General » Top Stories

Does ‘merit pay’ have merit?

The LR School District wants to give it a try. Teachers say the School Board hasn’t done its homework.

by and


The odds are not particularly good that the Walton Family Foundation’s merit pay proposal, approved by the Little Rock School Board Feb. 23, will make it past the required vote of all affected classroom teachers in about a month. The bar set by the teachers’ contract — 75 percent approval for any change in the compensation structure — is just too high.

But it’s probably just as unlikely that a “no” vote would be the final word on the subject. The proponents of merit pay — the district’s top administration, almost all the school board members, some business leaders and the Walton foundation, among others — will start working on another proposal.

Merit pay plans have been around in varying forms for more than 20 years, and are becoming more and more common across the country. Four Little Rock schools already have pay-for-performance programs, all financed originally by outside money, although one now has school district support. The idea of treating teachers more like other professionals — basing at least part of their salary on some measure of their skills or effectiveness — makes sense on its face to enough people interested in education reform, and has caught enough momentum nationwide, that it’s not simply going to disappear.

The problem hardly any district in the country seems to have solved at this point, though, is how to structure a merit pay plan so that it meets everyone’s idea of fairness, results in higher student achievement and doesn’t encourage teachers to, as one speaker at last week’s board meeting put it, teach the test instead of the curriculum. There is also the documented problem of increased cheating on standardized tests — in Houston and Chicago, among other places — by teachers who stood to benefit financially if their students’ scores went up.

The Walton proposal, designed to be a two-year pilot study, would randomly select 50 teachers from a pool of interested applicants. They would be eligible for bonuses of up to $10,000 each, depending how much their students improved on a standardized test that compares them with other students nationwide. Teachers who applied but weren’t picked would be the “control” group — evaluators would compare their students’ improvement with that of the teachers in the program to try to determine whether merit pay alone can boost test scores. Another 50 teachers would be picked the second year. The Walton foundation is budgeting $250,000 for the program’s first year, although if all teachers earned the maximum bonus they’d be on the hook for half a million. The school district would have to assume that cost in the second year. The foundation would also fund bonuses for the 50 teachers chosen in the second year.

If the Little Rock proposal fails the teacher vote, it will be for two primary reasons: It would reward only a few teachers based on a single measurement of student achievement, and it was handed down by district administrators without any input from teachers, community members, parents, or, for that matter, members of the Board of Education.

“Our feedback is not important,” Little Rock Classroom Teachers Association President Katherine Wright Knight said after last week’s board meeting. “This is just another example.”

It’s common sense that teachers would be more likely to support a proposal they had input into, and the experience of other school districts — most recently and most notably Denver, Colo. — backs that up.

In 1999, Denver’s school board and teachers union came almost to an impasse because the board wanted a merit play plan based on test scores. The two sides instead agreed to study the issue together, and took two years to develop a more broad-based plan with input from both sides. The plan, called ProComp, operated as a pilot for several years, and in 2004 Denver voters approved a $25 million tax increase to expand the program district-wide.

“The involvement of the union from the get-go had a lot do with why teachers and the public have gotten on board,” said Jeff Buck, a Denver high school teacher who was on the 10-member committee that designed the plan. New teachers have to participate, but veteran teachers can choose to stick with the old salary schedule.

The ProComp system rewards teachers for performance in nine categories, including student achievement, but standardized test scores were intentionally de-emphasized. All teachers are eligible, in contrast to the Walton proposal, which limits participation to classroom teachers in kindergarten through fifth grade — those whose students take the Iowa Test of Basic Skills, the standardized test used as the sole criterion for bonus pay.

Teachers weren’t consulted on this proposal because they weren’t the ones putting up the money, said Deputy Superintendent Hugh Hattabaugh.

The result, though, is a clear difference of ideas about the point of implementing such a narrow program.

