Thursday, Feb. 19, 2009 | In all parts of society and in nature, competition produces better results while the lack of incentives leads to a lack of progress. This is the basis of the natural forces of evolution which have produced such a variety of successful organisms, and of the competitive market system which all successful nations use today. The question is not whether teachers should be paid based on their performance — of course they should — but how to implement pay for performance fairly and without destroying something more valuable in the process.

A reasonable way to do this is to base the merit pay portion of a teacher’s salary on a combination of student evaluations, test score improvements, and the evaluation of a “benevolent dictator.”

Student evaluations may seem at first to be a dubious way to evaluate a teacher’s merit, like setting a fox to guard the chicken coop or having the inmates run the asylum. However, evaluations are in wide use already, determining college professors’ promotion. For example, at UCSD, student evaluations are collected in all large courses, and they are the only measure of teaching merit used when a professor is being considered for promotion.

This gives students a lot of power, because academic rank is the major determinant of a professor’s salary. In fact, positive student evaluations are correlated with the amount and quality of actual learning. When taken into account along with the other factors described below, student evaluations are a reasonable starting place for measuring merit.

Test score improvements are the closest thing we have to an objective measure of teaching progress, and their only pitfalls are that test scores may be subject to vagaries not under the teacher’s control, such as changes in demographics, and that using test scores encourages ‘teaching to the test.’

By using an average of the previous several years as the baseline, and by recalibrating the scores if a significant shift in student demographics does occur, the first problem can be eliminated. And if the test measures something worth measuring, then teaching to the test is in fact a desirable outcome. These scores are also complemented by the student evaluations, since a teacher who only teaches to the test without, for example, going off on interesting tangents when appropriate, may get the high test scores he was looking for but low student evaluations.

A “benevolent dictatorship” system is one in which supervisory personnel such as department heads and principals have some measure of arbitrary control over merit pay. These supervisors can synthesize student evaluations and standardized test scores with other, harder to quantify, qualities. These qualities might include a teacher’s reputation and value in the local context, service to the school, curricular innovation, etc. — things someone on the scene would be best qualified to judge. Supervisory determination is used already in most corporations, as in the traditional model in which the corporation determines one’s salary but one’s boss determines the bonus.

The obvious pitfall here is that personality conflict or sycophancy can sway a supervisor’s opinion. The solution, which is in current implementation in the U.S. Department of Veterans Affairs, is having performance pay recommended by an immediate supervisor but acted upon by a more distant board. This provides a check and balance between two different “benevolent dictators,” helping to ensure benevolence and limit dictatorship.

These three metrics — student evaluations, test score improvements, and overall evaluation by supervisors — can be divided into approximately even thirds of the possible merit bonus, or the supervisor can adjust the proportional representations of the other two metrics within a reasonable range.

This would allow supervisors to account further for local conditions, such as good teachers in a failing school, or complacent teachers in a top school. A supervisor could designate student evaluations or test scores to be the dominant but not the only determinant of merit pay.

A question that vexes business leaders is how much of an employee’s pay should be based on merit, and how much should be ‘base pay,’ predicated on traditional non-merit factors such as seniority. The program outlined above could be suitable for somewhere around a quarter of a teacher’s total salary. Thus the traditional determinants of pay, namely seniority and responsibility, would continue to guide the majority of a teacher’s remuneration, but merit incentives would play a role too significant to be overlooked.

Each of the three measures of merit has its shortcomings, but most of these can be avoided through careful implementation. Combined, they form a system for measuring merit accurate enough to be put into immediate use in schools around the nation. What valuable feature of the present system would be lost by implementing such a system? Some argue that differing salaries among teachers who work together would cause bitterness and resentment, destroying the harmony and teamwork so necessary for running a school.

However, differing salary for differing work is already the standard in every sphere but public employment, because the increase in productivity and quality that results from a reasonable incentive program is of far greater importance than any discord it may cause. In fact, a successful teacher with high merit ratings and corresponding higher pay might just as likely be a model to aspire to as a target for jealousy. Thus the effects on workplace morale are uncertain but would probably be a net positive.

Merit pay for teachers is an obvious and long overdue policy. Each component of the system outlined here has been shown to work in various real-world environments where quality improvement rather than job or pay security is a significant goal. Such a system should be implemented and reviewed annually, so that our teachers can be paid something closer to what they deserve, and a culture of excellence can take root across the public school system.

Benjamin Cosman is a junior at La Jolla High School. His essay reached the finals of the 2009 Essay Contest. The other finalists’ pieces will run this week with the winner’s appearing Friday, Feb. 20.

Leave a comment

We expect all commenters to be constructive and civil. We reserve the right to delete comments without explanation. You are welcome to flag comments to us. You are welcome to submit an opinion piece for our editors to review.

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.