The Kind of Control You Are Attempting…

May 11, 2015

(Guest Post by Matthew Ladner)

Is there much at stake in the fight over academic standards? Studies from no less than Hanushek and Loveless basically show that the standards movement has largely been pushing on a string. There is some evidence that suggests that states that were doing absolutely nothing on testing before NCLB saw above average math gains, but the fact is most states were testing before NCLB, the gains may have been a one time step increase, and evidence linking the quality of standards and/or tests to academic gains is in short supply.

NCLB’s attempt to test the nation’s kids to 100% proficiency (or as Andy Rotherham insists something more like close to it if you read the fine print, which few outside of Andy did) by a date certain ended in tears waivers.

My impression is that the standards movement basically hangs its hat on the Massachusetts experience. Massachusetts has the highest NAEP scores and thus is a good example to study. Massachusetts however introduced a multifaceted reform strategy in the early 1990s, but scholars seem remarkably incurious about which policy changes helped to drive how much improvement. Of course, like the Florida experience, we can never know what policy changes drove aggregate level improvements, but we have a great deal of micro-level evidence on the impact of individual policies. If any of this exists for Massachusetts, I’ve not seen it discussed. Even if we did have a good sense of this based upon a large body of studies, the question of external validity must be considered. Last time I checked MA was one of four states with an average family income for a family of four in the six figures and I’d wager draws an unusually high number of teachers from selective universities.

Why has the standards movement been pushing on a string? No it is not just that states set the test cut scores at incredibly low levels, although they did that:

It’s not just that states held a repulsive 35% of schools responsible for the scores of their special education kids scores in 2009-10, although they did just that:

After all of those things and others most states took the further step of obscuring the results behind a set of fuzzy labels, like Texas:

Some states have pulled this off much better than others, and a high quality system of transparency should be every policymakers goal. The idea that the country has meaningful, widespread “accountability” through state testing is a demonstrably simplistic notion. The greatest trick the devil ever pulled was conflating minimal skills testing in math and reading with robust accountability. While this is obviously absurd given a moment or two of reflection, it is also deeply ingrained in people’s thinking that you can do things like show a legislative committee a chart like the one immediately above, only to have a member of that committee berate you a mere few minutes later that private schools “lack accountability.”

Er, lack accountability compared to what? I may have missed it but I’m putting the number of people in Texas having been held responsible for the state’s 28% reading proficiency rate over/under at zero unless you want to blame it on the kids themselves, most of whom have been labeled “proficient” on state tests that the Wall Street Stock Picking Chicken might pass on a good day (see Figure 1).

Well yes, but the Common Core will fix all of this. Except of course it won’t. If you’ve been paying attention, you may have noticed that states all over the place have been adopting their own tests and cut scores and discussing withdrawing all together.  Meet the new boss, same as the old boss?

The current chaos shares an origin with the wrecking of the NCLB-era state tests. It is the same reason your tax dollars get used to pay farmers not to grow food so that you can pay higher grocery bills.  Agribusiness is organized and politically active, while eaters are disorganized and politically inactive.  Organized/active beats disorganized/inactive 99 times out of a 100.

So in theory, the state sets out grade level academic standards, and then tests children against those standards. Schools thereby follow a coherent flow of content such that you do simple addition before complex addition etc. In theory teachers and schools that fail to teach the standards get held accountable. In theory, there is no unauthorized breeding on Jurassic Park, but…

As long as you are going to have academic standards and tests, you ought to fight not to have horribly deceptive systems. You should rather fight for informative tests and clear labels, but with the full knowledge that the dinosaurs on your island will constantly be breaking out of your fences in any number of ways. They may even convince some people in the leafy suburbs that the substitution of one set of standards and tests for another constitutes oppression, er, somehow…how? I’m not entirely sure but…ah…stick it to THE MAN!

