Right on the Money

Despite repeated efforts to reward teachers based on performance -- both theirs and their students' -- many experts say this incentive doesn't improve education.

[caption id="" align="alignleft" width="200" caption="Illustrations by James Yang"]

[/caption]

Offering financial incentives to improve education -- providing money rewards to students, teachers, schools, or districts as a way to motivate them to try harder and do better -- is one of the hottest topics in education today.

On the student side, schools in cities like New York, Chicago, and Washington, D.C., are experimenting with financial rewards, including cash payouts to students who make good grades or show other achievement. The new competitive incentive grants from the federal Department of Education -- the so-called "Race to the Top" money -- hand out financial remuneration to states that comply with certain requirements, including improving academic results.

But the greatest focus has been "pay for performance" initiatives for teachers whose students make the most academic progress, typically measured by results of standardized tests. The concept is simple: A series of influential studies in recent years have shown that teacher quality is one of the most important factors in student achievement, so "good" teachers -- as reflected in growth in student test scores -- should be paid more than their less able colleagues. Financial incentives will encourage teachers to try harder in their jobs, the theory goes, and those who don't should leave the field and seek other careers. Pay for performance will rid schools of mediocre teachers, proponents say, leading to higher student achievement, betters schools, and, in the long-run, a more productive workforce in the United States.

In the ongoing effort to address the complicated issue of improving American education, pay for performance seems to make sense, and so the movement has caught on across the country. In the past decade, at least 20 states and a large number of districts have instituted some form of pay for performance for teachers, including California, Florida, Minnesota, Texas, and the cities of Cincinnati, Denver, New York, and Charlotte, N.C., according to Donald Gratz, Ed.M.'76, author of the new book, The Peril and Promise of Performance Pay. And President Obama has announced that the federal Teacher Incentive Fund, a competitive grant program to support pay for performance plans, will increase five-fold, from $97 million to $483 million.

But does pay for performance really work? According to many experts, the answer is a resounding no -- especially when teacher ability is measured solely or primarily on student scores on standardized tests.

"There has never been any research that shows that this works, although it's very fashionable to think that it should work," says Richard Rothstein, the former education columnist at The New York Times and the author of a number of books on education, including Grading Education: Getting Accountability Right.

"When it comes to the sexy reform du jour -- basing teachers' pay on student performance -- the research doesn't support it at all," concurs Bella Rosenberg, Ed.M.'72, an independent education consultant based in Washington, D.C., who worked for more than 20 years for the American Federation of Teachers. This year, Rosenberg did a project that required her to read "just about every piece of research available on this, including from the advocates," she says. She found no evidence that pay for performance improves education. "It's not there -- it's just not there," she says.

Indeed, since the idea of pay for performance first was born, in the 18th century, it has failed every time it's been tried, says Kitty Boles, Ed.D.'91, a senior lecturer at the Ed School. As early as 1710, in England, teachers were paid based on their students' test scores in reading, writing, and arithmetic. But problems with this approach quickly became apparent, she says. The curriculum narrowed as arts and science classes were no longer taught. Teachers focused on drills aimed at improving test scores, and "teaching to the test" was born. There were even scandals with teachers faking test scores. For these reasons, pay for performance -- also known as merit pay -- was abandoned. Over the past three centuries, it has been resurrected numerous times, and in each instance, Boles says, it has failed to improve education and was eventually dropped. This cycle has been repeated each time a merit pay system has been launched, including one championed by President Richard Nixon but declared a failure not long afterwards, Boles says.

Professor Susan Moore Johnson, M.A.T.'69, Ed.D.'81, agrees. "There have been waves of merit pay initiatives in the past, and every time someone recommends it anew, it's as if it's never been done before," says Johnson, who recently coauthored Redesigning Teacher Pay: A System for the Next Generation of Educators, a book garnering much attention in the education world by advocating a radically different approach to teacher pay that encourages teacher career development through a four-tier system of promotion.

