For more than a decade, states, districts, schools, and teachers have devoted enormous energy to closing achievement gaps between rich and poor students and among students from different racial and ethnic groups. But how much progress has been made in narrowing these gaps? Although the specifics depend on which gauge of achievement is used, the basic answer is clear by any measure—not enough.

In a study for the Center on Education Policy, consultants Naomi and Victor Chudowsky and I looked at trends in achievement gaps from 2002 through 2009 at grades 4, 8, and high school on the state tests used for No Child Left Behind accountability. To address some of the well-publicized limitations of state tests, we analyzed not only gaps in percentages of students scoring proficient—the coin of the No Child Left Behind (NCLB) realm—but also gaps in average scores on both state tests and the main assessment of the National Assessment of Education Progress (NAEP).

We found some nuggets of positive news. Black-white and Latino-white achievement gaps on state reading and math tests narrowed during these years in a majority of states with sufficient trend data. And when gaps shrunk, it typically wasn’t at the expense of higher-performing students—usually both groups made gains, though the lower-performing group improved at a faster rate. Gaps for Native American students and poor students narrowed less often, however.

NAEP, which differs from most states’ tests in content, rigor, and other features, offers an additional take on achievement gaps. The latest results from the main NAEP show mixed progress over the past two decades in narrowing gaps. Our own study, which focused on 2005 through 2009, found that even in the same state, gaps sometimes narrowed on the state test but widened on NAEP or vice versa.

Yet another indicator, the long-term trend NAEP, provides a forty-year historical perspective. Since the 1970s, African American-white gaps and Latino-white gaps on the long-term NAEP have narrowed significantly in both reading and math (except for the Latino-white gap in reading for 13-year-olds). But much of this improvement took place in the ‘70s and ‘80s. Although there has been some waxing and waning of progress since then, many gaps on the long-term NAEP were roughly as large in 2008 as they were twenty years ago.

Our own study acknowledges that achievement gaps on state tests also remain wide and persistent. In 2009, black-white gaps amounted to 20-30 percentage points in many states, and Native American-white gaps were about as large. Latino-white gaps in proficiency often hovered around 15 to 20 percentage points, while gaps between low-income and more advantaged students were in the 25-point range. If current rates of progress were to continue unchanged, and if other critical factors stayed the same, it could take decades to close these gaps in many states.

All of these findings point to the need for more intensive, fresh, and far-reaching strategies to address achievement gaps. A wealth of ideas for reducing gaps—from improving access to veteran, well-trained, and high-performing teachers to addressing out-of-school factors—are explored in Narrowing the Achievement Gap: Perspectives and Strategies for Challenging Times published by Harvard Education Press. (This volume also includes a chapter summarizing our data from the CEP study.)

A final observation: For all its flaws, NCLB made every school keenly aware of the size of its achievement gaps and made closing them a high national priority. Let’s make sure the next version of the federal Elementary and Secondary Education Act doesn’t leave us without the tools to monitor achievement gaps or the pressure to do something about them.