Bureaucratic accountability, in short, will always face severe political limitations, and even under the best of circumstances is no substitute for parents possessing an exit option. Even under the best theoretical systems there will always be kids who would be better off somewhere else for both academic and non-academic reasons. Decentralized accountability works best with transparency to inform choices, but centralized accountability without choice will inevitably face the gravity well of regulatory capture.

The level of control you are attempting is not possible.



Common Core Sophistry is Fun!

May 7, 2015

On Tuesday, the Wall Street Journal published a piece by Jason Riley on “The Soccer Mom Revolt Against Common Core.”  In it, I was quoted offering the analysis that upper-middle class moms were accustomed to having significant control over their kids’ schools.  NCLB may have annoyed these moms with tests that had little meaning for their kids, but by remaining largely agnostic on standards, academic content, and the method of testing, NCLB didn’t interfere with the operational control of suburban moms.

The over-reach of Common Core and federally-sponsored aligned tests is that they impinge to a much greater extent on the operations of schools.  When soccer moms come to school to complain about Rome and Juliet being cut to make way for informational texts, they are being told that the school had no choice in the matter.  Common Core made them do it.  And if they want to do well on the PARCC or Smarter Balanced tests, they have to make these changes.  It doesn’t matter whether CC really requires this change or not.  The issue is that large numbers of upper-middle class parents are being told that they no longer have the same kind of influence over their schools that they are used to having.  And they are pissed.  So, they are starting to boycott the tests.

Not so fast, says Mike Petrilli.  In a post today he argues:

Here’s where Jason’s argument falls apart: Common Core is almost everywhere. Soccer moms are found almost everywhere. Yet the rebellion he describes is limited to one specific area.

As for Jay, maybe the loss of parental control is a real issue, but why do parents in Montclair, for example, feel that their power is being usurped much more so than parents in other states? Again, it can’t be Common Core, or testing, or school accountability policies, because those are almost universal.

Common Core couldn’t explain the opt-outs because they are concentrated in NY and NJ while CC is spread across the country.  The culprit must be the unions, he argues, since they are strong in NY and NJ and managed to enroll these parents in their general fight against accountability.

Let’s try Petrilli’s argument on another situation to see how well it stands up.  The Baltimore riots couldn’t be caused by police abuse, he would have to argue, because the riots are concentrated in Baltimore while police abuse is widespread.  Convinced?

Let’s try another one.  The unions can’t be responsible for the opt-outs because their opposition to accountability is longstanding while the opt-outs are a new phenomenon.  Common Core sophistry is fun!

Of course, mass protests, like opting out or rioting, have to start somewhere even if the source of complaint is widespread.  In addition, agitators typically play a role in motivating and organizing mass protests, but the underlying injury needs to be present or the agitation fails to gain traction.  The unions couldn’t get the soccer moms to opt-out unless they were upset about something.  Before Common Core, the unions tried but failed to elicit upper-middle class action against accountability tests.  Now they are finding a receptive audience.

No amount of sophistry is going to change the political challenge Common Core faces by interfering with soccer moms’ control over local schools.  And no amount of blaming those soccer moms for failing to care about poor and minority students is going to guilt them into surrendering that control.

Begun the First Amendment War Has…

May 5, 2015


(Guest Post by Matthew Ladner)

up two-zip, Texas is.


The Rich Get Richer under Tax Credits-Public School Tax Credits that is

May 4, 2015

(Guest Post by Matthew Ladner)

Arizona passed individual scholarship tax credit donations for children to attend private schools, and for public school extracurricular activities, in 1997. Since that time the newspapers have felled many trees and spilled much each printing columns and letters bewailing the injustice of the private side credits- they are destroying public education, they are going to help rich people send their kids to private school, they are engaged in dark rites to bring Cthulu back into our plane to wreak his horrible revenge on all living things, etc.

