Insight
Pay for performance not performing well
Apr 12, 2007 3:55 PM
Places using such models have run into snags
Guess what? New teacher pay-for-performance plans in Florida and Texas have run into big problems.
Not surprised? Aha. You may be a teacher.
The Florida and Texas plans represent perhaps the worst model of performance pay, also called “merit pay.” Bonuses are based primarily or exclusively on students’ standardized test scores and then awarded to a predetermined percentage of top performers. There are better ways for districts to pay for teacher skills, but those have been buried in the rush to implement the simpler models.
Those models are already showing wear. In Florida, the Republican-controlled legislature is now rewriting the $148-million system it put in place just a year ago, after teachers reacted with fury and some school boards refused to carry it out. In Houston, Texas, teacher awards proved “embarrassing,” according to Houston Federation of Teachers President Gayle Fallon. “We have teachers who are on ‘growth plans’ — which isn’t a good thing — getting bonuses,” she said. Meanwhile, several “teachers of the year” got nothing.
Then on March 29, Texas legislators rebuffed their Republican leadership and voted to end the state’s merit pay plan, using the money instead for across-the-board raises for teachers and staff.
What lies on the surface
Despite extensive evidence of failure, the test-score-based plans are persistently popular in business and government circles. The federal government is forging ahead with a Teacher Incentive Fund which subsidizes merit pay based on test scores. And close to home, Chancellor Joel Klein, despite failing to get UFT agreement for such a pay plan in the latest contract round, is reportedly seeking out private funding for something similar.
What drives this obsession with performance pay?
To be fair, performance pay is a response to a real problem. Many great teachers go completely unrecognized throughout their careers. There are few avenues for advancement. If teaching is indeed a profession, then shouldn’t the wage structure reflect career advances?
In addition, teacher salaries are relatively low. It’s hard to attract and keep top teaching talent when a law degree or an MBA would easily triple the paycheck. Performance pay enhances a few salaries, but at the same time keeps district costs down.
For those in government and business, such pay for performance also has a knee-jerk appeal. Trained by traditional economics to “incentivize” for the results you want, many school system leaders find performance pay a no-brainer way to reward workers for raising test scores.
Former Florida Gov. Jeb Bush told a New York audience last September, “We have the database and we can put together pay for performance. It is the single most important thing we must do, and the hardest. This is controversial, but we have to pay teachers more for raising student achievement.”
A checkered past
Controversial? Pay for performance, especially the ham-fisted merit-pay versions used in Florida and Texas, have been tried and discarded in dozens of school districts over the last several decades. “These systems were often arbitrary, using poorly defined measures of performance and providing inadequate opportunities to observe actual performance in the classroom,” writes Allen Odden, a University of Wisconsin school finance expert.
The littered trail of failed merit pay plans does not mean that nothing can change in teacher compensation, but it does contain warnings. Arbitrariness or bias on the part of administrators, statistical errors, unfair treatment of teachers, a lack of funding and, most importantly, a lack of clear results have all dogged merit pay plans of the past.
Many violate the spirit of teacher collaboration. Twenty years ago, Harvard professor Richard Murnane and Michigan State professor David Cohen explained that schools may have very clear goals but “the goals are multidimensional — for example, raise the average reading level in each class, and eliminate drugs and violence from the school.”
While it might be possible to credit reading improvements to a particular teacher working behind closed doors, they wrote, “It is not possible to measure each teacher’s contribution to the second goal. In fact, eliminating violence and drugs from a school requires that teachers open their classroom doors and work as a team. … Managers (i.e., school principals) as well as teachers realize that much of the important work in schools must be done by teachers working together.” This may explain, they said, why paying teachers on the basis of student test scores is “extraordinarily rare in American education.”
Digging down for what’s real
Some more carefully constructed types of performance pay are being tried in Denver, Minneapolis, Chattanooga and several districts in Arizona. Their approaches are multilayered and the implementation is cautious.
In some states and school districts, new grade-by-grade “value-added” test data is being used to assess teacher performance. However, the pioneer in this area, Tennessee, requires three years of data before teachers can be evaluated on the basis of their students’ test score gains. Even then, Tennessee does not link the evaluation to teacher pay at all. (The Florida plan calls for bonuses to be awarded this June, after just one year of data.)
The Arizona-based Teacher Advancement Program gives value-added test data no more than 20 percent of the weight in teacher performance pay. The rest of the evaluation is based on peer and administrator observation and whole-school improvement.
Points of light
Pay-for-performance programs that have had some success incorporate limited value-added assessments, but they go about things differently.
- They focus on increased skills and knowledge, or demonstrated teaching expertise such as certification by the National Board for Professional Teaching Standards.
- They reward teachers for taking on added responsibilities, such as the lead teacher program in New York City, which was developed by the UFT.
- A plan can reward an entire school for improved performance. The UFT agreed to an experiment like this — the Breakthrough for Learning program in Districts 19 and 23 — several years ago. It proved ineffective and was eventually halted by its business sponsors, but similar plans have had more success in other states.
- A career-ladder plan, by which teachers are promoted through stages of professional growth based on mastery of teaching skills, was developed by a UFT membership committee and adopted by the Delegate Assembly in 1999. A similar career ladder was linked to pay in Cincinnati, though the plan has since been suspended.
Even the better pay-for-performance plans have a ways to go. But paying teachers based on test scores alone will never succeed. This may not be what Bush or Klein and Mayor Michael Bloomberg want to hear, but the truth is schools are more than the sum of their student test scores. They are living communities where “merit” has multiple dimensions.
