With President Obama and Secretary of Education Arne Duncan hinting that school reform could be the big bipartisan focus of 2011, it’s an exciting time to be an education writer. At Time, Andy Rotherham outlines why it could be difficult for Democrats and Republicans to work together on education nevertheless; in short, the standards and accountability movement that rose to prominence in the 1990s and coalesced around No Child Left Behind is under attack from both the conservative wing of the GOP and (to a lesser extent) the liberal wing of the Democratic Party, both of which are skeptical of standardized testing and top-down education mandates.
But that doesn’t mean there won’t be movement. For reasons I explained previously, if an education policy compromise does emerge between the Obama administration and the John Boehner-led Republicans, it will probably be around merit pay for teachers. So it’s worth explaining what the trends are in merit pay and what the best social science research tells us about whether the policy raises student achievement.
What’s happening on merit pay for teachers? The Obama administration has offered states billions of federal stimulus dollars if they agree to change the way teachers are paid, evaluated and trained. Performance pay is a big part of that push, and states including Colorado, Louisiana and New York have responded by passing laws promising to tie teacher evaluation and pay to how well students perform academically.
Meanwhile, Washington’s outgoing schools chancellor, the (in)famous Michelle Rhee, last year instituted the nation’s most aggressive merit pay program to date, which uses private philanthropic funding to offer public school teachers achievement bonuses of up to $10,000 in exchange for weaker tenure protections. The teachers’ unions in Seattle, Pittsburgh, Denver and a number of other cities have also agreed to contracts that include some form of performance pay, a major break from the traditional lock-step salary ladder for teachers, based on degrees attained and years on the job.
How is teacher performance measured in these plans? This is undoubtedly the most controversial aspect of merit pay. Most of the new performance pay laws and union contracts measure “effectiveness,” at least in part, by looking at how well a teacher “grows” his or her students’ test scores from one year to the next. The best and fairest way to measure this growth is through a statistical tool called “value-added measurement,” which is championed by economists who study the teacher labor force. Value-added equations attempt to control for factors such as class size, years on the job, and students’ race and poverty while measuring how much of an impact teachers have on kids. (Click here to see a sample New York City value-added report.)
But value-added is a relatively new technique and hotly debated in academic circles; even its defenders, innovative young economists likeDoug Harris and Thomas Kane, have urged caution, as have theEducational Testing Service, the National Research Council and the Department of Education’s own Institute for Education Sciences. Outright opponents, such as a coalition of education experts convened by the liberal Economic Policy Institute, raise a number of powerful objections: that value-added measurements are based on flawed standardized tests; that the ratings are particularly volatile (a teacher who scores very well or very poorly has only a one-third chance of getting a similar score the following year); and that the technique gives the impression that the teacher is the only factor that matters to student achievement, ignoring parental involvement, afterschool tutoring and other “inputs.”
There are also a number of practical problems in using value-added measurement as a way to reward or punish teachers. Many educators do not teach tested grades or subjects, including early childhood teachers, art teachers, music teachers and the like. It is also unclear what effect competing for bonus pay will have on the climate within schools, where students are best served when teachers share best practices. Then there is the overarching concern that paying teachers according to student test scores will lead to “teaching to the test” — a real problem, especially considering years of research showing that NCLB severely limited the curriculum in low-income schools, leading to lots of reading and math drills, and not so much fun science experiments, art projects and journal writing.
So does merit pay work? It’s too soon to say whether merit pay for teachers based on value-added measurement really attracts and, more important, retains high-quality teachers in low-income schools. Surveys consistently report, though, that factors other than money — such as an orderly school, a supportive principal and good facilities — are what teachers value most.
There is more to work with when it comes to the effect of merit pay on student achievement. A study of the pay system in Nashville echoed earlier results found in Colorado, Texas and Florida, finding that bonuses alone do not make teachers better at raising student test scores. The implication is that most teachers are already working as hard as they are able to, so throwing more money at them won’t improve performance. Proponents of merit pay, however, say these studies don’t tell us much, because so far, most bonuses are not being distributed along with high-quality teacher training and improved professional development.
In short (and I know this has been a long post), the federal investment in teacher merit pay based on value-added measurement is based in large part on faith in data and on the hope these programs will make teaching a more attractive profession to ambitious, results-driven people. There is no hard evidence that these policies have been transformative in the places where they have been deployed. But neither is there proof that, especially when used in conjunction with other, more subjective measures, value-added measurement of teachers will have a corrosive effect on teaching and learning. The key, as my Teachers College adviser Luis Huerta told me recently, is not to allow value-added measurement and merit pay to “over-rationalize” education, which is, at its heart, a deeply subjective, cultural, emotional, and even artistic process.