This is all nonsense of course– but I think I see now the origin of the “mostly benefiting the rich” narrative- projection. Benefiting the rich far more than the poor is in fact precisely how the public school credit operates. The public school credit goes to support sports, arts, field trips and all of the various things that Jay has been researching lately. The Center for Student Achievement very helpfully crunched the numbers in Arizona Department of Revenue reports and found the following:

So if you are having to squint at your Ipad, the chart has data from both 2005 and 2013, and calculates tax credit revenue by quartile of public school- from the poorest schools (75% and up FRL) to the lowest (< 25% FRL). In 2005, the poorest schools raised a meager $14 per child in tax credit donations, while the wealthiest raised more than 4 times as much at $57 per child.

By 2013, the poorest schools raised a smidge more per student ($16) which is not enough to keep up with inflation. Meanwhile, out in the leafy suburbs, schools collected $96 per pupil. Thus the gap went from $4 for rich kids for every $1 for poor kids, to $6 for rich kids for every $1 for poor kids.

Hmmmm…so the public credit gives to the most to the kids who have the most, gives the least to the kids who have the least.

Well the private school credit might be even worse! Except, it isn’t. All of the corporate scholarship tax credit money is means-tested in Arizona, and some of the individual credit is as well. Even among the individual tax credit money that is not means tested de jure is means tested de facto by the Scholarship Tuition Organizations (STOs). Page 49 of this Arizona Department of Revenue report shows that 70.4% of the original individual tax credit funds (the non-means tested program) go to students with a family income (family of four) of less than $79,000 and 38% of that goes to families making less than $45,000. All of the rest of the money goes to either low/middle income or kids with disabilities.

In fiscal year 2013 STOs raised about $108m from all credits, and we can safely estimate that between 80% to 90% of scholarship funds went to low and middle-income children, which beats not only the stuffing out of the public school credit, but also out of AZ public school system’s spending overall.


Go down or I’ll put you in Expendables 4!


Testimony to the Arkansas Common Core Council

May 4, 2015

Below is the text of the testimony I intend to present to the Arkansas Common Core Council on Wednesday.  The Council is chaired by Lt. Governor Tim Griffin and was charged by the legislature with providing advice on the future of Common Core Standards and PARCC testing.  You can watch videos of the Council testimony and discussions here.


Jay P. Greene’s Testimony to the Arkansas Common Core Council
May 6, 2015

Standards are about what we value. They communicate what we think is important for our children to learn, when they should learn it, and ultimately what kinds of adults we hope they will grow up to be.

Because standards are about values, their content is not merely a technical issue that can be determined by scientific methods. There is no technically correct set of standards, just as there is no technically correct political party or religion. Reasonable people have legitimate differences of opinion about what they want their children taught. A fundamental problem with national standards efforts, like Common Core, is that they are attempting to impose a single vision of a proper education on a large and diverse country with differing views.

National standards can try to produce uniformity out of diversity with some combination of two approaches. They can promote standards that are so bland and ambiguous as to be inoffensive to almost everyone. Or they can force their particular vision on those who believe differently. Either way, national standards, like Common Core, are inappropriate and likely to be ineffective. If national standards embrace a vague consensus, then they make no difference since almost everyone already believes them and is already working toward them. If, on the other hand, national standards attempt to impose their particular vision of a proper education on those with differing visions, then national standards are oppressive and likely to face high levels of resistance and non-compliance. So, national standards are doomed to be either unnecessary or illiberal. Either way, they are wrong.

Some of you may be thinking that education is not entirely about values. Can’t we at least agree, you might be thinking, that all children need to acquire basic competency in literacy and numeracy? And if so, might not standards be helpful in addressing these more technical issues even if they cannot address broader issues of values?

Unfortunately, even when it comes to some of the narrower goals of education, there is no evidence that standards deemed to be higher quality are effective in producing higher levels of literacy and numeracy. I’m aware of four analyses that have examined whether states with standards judged to be better have greater academic achievement. I’ve provided references to these four analyses in the written version of my testimony. None of them show any relationship between the ratings of state standards according to the Fordham Institute and Education Week and each state’s performance on the National Assessment of Educational Progress. I’m not aware of any empirical analyses that show that “better” standards lead to better outcomes for students.