Despite the history of merit pay, these plans continue to be reborn, including in various waves in the United States over the past century. Most recently, the passage in 2001 of the No Child Left Behind (NCLB) Act reignited the movement. By mandating that all states develop annual standardized tests to measure student performance, NCLB created objective standards that could be used for other purposes, too -- including as an ostensible means of judging teacher effectiveness. Merit pay gained real traction when the federal government instituted the fund that distributes awards to states and districts that create pay for performance plans in high-needs schools.

Proponents, insisting that tying teacher salaries to measurable standards will improve schools, have instituted a wide variety of incentive plans across the country: Some evaluate teachers based solely on standardized test scores, some on teacher skill development; some offer more pay to teachers working in at-risk schools or with at-risk children, or for teaching certain subjects. Some favor subjective measures such as a principal's evaluation of the teacher, which has its own critics who fear favoritism, and some rely on a combination of these and other factors.

To Boles, the format doesn't matter, whether it's purely objective or not; merit pay misses the point. "I'm not ready to say it will never work, but I doubt it will work because it's not the way we should be assessing teachers' abilities or skills," says Boles, who instead advocates better teacher training and a career path that involves mentoring and being mentored.

Plans that rely solely on student test scores have the most opponents, including many parents, who scorn "teaching to the test," in which students are drilled to increase their test scores rather than taught to understand the underlying material and learning skills to last a lifetime. Teachers' unions are strongly against these plans for a variety of reasons, including that they say it's nearly impossible to accurately measure an individual teacher's contribution to a student's success, since a child's achievement is cumulative over a period of years and the result of the efforts of many people. Some plans only reward the teachers whose subjects are tested; namely, reading and math teachers, thereby excluding others who also influence student achievement.

"We are opposed to any form of merit pay where pay goes to individual teachers based on student test scores," says Ed Doherty, Ed.D.'98, assistant to the president of the American Federation of Teachers in Massachusetts, which has 20,000 members. Not only is this the official position of the union, Doherty says, but in a survey of the 40,000 teachers in Massachusetts, about 90 percent oppose merit pay.

A related problem is the emphasis on subjects in which student performance is easiest to measure; namely, math and reading. "There's no way to measure performance other than in math or reading, other than by observing teachers in the classroom, but that's extremely expensive, so no one is talking about that," says Rothstein. By focusing on math and reading to the exclusion of other subjects, he says, "you create incentives to further distort and narrow the curriculum, which is disastrous."

He adds, "In any institution, if you have multiple goals and you create incentives to pursue only one or two, you get people abandoning other things they should be doing in order to focus on things for which they're held accountable." In his opinion, he continues, "this is one of biggest shortcomings of No Child Left Behind -- schools have abandoned science, social studies, history, arts, and physical education, which is particularly disastrous in low-income communities."

Teachers and many others are particularly offended by the underlying assumption of merit pay: namely, that teachers would work harder if only they were rewarded with even a minor financial bonus (the pay differential in most plans is typically quite small, only a few thousand dollars tops.) Most people who go into teaching are motivated by intrinsic rewards -- the value of the work they do -- rather than extrinsic motivators, such as money, many educators believe. "There's an assumption under this that teachers would try harder if they were paid more," says Gratz. "The corollary is that they're not trying hard enough now, which means they care more about money than kids. Frankly, teachers find that insulting."

Rothstein agrees. The flawed theory behind pay for performance is "that student achievement is not as high as you'd like it to be because teachers, to use the economists' term, are shirking, are not doing as well as they could, so they need incentives to work harder or better. That assumes that reason student achievement is poor is that teachers know what to do and just aren't doing it." To the contrary, Rothstein says, poor achievement in school is a larger problem that can't be laid in the laps of teachers. "The assumption is that all our problems are due to teachers, so we don't need to pay attention to social conditions students come from," he says.

