When you take out the opinion and cant, Diane Ravitch’s evidence for “killing off” the No Child Left Behind Act (“Time to Kill ‘No Child Left Behind,’” June 17, 2009) really comes down to her shrewd and misleading comparison of student-performance data on the National Assessment of Educational Progress from 1999 to 2004 with the NAEP data from 2004 to 2008.
Recently, Ms. Ravitch wrote an indictment against New York City officials for using test statistics as “damn lies” to justify conclusions she believed to be false claims. Yet that unacceptable approach is exactly the one she has taken in this opinion piece on NCLB. And here’s how she’s done it.
First, and less important, she fails to mention in using 1999-2004 data that NCLB was passed by Congress in 2001 and began to be implemented in early 2002. So, two-plus years of NCLB are in the base period she’s comparing against the NCLB period.
Second, and more important, her argumentation seriously breaks down when she fails to offer any sort of hypothesis for why 1999-2004 had strong results. I can’t remember Ms. Ravitch writing anything in the late 1990s that was supportive of policy at the time. And I see no explanation here.
Lacking one from her, let me lay out a hypothesis for all the improvement we’ve had since the late ’90s: The standards-based reform movement kicked into gear in the mid-’90s with the early steps in accountability in North Carolina, Texas, and Massachusetts. It picked up steam in the late ’90s, partly spurred on by the Improving America’s Schools Act. And it flowered with the adoption of the No Child Left Behind Act and the implementation of standards-based reform in many of the remaining states.
We could compare 1999-2004 with 2004-2008, if Ms. Ravitch insists. The earlier period is ever-so-slightly better, but I don’t know, nor does she, how to allocate credit for pre- and post-NCLB between 2002 and 2004. So, perhaps it’s safest either simply to say both periods were relatively strong or, even more to the point, to conclude that both periods should be joined together for analytical purposes, since similar policies were being administered throughout the decade.
We need fixes (not weakening), and we need to make important next steps. But before we listen to those whose failed policies have not worked in the past telling us to kill NCLB, let’s be very, very careful.”
We could compare either or both of these periods to the 1990s, when Ms. Ravitch happened to serve at the U.S. Department of Education. In fact, this is the most appropriate comparison. After all, the late-’80s/’90s were a discrete period of time between the 1970s and early ’80s, which bore all the markings and effects of the civil rights movement, and the late 1990s and 2000s, the period of the fruition of the standards-based-reform movement.
So, Ms. Ravitch, while we’re considering “killing off” NCLB, let’s really compare data so we can understand the effect on student results between the recent policies and those that were in effect in the late ’80s and ’90s. The data that follow come from the following sources: trend in NAEP mathematics average scores for 9-year-old students, by race/ethnicity; trend in NAEP mathematics average scores for 13-year-old students, by race/ethnicity; trend in NAEP reading average scores for 9-year-old students, by race/ethnicity; and trend in NAEP reading average scores for 13-year-old students, by race/ethnicity.
And, while we look at this data, let’s remember a lesson I once learned from Grover J. “Russ” Whitehurst, the former head of the federal Institute of Education Sciences, that a gain of approximately 10 scale score points is roughly equivalent to a grade level.
From 1990 to 1999, scores for black 9-year-olds were essentially flat, going from scale scores of 208 to 211 in math. From 1999 to 2008, these students gained more than a grade level, with scale scores going from 211 to 224.
From 1990 to 1999, black 13-year-olds’ scores were also flat, moving only from 249 to 251. Yet from 1999 to 2008, they, too, gained more than a grade level, going from 251 to 262.
From 1990 to 1999, scale scores for Hispanic 9-year-olds went down, from 214 to 213. From 1999 to 2008, they went up: a miraculous jump from 213 to 234. The gap with non-Hispanic whites closed here, from 26 points to 16 points. Indeed, Hispanic 9-year-olds are now performing as well in math as non-Hispanic whites were when Ms. Ravitch was serving in government.
From 1990 to 1999, Hispanic 13-year-olds’ scores went from 255 to 259. From 1999 to 2008, they went from 259 to 268.
Here are the reading data.
From 1990 to 1999, the scores of black 9-year-olds went from 182 to 186. From 1999 to 2008, they went up: from 186 to 204. This represents a closing of the white-black gap from 35 points to 24 points and represents the best growth ever, alongside that of the 1970s.
From 1990 to 1999, black 13-year-olds saw their scores actually go down, from 241 to 238. From 1999 to 2008, they went up, from 238 to 247.
Hispanic 9-year-olds’ scores were flat in reading from 1990 to 1999, going from 189 to 193. But from 1999 to 2008, they gained virtually a grade level and a half, going from 193 to 207.
So, while black and Hispanic reading and math results were virtually stagnant in the 1990s, these students gained a full grade level, and sometimes more, in both reading and math, from 1999 to 2008. You get the point.
Standards-based reform works. The No Child Left Behind Act has worked. We need fixes (not weakening), and we need to make important next steps. But before we listen to those whose failed policies have not worked in the past telling us to kill NCLB, let’s be very, very careful. That would be the absolutely wrong move for our children, particularly poor children and children of color.
A version of this article appeared in the June 17, 2009 edition of Education Week as Let’s Not ‘Kill Off’ NCLB