Performance Bonuses Bring Uneven Payoffs for Teachers

By Kenneth J. Cooper
Washington Post Staff Writer

The Washington Post
Sunday , June 18, 2000 ; A02

A groundbreaking experiment with performance bonuses for teachers in Denver has shown uneven results for students and faculty in its first year--raising questions about an increasingly popular approach to lifting student achievement.

The four-year test is an important aspect of the growing trend of paying teachers at least partly based on their classroom performances. The bonuses reward good instructors and provide other teachers a financial incentive to work harder toward improving student scores.

Only in Denver, where teachers in 12 elementary schools volunteered to participate in the program, has "pay for performance" been based on test scores in individual classrooms, instead of entire schools.

At some of the Denver schools, all or nearly all the teachers met achievement goals negotiated with principals at the start of the school year, and received an extra $1,000 in their paychecks. Standardized test scores rose modestly, but the gains were not necessarily larger than in previous years.

But in at least one school, fewer than half the teachers qualified for the full bonus, causing hard feelings among those who didn't. They blamed a high turnover of students and the poor scores of a few.

Critics of altering the traditional salary structure, which is based primarily on seniority, worry that teachers whose students fail to improve because of social problems beyond a teacher's control will unfairly lose out on bonus money. They also are concerned that teachers competing for bonuses won't work together as a team. Instances of both were reported in Denver's trial run.

But the idea of transferring the concept to public school teachers has gained political support.

Vice President Gore has specifically endorsed the Denver experiment, while a spokeswoman for Texas Gov. George W. Bush said he would permit states to use federal funds for various forms of teacher compensation.

In the past year, the House and Senate have voted for versions of performance pay. Governors meeting at an education summit last fall also favored the practice.

About 20 school districts and states--from Boston and Dallas to Kentucky and California--pay bonuses to all teachers in a school if its test scores rise. This approach appeals to national teachers unions.

"There's a pretty good consensus to reward groups, teams or schools, rather than individuals, because so many individuals are responsible for increasing achievement in schools," said Sandra Feldman, president of the American Federation of Teachers.

The National Education Association, the larger of the two teacher unions, is scheduled to take a formal position on the issue next month at its annual convention. A study committee has recommended that it deem "group-performance incentive systems" as "worthy of experimentation."

More states and districts are considering some form of performance pay. Education leaders in New York, Los Angeles, Cincinnati, Iowa, Maryland, Prince George's County and elsewhere have advocated or are working toward teacher bonuses.

"This is popping up all over," said Allan Odden, an education professor at the University of Wisconsin who has been preaching the value of performance pay for a decade.

Besides incentive bonuses, Odden said, the concept includes "pay for knowledge and skills" based on a widely accepted theory that students learn more if their teacher has grown professionally.

Teachers who undergo rigorous certification by the National Board of Professional Teaching Standards, for instance, receive bonuses or salary increases in 25 states and 110 districts, including Virginia, Prince George's County and Montgomery County. Both national teachers unions support what some call "knowledge-based pay."

In Denver's experiment, four schools awarded bonuses if teachers took courses, read academic texts or developed new curriculum. The other eight schools either rated teachers based on student standardized test scores or other tests, including some developed by teachers. In all, teacher goals were based on 100 different tests.

Those goals varied by teacher but generally were aimed at raising scores by 5 to 10 percentage points or bringing lagging students up to grade level. The 350 teachers received $500 up front for participating in the experiment and an additional $500 for reaching each of two achievement goals.

The Denver school board is scheduled to review initial results of the experiment at its meeting Thursday.

At Colfax School on the city's far west side, 21 teachers reached both their goals and the other two met one, according to principal Mary Romero.

For Colfax's predominantly Mexican American students, the picture was mixed: In two grades, scores rose more than they had the year before, while in two others the increase was smaller than the previous year. In another grade, scores went down as much as they had a year earlier.

Not all of the school's teachers set goals to raise average scores. Some had specific goals for individual students.

Joyce Simmons, principal of Smith Renaissance School of the Arts in northeast Denver, sounded glum as she described the results at the mostly African American school. Scores increased, but not any more than they had in recent years. Fewer than half the teachers met both their goals, and some failed to accomplish even one, creating tensions among the staff.

"This year, it really has not worked in the way I wanted it to," Simmons said.

Esther LeMire was on a three-member team of first-grade teachers who came up short--because three students got low scores.

"Maybe if we had been better goal writers we would have met our goal, or if we had put our goal a little lower," said LeMire, the school's representative to the local teachers union.

Principals and teachers participating in the pilot project reported a number of fairness and technical issues.

Is it fair to judge teachers of pre-kindergartners and kindergartners, for whom no standardized tests exist, based on scores on informal assessments? How fair a measure are the test scores of bilingual students in the third grade, when they make a language transition and take a test in English for the first time? How do you set achievement goals for specialists who teach the arts or physical education? For support personnel such as social workers and librarians?

And there were some doubts that $1,000 was enough of an incentive to make a difference to teachers who earn an average of $42,000 a year.

"It was probably enough to motivate a bit of extra work," said Stephen Levin, a first-grade teacher who met his goals at Edison School. "I don't think it was enough to change the program."

© 2000 The Washington Post Company

home | school of education

by Mark Whitaker
© Copyright 2003 The Department of Educational Leadership and Policy Analysis, UW-Madison. All rights reserved.