Ignoring pleas from teachers' unions, the Los Angeles Times went public last week with an online database of 6,000 local third-, fourth-, and fifth-grade teachers, ranking -- and naming -- each by their efficacy in raising student test scores. Thanks to the paper's feat of selfless journalism, now countless parents, once naively confident that their child was getting a good education, can plug in a teacher's name and see if their child's instructor is in fact incompetent.
The teachers' unions were not pleased. A.J. Duffy, president of the United Teachers of Los Angeles, said he was "outraged" the paper would "put this out and put people in harm's way." But many -- including education officials in the Obama administration -- welcomed the exposé as a refreshing bit of transparency for a system that eats up over half a trillion dollars of taxpayer money a year -- yet provides little accountability. "What's there to hide?" asked U.S. Education Secretary Arne Duncan.
But while rooting out bad teachers and excoriating unions for protecting them has become the cause célèbre of education reformers, this isn't simply a case of teachers' unions trying to protect their own. Since the passage of the No Child Left Behind Act (NCLB) in 2001, schools have been required to test students' progress on a yearly basis. There's been an ongoing debate not only about what exactly these tests measure but whether -- and how -- the resulting data should be used to evaluate teachers. For the first time, education-policy experts have a big heap of standardized statistics gathered over several years -- instead of a hodgepodge of sporadically taken samples -- and no one knows quite what to do with them.
On one hand, NCLB has provided experts with more information about how students and schools perform on standardized tests, inaugurating a new era of data-driven education policy. It's allowed states to identify and improve problem schools and -- because tests quickly revealed how wide the achievement gap was between white and minority students -- led educators to focus on improving education for the neediest kids. The law makes inroads on the accountability front by requiring failing schools to provide parents with an annual report card explaining what they are doing to improve test scores. Standardized testing also provides a tangible, quantifiable way to evaluate how teachers are doing.
But as countless commentators have pointed out, there are glaring problems with the use of standardized testing in education. The most salient controversy has been over how these tests measure learning. While they provide a basic assessment for how students are doing in math and reading, they exclude all other subjects, which has led critics to charge that NCLB hasn't improved education as much as it's narrowed it.
Teachers' unions further argue that the use of student test scores to evaluate teachers has encouraged educators to "teach to the test" rather than provide kids with a broad, well-rounded education. This may sound like a quaint criticism, but can a narrowly focused teacher who produces results on a formulaic test also be the type of inspiring educator an adult will look back on and say, "That person really changed me"? Are we now cultivating an army of dull SAT-prep teachers? And, as George W. Bush put it so poetically, "Is our children learning?"
To the chagrin of many teachers' groups, the Obama administration has largely continued the previous administration's testing regime, though it's made more of an effort to improve failing schools rather than siphon off their funding with vouchers. Under President Barack Obama, the Department of Education has also stepped up efforts to use testing data to evaluate instructors. Race to the Top -- Obama's signature education initiative -- provides states with $4.5 billion in competitive grants, but to qualify they must allow their teachers to be assessed based on their ability to raise test scores. In response to the competition, 12 states have passed laws doing just that.
Any expert will tell you that teacher evaluations based on standardized tests are only as good as the tests themselves, but even assuming they provide a suitable barometer for learning, the relevant question is how well test scores reflect teacher quality. After all, kids' scores are also influenced by social factors such as their race, English language ability, individual motivation and talent, and peers, as well as their parents' income and education. A teacher in a wealthy district is almost sure to have higher-performing students than one in a poor district. In response to this dilemma, education reformers have introduced value-added assessments, which the Los Angeles Times used in its teacher rankings (the paper hired RAND Corporation economist Richard Buddin to conduct it). Value-added assessments -- VAAs for short -- use sophisticated statistical models to strip away the effect of different social variables, ideally leaving a neat indicator of teacher quality. While these assessments have been heralded as a breakthrough in evaluating teachers, their use is not without controversy.
A recent report from the Economic Policy Institute reveals that VAA results vary widely depending on the methodology used and produce inconsistent results from year to year. One study found that, even when using the same methods, a third of teachers ranked in the top 20 percent one year dropped to the bottom 40 percent the next. Scholars also routinely point out that these models apply to idealized circumstances in which students are randomly assigned to classrooms and students at all levels learn at the same pace. Because these kinks still need to be ironed out, various scholarly organizations -- including the National Academy of Sciences and even the RAND Corporation -- have said VAAs shouldn't be used as the sole, or even the primary, tool for evaluating teachers.
Which makes it strange that the Times would publish the results of a single VAA analysis under the guise of providing "objective information about instructors' effectiveness." As Randi Weingarten, president of the American Federation of Teachers, put it, VAAs are "not ready for prime time." Experts -- and even education reformers who have no vested interest in protecting teachers -- are still debating how best to model teacher quality based on test scores. In passing off its results as a definitive assessment instead of reporting on the various angles of the issue or presenting the raw, unprocessed data graphically as other papers have done, the paper essentially laid down a stake in the debate -- and joined the blame-the-teacher bandwagon.
Still, the Times' decision is understandable. Every year, taxpayers pay hundreds of billions of dollars into the public education system, and it's tough to tell whether we're getting our money's worth. But those who are impatient with the pace of school reform and eager to have a concrete way to evaluate their efficacy -- a group that we can now safely say includes the Los Angeles Times -- tend to overlook the fact that teaching and learning are complicated endeavors. Yes, more data is now available. But it will take time to figure out better ways of collecting and analyzing it.
To respond to Arne Duncan, the objection is not that there's "something to hide"; it's that we still don't know what this data is revealing.