More evidence on teacher merit pay

Individual Teacher Incentives And Student Performance
by David N. Figlio, Lawrence Kenny 
This paper is the first to systematically document the relationship between individual teacher performance incentives and student achievement using United States data.  We combine data from the National Education Longitudinal Survey on schools, students, and their families with our own survey conducted in 2000 regarding the use of teacher incentives.  This survey on teacher incentives has unique data on frequency and magnitude of merit raises and bonuses, teacher evaluation, and teacher termination.  We find that test scores are higher in schools that offer individual financial incentives for good performance.  Moreover, the estimated relationship between the presence of merit pay in teacher compensation and student test scores is strongest in schools that may have the least parental oversight.  The association between teacher incentives and student performance could be due to better schools adopting teacher incentives or to teacher incentives eliciting more effort from teachers; it is impossible to rule out the former explanation with our cross sectional data.

In other words, schools with merit pay have better outcomes, but we can’t be sure there’s a causal effect. Another approach would be to find a natural experiment, in which some schools switch from uniform salary schedules to merit pay. Does anyone know of a good natural experiment?

This entry was posted in Economics of Education. Bookmark the permalink.

5 Responses to More evidence on teacher merit pay

  1. Bring Back EP at LP says:

    Andrew, this brings up the old problem of whether you can measure performance.

    If you merely take student scores then teachers at selective schools will be jumping for ‘performance pay’ whilst those at comprehensive schools would be completely cheesed off.

    Oh to be in a ‘private’ school where you can prevent any pupils that may have ‘problems’.

    The best papers I have seen you offer ( and thanks a lot for that) do not measure performance per se but where we get an indirect measure where there is a collective agreement on teaching standards.

  2. Michael Moriarty says:


    I recall this issue being raised last year. On tracking back to my original post I realised that I paraphrased some thing with out giving due credit. This is my opportunity to correct this.

    Aaron Sorkin no he is not a political leader, he was the Head Writer for a TV Show called the West Wing had this to say about schools:

    “Education is the silver bullet. Education is everything. We don’t need little changes. We need gigantic revolutionary changes. Schools should be palaces. Competition for the best teachers should be fierce. They should be getting six-figure salaries. Schools should be incredibly expensive for government and absolutely free of charge for its citizens, just like national defense. That is my position. I just haven’t figured out how to do it yet.”

    Spoken by Sam Seabour (AKA Rob Lowe) Hopefully that is giving full credit.

  3. Michael Moriarty says:

    That being said who is going to actually do this ? Well it is being done…well at least considered. This is well worth a read :


  4. Michael Moriarty says:

    Does anyone know of a good natural experiment?

    You probably have read this already. Interesting just the same:

    Click to access w9671.pdf

    Underdeveloped V Developing V Developed Countries would in my opinion all yeaild differing results.

Comments are closed.