Repeatedly at last week’s board meeting, teachers opposed to the plan said it was unfair to give bonuses to such a small number of teachers, and to completely disqualify teachers in middle and high schools, teachers of subjects like music and foreign language that aren’t on the Iowa test, non-classroom teachers like reading specialists, and non-certified personnel like instructional aides who assist teachers in the classroom.

One, Parkview math teacher Phillip Wilson, criticized board members for trying a program that doesn’t have evidence to back up its effectiveness.

“I am only asking you board members to do what you’ve demanded of us teachers for years — research-based strategies,” he said.

But the lack of evidence is kind of the point, said Gary Ritter, an education researcher with the University of Arkansas’s College of Education Reform (funded largely by Walton money) who helped design the evaluation part of the proposal. The foundation would pay the college $97,000 for the evaluation.

“We know people have strong theories about why it might and might not work, but we don’t have strong evidence,” he said. “It appears to be a path that many folks are going down anyway. I would prefer going down it with some evidence.”

It’s not possible, Ritter said, to get that evidence from programs like the one funded by Democrat-Gazette Publisher Walter Hussman, in which all instructors at two schools — Meadowcliff and Wakefield elementaries — are eligible for bonuses if test scores go up. It’s too difficult to separate out other factors that might affect student achievement at an entire school, such as a strong principal or a new schoolwide educational program, he said.

“We experiment with kids all the time,” Ritter said. “The problem is we don’t keep track of what happens.”

And the program is meant to be a limited pilot, not the district’s permanent, final merit pay program, Superintendent Roy Brooks emphasized at the school board meeting.

“Gathering data is one thing,” he said. “Implementing a pay-for-performance plan district-wide is another. That would require at the very least careful thought, research, community involvement and communication with teachers.”

The board’s 5-2 vote in favor of the Walton proposal can’t be taken as a face-value approval of the plan itself. Two of those who voted for the plan — Baker Kurrus and board President Micheal Daugherty — said they did so because they thought teachers should make the decision. (Kurrus downplayed the importance of the whole issue, saying that teacher compensation is a state law matter under the state Supreme Court’s Lake View ruling, and that the board and the teachers union should be working together to influence lawmakers.) And Tony Rose, who with Katherine Mitchell voted against the plan, said he was actually voting against the process by which it came to the school board — although he called basing merit pay on a single test score “foolish.”

“This plan has been handed to us,” he said. “I don’t know if it was designed by educators or not. I don’t know how it’s going to be evaluated and the Walton Family Foundation is not going to tell us. It’s going to be evaluated by a school of education reform that they created.

“The point of this is we need a comprehensive vision for what merit pay is going to look like in this school district. It needs to be something we put together in the very near future. … I’m going to vote against this because it’s the wrong approach. It may be the right plan a year from now after we’ve developed a plan.”

Rose said he didn’t see the point of the Walton proposal because it’s so similar to the merit pay program in place at Meadowcliff, and the results from Meadowcliff have been “ambiguous.” (This is Wakefield’s first year in the program, so there are no results from that school yet.)

Classroom teachers and other instructors at Meadowcliff got close to $135,000 in bonuses last year for their students’ improvement on the SAT-9 standardized test, which students took in addition to the state-required Iowa Test of Basic Skills and Benchmark exams. But students’ scores on the Benchmark exams — the ones that matter in terms of the state’s accountability system, which can lead to a school being “reconstituted” if scores don’t improve from year to year — landed Meadowcliff on the state’s “academic distress” list.

Meadowcliff Principal Erin Carter defended the merit pay program as a success. Improvement was what they were after, she said, and that’s what they produced. The Benchmark exams, she said, compare one class of students to the previous year’s students, while the merit pay program compares students’ scores at the beginning of the year to the same students’ scores at the end of the year.

Testing is a huge part of the merit pay issue: which tests to count, whether test scores are a reliable reflection of a teacher’s work over the course of a school year, how to measure the contributions of teachers whose subjects aren’t part of the standardized testing system — not just art and music, but foreign language, history, creative writing.

