Newest Test Scores are Bad News for Centralized Education, Common Core

This morning I read an op-ed by Douglas Holtz-Eakin tackling the chasm between what it takes to enroll in college and how ready for college students actually are. It is a yawning gap, and Holtz-Eakin rightly laments it. But then he pulls the ol’, “Common Core is a high standard,” and suggests that it will bridge the college prep divide. He even writes that the Core has been “shown” to be “effective.”  

Not only has there been no meaningful evidence of the Core’s effectiveness, but right after I read Holtz-Eakins’ piece I saw that the latest National Assessment of Educational Progress scores had come out – indeed, for the very 12th grade students on the verge of college – and they had dropped in both reading and math between 2013 and 2015, and some dropped going back to 2009. This was, of course, as Common Core was being implemented nationwide. And not only did aggregate scores drop, but also scores for numerous racial and ethnic groups.

Do these results prove that Common Core is either impotent, or worse, a negative force? Certainly not. For one thing, as presented we can’t even break the 12th grade scores out by state as we were able to do with the 4th and 8th grade scores released several months ago. And even that was only able to furnish slightly more nuanced evidence than looking at aggregate national scores. But all these scores do undermine any proclamations of proven Core effectiveness.

Of course, lots of things affect test scores – federal policies, state policies, local policies, economics, demographic changes, etc. – and we can’t ignore all those things and just declare whatever policy we happen to dislike the undisputed villain. But one thing is clear, no matter how you feel about Common Core or anything else: NAEP tests continue to produce awful results for the students who are about to finish K-12 education, whether it is stagnant 17-year-olds’ scores on Long-Term Trend NAEP exams, or these scores for 12th graders on the “Main NAEP.” And this, as I tackle in a new, big update to the Downsizing the Federal Government K-12 page, despite huge increases in spending over the decades, as well as heavily centralized control.

Do the latest NAEP results prove that the Common Core, or centralization more broadly, are bad for American education? No. But they sure don’t help the narrative that centralization, including the federally driven Core, has helped it.

Testing for Core Disruption

It’s been a day since the disappointing “Nation’s Report Card” results came out, and it has given me a chance to crunch some numbers a bit. They don’t tell us anything definitive – there is a lot more that impacts test scores than a policy or two – but it is worth seeing if there are any patterns that might bear further analysis, and it is important to explore emerging theories.

Not surprisingly, while many observers have been rightly hesitant to make grand pronouncements about what the scores mean, some theories revolving around the Common Core have come out. The one I’ve seen the most, coming from people such as U.S. Secretary of Education Arne Duncan and Karen Nussle of the Core-defending Collaborative for Student Success, is that the Core will bring great things, but transitioning to it is disruptive and we should expect to see short-term score drops as a result.

That is plausible, and we can test it a bit by looking at the performance of states (and the Department of Defense Education Activity) that have demonstrated some level of what I’ll call Core aversion. Those are states that (1) hadn’t adopted the Core at the time of the NAEP test; (2) had adopted but had moved away by testing time; and (3) were still using the Core at test time but officially plan to move away. They are broken down in the following table, which uses score changes in the charts found here:

“Nation’s Report Card” Rapid Reaction

This morning the latest scores from the 4th and 8th grade National Assessment of Educational Progress – the so-called Nation’s Report Card – came out, and the story isn’t very good, at least upon first examination. Average scores in 4th and 8th grade math, and in 8th grade reading, were down from 2013, and essentially stagnant in 4th grade reading.  

Of course, there is a lot you cannot tell about school systems from looking just at NAEP scores. Numerous variables that affect academic outcomes, ranging from demographic changes to cultural shifts, can have important impacts on scores. But it is sobering to see national test scores stagnate or drop, and at the very least the scores should put a damper on some of the declarations of success we’ve seen in the past from people like U.S. Secretary of Education Arne Duncan, who in 2013 credited state transitions to the Common Core national curriculum standards for upticks that year.

