The NAEP Data: Does It Show That ‘Education Reform’ Has Worked—Or That It Hasn’t?

Cross-posted at Education Week.

The April 1st Washington Post carried an op-ed by Arne Duncan, President Obama’s Secretary of Education titled, “People are saying education reform hasn’t worked.  Don’t believe them.” On April 10, Education Week announced the most recent results from the National Assessment of Educational Progress with the following headline: “Achievement Flattens as Gaps Widen Between High and Low Performers.”   I wish also to call to your attention another Education Week headline: “DeVos Calls School Choice-Friendly Fla. A ‘Bright Spot’ in ‘Stagnant NAEP Scores.”  In my blog this week, I will parse the claim, counterclaim and counter-counterclaim.

First up is Arne Duncan’s claim that education reform has worked.  Duncan’s exhibit A is NAEP.  Since 1971, he tells us, 4th grade reading and math scores are up eight points and 19 points respectively and much of the gains have been driven by students of color.

Yes, there have been great gains made by students of color since 1971 as recorded by NAEP. The former Secretary claims that the “education reforms” he championed when in office were responsible for the gains he cites.  But more than half of those gains cited were made in the 1970s and 80s as a result of the ESEA legislation, long before Arne Duncan came to Washington.

Duncan notes that NAEP does not record comparable gains at the high school level, and wonders out loud whether that is the result of excessive resistance to change among the nation’s high schools.  Duncan might want to consider the possibility that since most of the ESEA money went to the lower schools, not to the high schools, that’s why the lower schools did so much better.

Here’s what Education Week had to say about the just-released NAEP results: “The results… showed no change at all for 4th grade in either subject [referring here to mathematics and reading] or for 8th graders in math since the tests were last given in 2015.  Eighth graders on average made only a 1-point gain in reading….The meager gain in reading was driven entirely by the top 25 percent of students.  During the last decade, 8th grade reading was the only test in which the average score for both high and low performers rose.”  Since then, they have been diverging, thus showing a widening gap between the high and low performers.  Many commentators spoke of continued “stagnation” in NAEP scores and highlighted the growing gap in performance between the high and low performers.  A few noted the parallel between the stagnation in the performance of American 4th and 8th graders on NAEP and the flat performance of American 15-year-olds on PISA for years on end, which, in turn, mirrored the flat performance of American high school students on the ‘historical’ version of NAEP until it was discontinued.

The picture that Secretary DeVos paints with the NAEP data is, if you don’t mind my saying so, a bit bizarre.  She uses the release of that data to point to Florida as the “bright spot” in an overall picture of “stagnant” NAEP scores.  Florida, she points out, was able to show “significant improvement” in 4th and 8th grade mathematics and 8th grade reading.  What she failed to point out is that Florida ranked 30th in Education Week’s annual Quality Counts K-12 state ranking for 2018 and its meager gains on this administration of NAEP only restore it to its performance levels in 2013 after dipping in 2015. Indeed, the state is doing no better now than it did in 2007.  She touted Florida’s scores, of course, because it is a poster state for the reforms that the Secretary favors.  Just as Secretary Duncan used 20th century school performance to create the impression that the reforms he favors produced 21st century progress on NAEP scores, Secretary DeVos would have you believe that the current NAEP report shows that the reforms she favors work.  If Secretary DeVos wants to use Florida as proof that her reforms works, Florida will have to perform at much higher levels than it does and show consistent growth over several years.  The truth is that the NAEP numbers don’t validate the reform agendas of either secretary.

If you want a more balanced, nuanced analysis of state performance on NAEP as revealed by the data, you should read the op-ed piece by John White, Louisiana’s Commissioner of Education, that appeared in the Washington Post about two months ago.

Regular readers of this blog will not be surprised by my take on this subject.  The most important indicator of American student performance is not 4th grade NAEP or 8th grade NAEP.  It is American students’ performance on PISA, an international comparative measure of the performance of 15-year-old students in 72 countries, including virtually all of the wealthiest countries as well as many others.  The performance of American students on this assessment has been essentially flat since it was first administered in the year 2000.  In math, 20 countries scored higher than the U.S. in that first year.  Now, 36 countries have statistically significantly higher scores than the U.S in math. Students in the top-performing countries leave high school two to three years ahead of our students.  A considerably larger fraction of their students performs in the top quarter of the PISA performance scale than is true of American students.  And a larger fraction of American students performs in the lowest quarter of that scale than is true for these top performers.  Which is to say that both our top performers and our lowest performers are way behind our competitors.  And, by the way, remember that big gain in mathematics performance on NAEP that Arne Duncan referred to with such pride?  Well, American students are further behind their counterparts on the PISA survey in mathematics than any other subject.

