The Magnificently Complex World of Teacher Pay & Other Incentives

Perhaps the most important message from the teacher merit pay study, released this week by the National Center on Performance Incentives at Vanderbilt University, came in a short, four-sentence paragraph.  On the last page of the summary section, the researchers conclude:

The implications of these negative findings should not be overstated. That POINT did not have a strong and lasting effect on student achievement does not automatically mean another approach to performance pay would not be successful. It might be more productive to reward teachers in teams or to combine incentives with coaching or professional development. However, our experience with POINT underscores the importance of putting such alternatives to the test.

Stated another way: Designing effective incentives for teachers is a mighty complex task, full of many subtle decisions and much uncertainty.  This study, like the ones before it, and the ones that will surely follow it, is highly inconclusive.  It is merely one experiment out of an endless number of possible experiments that could have been conducted.

Following a logical procession of incentive design questions reveals just how many choices the person or group configuring the incentive system must make.  Altering any one decision can greatly impact teachers’ attitudes towards the incentive system, how the incentives interact with other teacher incentives, and how the incentives end up influencing teacher behavior.  Each choice made in the design process affects other choices.  Changing just one piece of an incentive system’s design creates an entirely new system.

Most people who design incentive systems (e.g., managers, researchers, and teachers themselves) begin by determining what end outcome they hope to achieve, and then deciding how they can use the resources they have to influence individuals or groups to accomplish that end aim.  Here is a series of questions that architects of teacher pay incentives usually must answer:

What is my end aim? Am I trying to encourage current teachers to better teach to a set of standards and raise student achievement?  Am I trying to attract people with different characteristics and values into the teaching profession?  Or do I wish to create a more just compensation model for teachers?

Should I reward or punish teachers to change their behaviors? Of the resources I control, is it best for teachers to look to the promise of gaining something if they act in desirable ways, or is it more effective for teachers to fear losing something if they act in undesirable ways?

What resources will I use? Compensation, including both pay and benefits, comes in several forms.  Many teacher pay for performance structures use money as the primary incentive, mostly because it can be quantitatively measured and because it is assumed to be universally attractive.  But would it better to offer teachers prizes (like paid vacations to exotic resorts) or better working conditions (like more flexible schedules or advancements in position and authority)?  What if teachers with stellar test scores could work less hours or days at the same compensation level?

With what magnitude and frequency should the incentive be offered? That is, how much of the resource must I offer to or take away from teachers to alter their practice?  How immediate must the reward or punishment be to effectively change behaviors?

What should I use to determine how much teachers are rewarded or punished? Should I consider the percent of students who meet “proficiency” on a state standardized test?  Should I rely on evaluations from principals and other top-notch teachers?  Should I look at growth models and “value-added” models?  Most importantly, how do I ensure that the system of determination is carried out with integrity?

What other teacher incentives (and disincentives) are operating against mine? Do principals, teachers unions, parents, and colleagues place opposing demands on teachers that undermine the larger goals of the system?  Are other incentives stronger than the ones I have implemented?  How do I ensure that my incentives aren’t combining with other incentives to form strange, unintended consequences?

How do I decide if my incentive system is working? Sometimes end aims may be vague or difficult to assess.  What tools and standards should be used to determine the efficacy of incentive systems?  When will I know that my end aim has been reached?

This series of questions represents just a fraction of the design questions that have to be answered by those who are creating teacher incentive systems, including pay for performance schemes.  As creative new methods are devised for rewarding and punishing teachers towards a set of clear, achievable goals, much more research and analysis will be needed to determine what makes some systems more successful than others.  How design decisions are made and how those decisions influence the outcomes of other decisions matter tremendously.

In all of the magnificent complexity surrounding teacher merit pay and incentives, there is one obvious observation made by most people in education: The incentive and pay systems that are currently in place for most traditional public school teachers in this nation are not ideal.  Pay based entirely on years of experience encourages people to stay in the classroom, and not much more.  This observation alone justifies the bold quest by school leaders and education researchers to find something new, something workable that will finally give teachers (and potential teachers) a renewed purpose for ensuring an outstanding education for all children—including the most disadvantaged students.

Tagged as: Education Research, Merit Pay, National Center on Performance Incentives, Pay for Performance, Teacher Incentives