Perhaps a look at Kentucky, which has been held up as a success story for adopting the Core ahead of all other states and seeing increases on its state tests, is telling. Kentucky may well be seeing improvements, but the NAEP exams, for many people, serve as something of an external audit to see if states’ own tests are producing deceptive information. Of course there can be legitimate disagreements about what test is better – and if testing is even a good way to measures success – but many people who support the Core see state tests as dishonest if they differ markedly in their results from NAEP. So NAEP is important to them. Well, now, while seeing rising scores in 4th grade reading, Kentucky has seen falling scores in 8th grade math and reading, and stagnant scores in 4th grade math. Does that mean the Common Core, or anything else they are doing in Kentucky, necessarily doesn’t work? No. But it does furnish evidence that contradicts the simplistic message of, “Look at Kentucky – the Common Core works!”

There is much that NAEP is too limited to tell us definitively, but the same goes for any single measure of education. And we should be concerned whenever we see scores go down.

Do Racial Disparities Explain Flat Student Performance?

The latest results of the Nation’s Report Card for history, geography, and civics are out, and as usual they are depressing. The exam, formally known as the National Assessment of Educational Progress, is administered to a representative sample of U.S. students to give a snapshot of student performance in a variety of subjects nationwide. Education Week reports:

The nation’s eighth graders have made no academic progress in U.S. history, geography or civics since 2010, according to the latest test results from the National Assessment of Educational Progress (NAEP).

Fewer than one-third of students scored proficient or better on any of the tests and only 3 percent or fewer scored at the advanced level in any of the three subjects.

No significant changes since 2010

However, Chad Aldeman of Bellwether Education Partners argues that saying students “have made no academic progress” is “the wrong way to look at it” because of something called Simpson’s Paradox (which has nothing to do with the voice of Principal Skinner and Mr. Burns turing down a $14 million contract):

Which Schools Best Serve the Public? New U.S. History, Civics Scores Point to Them

The latest 8th grade U.S. history, civics, and geography results from the National Assessment of Educational Progress – the so-called Nation’s Report Card – have been released, and as usual, things seem bleak: only 18 percent of students scored proficient in U.S. history, 23 percent in civics, and 27 percent in geography. These kinds of results, however, should be taken with a few salt grains because we can’t see the full tests, and the setting of proficiency levels can be a bit arbitrary. Also, we don’t…

Oh, the heck with all that. As a fan of school choice, just tell me if private schools did better!

Based on the raw data, they did. 31 percent of private school students were proficient in U.S. history, versus 17 percent of public schoolers; 38 percent were proficient in civics, versus 22 percent of public schools kids; and 44 percent were proficient in geography, versus 25 percent of public schools kids. That said, to really know which broad swath of schools did better – and from a parent’s perspective, it is really only the individual schools from which they might choose that matter – you’d have to control for all sorts of characteristics of their students. From what I’ve seen, what was just released didn’t do that. Thankfully, others have.

What have they found? Controlling for various student characteristics and other factors, private schools beat traditional publics in terms of political knowledge, voluntarism in communities, and other socially desirable outcomes. Why?

There may be many possible reasons, but at least one seems to be intimately connected to choice: autonomous schools select their own curricula, and families willingly accept it when they choose the schools. That means chosen schools can more easily teach coherent U.S. history and civics than can public schools, which often face serious pressures to teach lowest-common-denominator pabulum lest conflict break out among ideologically and politically diverse people. Perhaps ironically – though not if you understand how a free society works – by not being public, private schools may actually serve the public better.

So no, you can’t conclude a lot from the latest NAEP scores. But that doesn’t mean they can’t point you in the right direction.

New National Test Results Released Today

The U.S. Department of Education has just released 2013 results for the National Assessment of Educational Progress—aka, “The Nation’s Report Card.” The scores are for 12th grade reading and mathematics, and neither has changed since the last time they were administered a few years ago. But of course what we really want to know is how well students are performing today compared to those of a generation or two ago. That would tell us if our education system were improving, staying the same, or declining in performance.

The trouble is, “The Nation’s Report Card” doesn’t go back very far.  The reading results reach back to 1992 (since which time, there has been a slight but statistically significant decline), but the math results only reach back to 2005 (since which time, there’s been a slight but statistically significant increase). It’s just not that long of a time period to assess trends.