The people who say that reform has not worked are both right and wrong.  They are right if what they mean is that, where it counts most, at the point at which students leave high school, performance has been essentially flat for decades.  They might also have noted that this essentially flat performance has been purchased at a steeply rising price in dollars-per-student.

But these critics are dead wrong if what they are saying is that reform does not work.  If reform does not work, there would be no list of close to 30 countries whose students are outperforming ours.  Those countries have for years been pursuing reforms that have worked very well indeed.  It is not that reforms do not work.  It is the reform agendas of both secretaries that has not worked in the United States.

Let’s look a little closer at the reforms that Duncan identifies as “the reform movement.”  These are the new standards—presumably he means the Common Core State Standards and the Next Generation Science Standards and the assessments intended to measure the Common Core—the various incarnations of choice among schools and teacher accountability, all vigorously promoted by Secretary Duncan when he was in office.

It is true that a large majority of states have adopted these standards and most, despite pressure to drop them, have kept them, often after calling them something else.  It is also true that many states have adopted the assessments the federal government helped develop to measure the extent to which their students have mastered the standards.  But, as I have repeatedly pointed out, none of the high-performing countries ever believed that the implementation of standards and assessments, by themselves, would produce big gains in student achievement.  Most use their standards to produce well-developed curriculum frameworks and course syllabi that teachers all over their country are supposed to use.  Their assessments are designed to measure what is in the syllabi. While in teachers college, teachers are taught how to teach the courses defined by the syllabi. So, the standards and assessments are only two parts of a tightly integrated system of instruction defined by the state.  But most states just conducted workshops for teachers designed to tell them what was in the standards.  Few states even attempted to give their teachers the knowledge, skills or even the high-quality curricula needed to teach the new standards like New York State when David Steiner was Commissioner of Education and Louisiana under John White.  No state has been able to implement the kind of full standards-based instructional system I just described.  Standards-based education has not failed.  It was never tried.

That leaves two of Secretary Duncan’s markers of “education reform.”  The first is choice.  As I have pointed out, there is not a single top-performing country that has used choice to propel its students to high achievement, with equity.  There are countries that used the same strategies that the other top performers have used to get high achievement and that have, in addition, embraced choice, but they do that because they value choice as a right in a democratic society, not because they believe it results in net gains in student achievement.  The fact is that, all over the United States, in study after study, choice and regular schools run neck and neck in student achievement, or choice comes up short, when students are matched on background.  If Secretary Duncan thinks that choice is an effective strategy for raising student performance at the scale of a state, he has yet to make his case.

The third and last arrow in Secretary Duncan’s “education reform” quiver is teacher accountability.  On this one, the critics are dead right. This “reform strategy,” a keystone of federal education policy while Arne Duncan was Secretary of Education, can only be described as a catastrophic failure.  Built on deeply flawed methods for connecting the performance of students to individual teachers, it sought to impose often draconian sanctions on teachers whose students were said to perform poorly and on principals who failed to get rid of such teachers.  While other countries put their efforts into producing a surplus of great teachers, the United States under Arne Duncan went looking for poor teachers, using methods for identifying them that had severe negative consequences.  Little wonder that survey after survey showed that teachers were telling their own children not to go into teaching, applications to teachers colleges plummeted and good teachers left teaching before getting to retirement age in ever-increasing numbers.  The result has been growing teacher shortages, which legislators have responded to by waiving the already very low requirements for teacher licensure and authorizing the granting of emergency certificates to people manifestly unqualified to teach.

Secretary DeVos has offered no evidence that her brand of school choice—that is, a choice between public and private providers—works, if by “works” one means that it is a lever for improved student performance.  She appears to be in the camp of those who believe that choice is a value in and of itself in a democratic society and does not need to be justified by any other criterion, such as effectiveness.  In her case, the choice that is more important is the choice between a public and a private—especially religiously affiliated—provider.  If I have that right, efficacy is simply not an issue, at least for her, though it is for me.

So I come back at the end to the point at which I started.  The NAEP data do not show that reform has failed, nor do they show that reform has succeeded.  What they do show is that the American model of school reform has failed, if you measure success as improved achievement and narrower gaps between the top and bottom performers among our students, at the point at which they leave high school, which is the only point that really counts.  If you want a model of education reform that could vault any state in the nation to the top of the global league tables, go and look at the education systems of the countries that are already at the top of those league tables.