Johnson concurs. "The essential assumption of pay for performance is that pay for performance is about effort, and that teachers who are offered a small sum of money -- and it's really very small, when you look at these plans -- will somehow redouble their efforts and solve problems [of student achievement] they don't know how to solve," she says.

Rob Stein, C.A.S.'93, Ed.D.'01, also believes that teacher motivation is not the core issue. Stein was named principal of an inner-city Denver high school when it reopened two years ago after being shut down for being the worst-performing school in the district. Denver's merit pay system, known as the Professional Compensation System (ProComp), is currently touted as the model system for merit pay because it had widespread support, including from teachers and parents when it passed about five years ago.

"Denver may be leading the nation, but it's still not a very good model," insists Stein, who describes himself as "agnostic" with regard to performance pay for teachers because he doesn't oppose it, per se, but believes it doesn't work. "Teachers probably are not drawn to the profession for financial incentives," says Stein, whose faculty receives additional pay because they work in his high-needs school. "These are people who already take a pay cut just by deciding to teach. Giving them a five or 10 percent bonus -- when they could earn much more in another field -- isn't a real incentive. Financial stakes aren't what they're in for, and the amount isn't enough to make a real difference anyway."

He says, "What I see is that people have a missionary zeal to want to work with kids who need them the most. I've never heard anyone say, 'I'm applying to your school because of the extra pay incentive.'" Still, he adds, if the district is going to offer rewards for things his teachers would do anyway, he's happy to ensure they receive them.

Gratz also believes merit pay doesn't incentivize teachers. But, he says, educators in Denver find that ProComp has some real benefits in getting all stakeholders -- teachers, principals, and parents -- focused on student success. In other words, he says, the schools benefited not from improved teacher commitment but as a consequence of everyone searching for ways to help students.

Merit pay may not compel teachers to try harder. But on the specific issue of attracting high-quality teachers to teach in at-risk schools or with difficult student populations, Jennifer Steele, Ed.D.'08, says financial rewards have an impact. Steele works for the Rand Corporation on projects related to pay for performance and teacher effectiveness; at Harvard, she wrote her dissertation on whether a $20,000 cash incentive in California would induce academically talented teachers to go to disadvantaged schools.

In fact, it did. The bonus increased by 28 percentage points the likelihood that gifted teachers would enter a low-performing school. "So far, we've found that you can influence the career choices of teachers with financial incentives," Steele says. Still, that's a different issue than rewarding teachers for student performance, she says. While test scores can be one measure, it's critical that they not be the sole measure. Rather, a broad set of factors should be evaluated in assessing a teacher's performance, including his or her lesson-plan portfolio. Another measure should be a principal's subjective evaluation of a teacher, which Steele says is a pretty good predictor of a teacher's effectiveness. While there are many critics of the subjective approach, it has an important role in order to balancing out the "teach to the test" and other negative consequences of relying solely on test scores.

What to Reward Eric Anderman, Ed.M.'86, and his wife, Lynley Hicks Anderman, who teach at The Ohio State University, are researchers who've studied and published in the area of educational motivation for 20 years. In their new book, Classroom Motivation, they argue that incentives can work in motivating students -- and teachers, too -- if they are properly structured. That means incentives should be awarded only if they are informational, meaning the student has really learned something, and if the reward is not perceived as controlling but provides the student some choice, such as deciding to read a book when not specifically asked to do so.

"You get rewarded not for doing something but learning something," says Anderman. "For example, you're rewarded for demonstrating to me that you know how to add a series of two-digit numbers and understand the process behind it, versus just completing a worksheet." The same principles apply to teachers: "If teachers are simply rewarded for following some kind of protocol or rule according to how it's mandated, that's not effective. But if the teacher did something creative, innovative, that would be great because it's coming from the teacher," he says.

The other kind of reward will work, he stresses, but has serious negative consequences. "The danger is people start doing things just to get the reward and lose interest in the activity itself. So the teacher might go through the appropriate behaviors in order to get a bigger paycheck instead of because he or she wants to teach kids. The same with k
ids: they'll do the worksheets in order to get the reward, but get bored with math and rule it out as a career," he says.