Wouldn’t it be great if there were a different set of NAEP tests, called the “Long Term Trends” series, that reached back all the way to the early 1970s! And wouldn’t it be even better if we could find out how much we’ve spent per pupil over that same time period, so that we could figure out if our schools are getting more or less efficient with our dollars? Well, what do you know, there is, and we can!

But here’s the thing. Some people look at that national trend chart and think: but my state is doing much better than that! Is it? Is it really? I decided to find out, for all fifty states. The result is my recent, mysteriously-titled paper: State Education Trends. Drop by and check out how your state has done over the past 40 years.

[Note to readers: The state charts look at changes in annual per pupil spending over time, whereas the national chart above looks at the change over time in the total cost of a full K-through-12 education, so the spending trend lines are not directly comparable].

No Compelling Evidence ‘No Child’ Worked

Over the last few days the Wall Street Journal has run two articles suggesting that the No Child Left Behind Act has been somewhat successful. But that’s not supported by the federal government’s own measure, the National Assessment of Educational Progress.

The WSJ’s first article appeared on Saturday, and while focusing on the stagnation of high-achieving students, it asserts that NAEP exams show “dramatic progress—sometimes double-digit increases—for the lowest achievers over the last two decades, especially after No Child Left Behind.”

Last month I debunked the idea that historically struggling groups have seen dramatic improvements under NCLB, laying out the data from numerous NAEP tests. Quite simply, looking at score gains per year, there were many periods before NCLB that saw faster improvements. Below are two more tables from the latest NAEP scores, released a couple of weeks ago. These are for the so-called “main” NAEP, which is not nearly as valuable as the long-term trends exam for seeing historical patterns, but the WSJ cites it and it does contain new information. The results are for the bottom 10 percent of performers.

As always, at what year one could start crediting results to NCLB is debatable. (Actually, you can never simply look at NAEP scores and attribute them to one factor because so many variables influence outcomes.) That date cannot be earlier than 2002, the year the law was enacted, and probably should be 2003, by which time most of the regulations were written and the law began to take real effect. To deal with this problem, the tables include only years that fully include NCLB or do not include it at all. Also note that there are two pre-NCLB time bands for reading because there are no 2000 8th grade reading scores.

Mathematics, 10th Percentile

Reading, 10th Percentile

Once again, there is is no pattern of faster improvement under NCLB than before it. Highlighting periods with greater growth than under NCLB, you can see that in 4th grade math improvements were faster before NCLB than after. In 8th grade math, it’s essentially a dead heat. In 4th grade reading, there’s sizable improvement under NCLB, and in 8th grade reading there’s an appreciable advantage before NCLB.

The second WSJ piece that gives NCLB undue credit is an op-ed from Kevin Chavous. Chavous, a tremendous advocate for school choice, implies that NCLB supplies “accountability” needed to make American kids competitive with their international peers. But as we’ve seen, there’s precious little evidence that NCLB has done anything to improve educational outcomes. Meanwhile, it has cost us a mint, with Department of Education k-12 spending rising from $27.3 billion in 2001 to $37.9 billion in 2011.

Unfortunately, Chavous’s piece seems more aspirational than reality-based, as is often the case in education policy. “We must try to make schools and teachers accountable,” he seems to be saying. “Heaven knows the states won’t do it!”

The need to deal in reality is why Mr. Chavous’ main concern—getting school choice—is so crucial. Government schooling will never be fundamentally changed because those who would be held accountable—teachers, administrators, bureaucrats—have by far the most motivation to be involved in education politics, the greatest ability to organize, and hence the biggest store of political power. Their livelihoods, after all, are at stake. And what do they want? What we’d all probably like: as much pay as possible with as little accountability.

The only way to end employee domination of education is to fundamentally change the system: instead of having politics control schooling, let parents control education money so they can take their children out of schools they don’t like and put them into those they do. Don’t force them to undertake the endless, hopeless warfare of having to form coalitions, try to get politicians’ ears, spur politicians to move and, if they can ever get decent changes, then force them to constantly fight to keep the reforms against opponents with full-time lobbyists and political machines. No, let them vote with their feet, right away, and get their children the education they need.

NCLB is, by most indications, an abject failure, and the very nature of government schooling doomed it to be so.