The Economics and Politics of Teacher Merit Pay

Editor's note: Andrew Leigh is the ALP Federal Member for Frazer in the ACT and prior to this had a high profile as an educational economist and blogger

The ' logic of merit pay schemes is described broadly as follows:

• Teachers differ markedly in their effectiveness
• Differences in teacher effectiveness has a direct and significant impact on student learning
• Current pay differentials based on experience and qualifications have no basis in evidence (apart from the first five years)

Teacher pay increments are much flatter than comparative occupations

The arguments against it are summarised up as follows:

• Because of other influences on student outcomes merit pay is inherently unfair - and will therefore reduces morale and effort;
• They will cause teachers to focus on a narrow range of activities - teach to the test (Campbell's Law);
• If assessed by principals it will lead to favouritism;
• It goes against teacher collaboration - on this matter, Leigh comments that compared to some other occupations, teaching involves relatively little team work. However, there are still instances in which teachers assist one another, through team-teaching, sharing classroom notes, or providing mentoring and advice in the staffroom. (Editor's note: This suggests that like most economist who confidently promote economic solutions for educators, Leigh does not understand this important issue); and
• Group based schemes will encourage free riders

Leigh also looks at the complex issue of assessment in merit pay schemes - noting that while there is a general tendency to agree that multiple measures are important most schemes using multiple measures end up being completely incomprehensible to anyone but the designers and thus lose credibility.

He offers the following example

"Our value-added measure was based on students' year-to-year growth on TCAP. To control for the possibility that students at different points in the distribution of scores are likely to make different gains, we benchmarked each student's gain against the average gain, statewide, of all students taking the same test with the same prior year score. Benchmarking was simple: we subtracted the statewide average gain from a student's own gain to find out by how much his growth had exceeded the state average. Finally, we averaged these benchmarked scores over a teacher's class-more precisely, over students continuously enrolled in the teacher's class from the twentieth day of the school year to the spring TCAP administration, and for whom we had the prior year scores needed for benchmarking. This average was the value-added score used to determine whether the teacher qualified for a bonus.'

He concludes that, if current teachers do not understand how a merit pay plan works, they are unlikely to alter their behavior in the same way that they would if they understood it perfectly. And if potential teachers do not understand merit pay, it is unlikely to affect their selection into the profession.

Leigh argues for more research on merit pay - particularly on whether it affects collegiality and opt-in schemes. 

Leigh concludes his presentation with ten questions all of which assumes that merit pay should still be on the agenda - we just need to finesse it more. 

Read More: http://www.andrewleigh.com/blog/?p=1874 and http://www.andrewleigh.com/blog/wp-content/uploads/2011/11/PerformancePay.pdf