Arkansas students take only the math and language arts sections of the Iowa test, which is all multiple choice. It’s a “norm-referenced” test, meaning students’ scores show how well they did compared to every other student across the country who took the same test. A score of 50 means the student did better than 50 percent of all the kids who took the test.

The Benchmark exams, on the other hand — called “criterion-referenced” tests — operate more like conventional classroom tests. There are some short-answer questions, and students are scored against a set standard (created by Arkansas educators), not against each other. But scoring the Benchmark exams takes months — teachers typically haven’t gotten students’ scores back before the beginning of the following school year — so it’s not practical at this point to use them in a merit pay system, supporters of the Walton plan explained.

But testing is the law of the land, literally, so pretending test scores don’t or shouldn’t count isn’t realistic, Kurrus said.

On the other hand, they’re definitely not the only measure other school districts and states have used to award performance-based pay.

In Denver, for example, teachers earn an extra $999 if they teach at a school with a high percentage of low-income, special-needs or ESL students, or if they’re certified in a hard-to-staff area like middle school math or special education. They get tuition reimbursement for professional development. They’re evaluated every three years by their principal or other “qualified designee,” with a 3 percent raise for a satisfactory rating. And each teacher works with his or her principal to set two objectives for student achievement each year. If students reach one, the teacher gets a bonus; if they reach both, the bonus becomes a permanent salary increase.

Other systems pay teachers more for taking on more responsibilities, or require them to show that they’ve earned a promotion to a higher rank of teaching rather than automatically ushering them along a traditional pay-for-experience salary scale.

More than half the 50 states have instituted some kind of statewide pay for performance program at one point or another since the early 1980s, although some have been repealed. Arkansas currently gives a salary supplement to teachers who become nationally certified — a rigorous, months-long process that requires hundreds of hours of work on the part of the teacher.

And the subject is bubbling just under the surface at the Capitol now, although it’s a pretty thick surface: the requirements of the Lake View Supreme Court ruling and a probable special legislative session to address them.

But several lawmakers involved in education policy said they think some kind of merit pay discussion is in the state’s future.

“There’s a growing awareness that a system based solely on tenure and level of education does not serve as well,” said Sen. Jim Argue, chair of the Senate Education Committee and committees overseeing interim educational adequacy issues. The original consultants’ study that legislators used in 2003-4 to retool the state education finance system included a recommendation to reward teachers for their individual knowledge and skills, but lawmakers didn’t adopt it or appropriate money to study it further, Argue said.

He said his ideal merit pay plan starts with a testing system that gives teachers frequent and instant feedback, and not only assesses where students are but shows teachers where they need to focus.

And it also includes buy-in from teachers themselves.

“It can’t be something we’re forcing on the system,” Argue said. “That doesn’t do us any good. But I really think it’s possible to design a system that teachers would accept as a resource for them to better serve their students.”

Rep. Joyce Elliott, a former teacher, said she thinks the current system of paying educators is “absurd,” but that linking pay to test scores is “very tenuous.”

“There’s something to be said for, if teachers for example voluntarily took on students who have the most challenges,” she said. “That’s worthy of being granted some kind of enhancement pay. What we tend to do now is put our best teachers with the best students, and that’s backward.”

Elliott said what she really wants is for teachers themselves to take over the discussion of merit pay.

“That’s what grieves my heart more than anything — that we’re just saying ‘No, no, no.’ ”

Parkview English teacher John Binyon shares that view. Speaking at the Little Rock School Board meeting last week, he said he didn’t like the Walton proposal, but that the Classroom Teachers Association — of which he’s a member — needs to come up with its own plan rather than simply opposing the administration’s.

“I’m more pleased with [the school board members] than I am with the CTA,” he said after the meeting. “… I wish we could all get together because [merit pay] is inevitable.”

Katherine Wright Knight, president of the CTA, said after the meeting that she doesn’t think research supports any kind of merit pay, but that the teachers union would be willing to work with administrators and board members on a proposal if they’re asked to. So far, though, she said, they’ve simply had changes pushed on them.

Add a comment