Teacher Opposition to Merit Pay: It’s More Than Risk Aversion

In Risky Business: An Analysis of Teacher Risk Preferences, Daniel Bowen et al. of the University of Arkansas test the hypothesis that teachers oppose merit pay systems because they are more risk-averse. (A free, earlier version of the working paper is available here.)

The authors used a lottery-based risk elicitation tool to measure participants’ risk attitudes. The tool asks participants to make choices in a series of 10 lotteries with different real dollar payouts; risk attitude is estimated based on their choices. The authors compared the risk attitudes of graduate-level, pre-service teachers at a major public university to graduate students in MBA and JD programs at the same university. They find that pre-service teachers are indeed more risk-averse than their peers seeking an MBA or JD. Interestingly, however, the study finds no correlation between participants’ risk attitudes and their preferences about teacher pay systems – in fact, they find strong overall support for individual merit pay systems from all participants.

The researchers suggest a number of explanations for this finding, including that pre-service teachers overestimate their ability to be high-performing and that new teachers earn the lowest salaries on traditional step-and-lane pay schedules, making the prospect of earning more enticing.

However, I think the authors overlooked another possibility: Merit pay systems sound good in theory, but not always in practice. Perhaps pre-service teachers like the idea of merit pay precisely because they are pre-service. They have not personally experienced merit pay systems and thus tend to think about them theoretically. Many current teachers, on the other hand, have personal experience that shapes their opinions and forces them to think about merit pay systems more practically.

I taught for five years and have seen a few really bad teachers – like showing-movies-every-day-while-sleeping-in-the-back-of-the-classroom bad. This particular teacher should have been fired, and the school certainly needed an avenue to do so. But what about the well-meaning teacher who spends countless evening and weekend hours visiting students’ homes, grading papers, and planning lessons, not to mention spending her own money on classroom supplies? Most would guess she is a good teacher, so when VAM data say the opposite – that she’s ineffective – it’s a hard pill for many administrators, families, students, and other teachers to swallow. (Stories like this probably exist in every district, to some degree.)

Ultimately, I don’t think it’s the element of risk that drives teachers to oppose merit pay systems. It’s how these systems are designed and play out in practice – how they define and measure “good” teaching, and how this aligns (or doesn’t) with the culture and understanding of good teaching in a particular school or district. When there is alignment between the system and the culture of the school, really good things can happen. But when there’s misalignment, tension and frustration can boil over. In some cases, it’s the culture of the school that needs to change (or be replaced). But for the vast majority of schools, the “right” answer has to lie somewhere in the middle. This means that policymakers cannot come to the table with a “right” answer for how to evaluate and compensate teachers; instead, they must enlist teachers in creating a common understanding and definition of good teaching, and build the system from there.