education in college are no better at teaching math and reading to elementary- and middle-school students than those who did not. Even those teachers who attended the most selective universities in Florida, such as the state’s flagship university, the University of Florida, are no better at lifting student performance in reading and math than those who attended Florida’s less prestigious institutions. Nor can we identify any benefit from earning a master’s degree, despite the fact that school districts tend to pay 8 percent to 10 percent more to teachers who hold such a degree.
We do show that teachers become more effective in the classroom over the first ten years of their teaching experience, but (except in elementary reading) we show no further gains—and some declines–in effectiveness as teachers enter their second decade of teaching. Yet salaries of teachers often climb steeply in their second decade of teaching, as I noted in a blog post last week.
Our results are based upon an analysis of student-level data provided by the Florida Department of Education covering students in grades four through eight who took the state reading and math tests between 2002 and 2007. Teacher effectiveness was measured by their impact on their students’ test scores, taking into account a wide variety of student, classroom, school, and, where appropriate, teacher characteristics.
The one bright spot in our data set is the finding that teachers who are certified by the National Board for Professional Teaching Standards are, in fact, better teachers. The board asks teachers to submit evidence of quality teaching. After a careful review of their work, they certify about half the applicants.
The National Board certifies teachers after they have proven their effectiveness. That they do a pretty good job of identifying effective teachers suggests that picking good teachers is a lot easier than training them. It’s not clear that Florida’s governor Charlie Crist had a good substantive reason to veto the merit pay plan the state legislature passed.