Of course, it's more time-consuming to base rewards on a larger portfolio of factors including subjective evaluations, which is why relying on test scores is popular. But in the end, it will backfire -- if the goal is to produce educated children. "If you keep rewarding for teaching to the test, teachers will keep doing that," says Anderman. "What would be a good thing would be if teachers were rewarded not just for making students achieve, but for specific ways of making them achieve" including learning critical thinking and other things harder to test but more valuable in the long run. Like Boles, Stein, and others, Anderman believes higher salaries across the board for teachers would be more useful than merit pay, as would better teacher preparatory programs, mentoring, and other ongoing supports.

A Multi-Tiered Approach Johnson is director of the Project on the Next Generation of Teachers, where she studies teachers' work and careers. Her latest book, Redesigning Teacher Pay: A System for the Next Generation of Educators, cowritten with John Papay, Ed.M.'05, an advanced doctoral student at the Ed School, is gaining attention from educators searching for better ways to approach teacher pay. The book grew out of two studies, the first of which took a broad look at pay for performance in four urban districts: Houston; Minneapolis; Charlotte- Mecklenberg, N.C.; and Hillsborough County, Fla. Without assessing these programs per se, Johnson explains, the book examines how these systems are set up, including whether they use student performance on standardized tests, professional evaluations, a hybrid model, and whether they used individual or group assessments.

But it's the second part of the book that is gaining attention in education circles. It offers a new and comprehensive approach for teacher pay that focuses on helping teachers develop their skills throughout their careers in order to benefit students and schools. Johnson and Papay's concept states that since money is not the primary motivating factor for teachers, it will neither attract nor retain them in the field. What's needed to cultivate good teachers -- and by extension, better students -- is a range of support including mentorship and the ability to learn and grow in a formal way, they believe.

"Teachers generally don't go into teaching for money, especially in these days when they have access to all other lines of work," in contrast to years past when women and men of color went into education because they were blocked from some fields, Johnson says. "Today, people who are choosing to teach are really choosing to teach, and it's with awareness of the limitations of salaries. No one expects to get rich. You hear this again and again in interviews with teachers." As most teachers will explain, she says, they're drawn to the field because they want to help students. "They'll say that again and again: It's the kids," she says.

For that reason, Johnson and Papay propose a system that would replace the common single-salary scale in teaching with a four-tiered pay structure that sets out goals and provides rewards in the form of substantially higher pay when teachers achieve them by being promoted to the next tier. Each of the tiers -- probationary teacher, professional teacher with tenure, master teachers and school-based leaders, and school and district leaders -- provide opportunities for career growth. And the system emphasizes career support that helps all teachers improve. Johnson and Papay also propose what they call a "Learning and Development Fund," created by diverting resources from the single-salary scale, to finance new learning opportunities for teachers, provide stipends for special staffing assignments, and give other support to assist teachers and schools.

Johnson believes that such an approach -- with its emphasis on investing in teachers' careers -- is the answer to a stable and successful teaching corps.

"Districts that implement the tiered pay-and-career structure and its companion Learning and Development Fund will fundamentally change how they recruit, compensate, assess, and develop teachers," she and Papay write. "As a result, their schools should achieve greater stability, steady improvement, and increased student success."

-- Elaine McArdle is a freelance writer based in Cambridge. Her last piece in Ed. explored rural education.

When we initially started this story, we wanted to look at incentives overall in schools -- for teachers and students. That turned out to be a huge undertaking, too huge for one short magazine story. In addition, Harvard Professor Roland Fryer's new research on student incentives wasn't being released until after the publication of our magazine. For those reasons, we decided to tackle the topic with two stories -- teacher incentives in this issue of Ed. and student incentives in a web-exclusive, "Earn to Learn?". As always, let us know what you think!