Cracks in the NCLB Foundation

In a speech earlier this week, Rep. George Miller (D-CA), Chairman of the House Education and Labor Committee, staked out his vision for the reauthorization of the No Child Left Behind Act. Much of the subsequent discussion focused on the Miller's proposal for what are commonly called "multiple measures." He said:

Our legislation will continue to place strong emphasis on reading and math skills. But it will allow states to use more than their reading and math test results to determine how well schools and students are doing.

This is one of those issues where a few words here or there can outweigh the hundreds of pages that comprise the rest of the law. It's so important that a group of NCLB supporters sent Miller a letter a few weeks ago (link via This Week in Education) saying, essentially, "Please, please don't screw this up."

As this gets discussed in the coming weeks, months, and possibly years, it's important to understand what's at stake. Nearly all the back-and-forth will be about what gets measured. But equally important--perhaps more important--is who does the measuring.

The multiple measures idea stems from one the most common--and correct--criticisms of NCLB: schools are rated almost exclusively based on state assessments in reading and math. This system can be inaccurate and reductive--not only are we limited to one way of measurement, via standardized tests, but we're also limited in what's measured. Subjects like art, music, social studies, etc. are left out, along with the non-academic skills and character traits that schools are charged with teaching students. By expanding school measures beyond once-a-year tests, the thinking goes, we can get a broader, more nuanced, more accurate sense of what schools are really doing for their students.

A worthy goal, to be sure. But here's the problem: in many multiple measure scenarios, it's the schools themselves that will be doing the measuring. And that undermines one of the great virtues of NCLB: the separation of those being held accountable from the process by which they're judged. That independence is based on a rock-solid understanding of human nature: people can't be wholly accountable to themselves.

The people here at Education Sector who handle accounting, for example, are scrupulously honest. Nonetheless, we're required to have our books audited by an outside accounting firm every year. Nobody disputes the necessity of this, just like nobody disagrees with having line judges call serves in and out at Wimbledon. When the stakes are high--as with money, championship tennis, and the educational lives of the nation's schoolchildren--measurement must be independent.

Local measurement will also inevitably create huge inconsistency and variance among schools and districts. People are already confounded by the fact that there are essentially 50 versions of NCLB, one for each state. What's the law going to look like if there's one version for each of the nation's 14,000 school districts, or 90,000 schools? A lot like having no accountability at all.

These are the reasons that we're stuck, for the moment, with standardized tests, their many flaws and limitations notwithstanding. And it's why this single issue has the potential to make various cliched dam-related metaphors come to life. Crack open NCLB with misguided multiple measures, and the entire vast enterprise will collapse.