The lack of relationship between the judged quality of state standards and student achievement should raise a number of concerns for this Council. First, it should make you doubt claims about the quality of Common Core standards. How does anyone know whether Common Core standards are good and will contribute to academic achievement if no one has ever found a relationship between Common Core (or any standards for that matter) and student outcomes? Many people claim to be expert judges of the quality of standards but no one’s judgment has been validated by actual improvement in student performance.

Second, perhaps the lack of relationship between expert judgments about the quality of standards and student achievement is explained by the fact that there is not a single path to academic success for all of our incredibly different children. Common Core or other standards might be good for some students in some circumstances, but bad for other students in other situations. The reason why expert claims about the quality of standards have never aligned with student achievement is that there is no single set of standards that could be optimal for promoting even basic literacy and numeracy for all students. Standards, particularly national standards, are then a fool’s enterprise of one size fits none.

Third, the lack of relationship between “better” standards and achievement might be caused by low levels of compliance by schools and educators rather than the unreliable judgment of experts. That is, standards are just a bunch of words in a document. Even if they are the right words and even if one set of words could fit what all children need, there is no assurance that schools or educators would teach to those standards. Schools and educators have their own ideas about the proper goals of education and little can be done to force them to change their practice.

Key backers of Common Core standards are aware of this problem and so the U.S. Department of Education funded the development of new tests that would be aligned with these national standards. If these new tests could detect whether schools and educators were changing their practices in the ways desired by Common Core and if rewards and punishments could be imposed on schools and educators for their compliance with the new standards, then perhaps the empty words of standards could be transformed into a real change in the education system.

The problem with trying to use PARCC or Smarter Balanced tests to drive Common Core changes is that it almost certainly requires more coercion than is politically possible and would be undesirable even if it could be accomplished. If Arkansas tries to use the PARCC test to impose strong enough sanctions on schools and educators to drive changes in their practice, we will witness a well-organized and effective counter-attack from educators and sympathetic parents who will likely neuter those sanctions. If, on the other hand, the consequences of PARCC are roughly the equivalent of double secret probation in the movie, Animal House, then no one has to change practice to align with the new standards.

And even if by some political miracle the new PARCC test could be used to impose tough sanctions on schools and educators who failed to comply with Common Core, it’s a really bad idea to try to run school systems with a test. All sorts of bad things happen when maximizing performance on standardized tests becomes the governing principle of schools. Schools and educators are likely to narrow the curriculum by focusing on tested subjects at the expense of untested ones. If we care at all about the Arts, History, and Science we should oppose trying to run schools with math and ELA tests. And within tested subjects schools and educators are likely to focus narrowly on tested items at the expense of a more complete understanding of math and English.

Common Core is unlikely to produce meaningful changes in practice without an aligned test that punishes schools and educators, but those types of harsh consequences are unlikely to survive the political opposition of educators and parents. And even if PARCC could impose tough consequences to drive changes in practice, the changes would produce a disastrous narrowing in the curriculum of schools.

So what should this Council recommend? Given that there is no technically correct set of standards and given that expert judgment about the quality of standards has never been validated by better student outcomes, there is no reason for Arkansas to defer to the small group of national experts who drafted the Common Core standards. Arkansas policymakers, educators, and parents know as much about effective standards as these self-proclaimed experts. So we should be empowered to write our own standards that reflect our own priorities and values in education. If standards are about values, they should be developed as close to the people to whom they apply as is practical.

But even standards that are developed in a decentralized way will fail to capture all of the legitimate diversity of goals and needs. For that reason, even standards that are developed locally should be humble about what they can accomplish and the extent to which schools and educators ought to change their practice as a result. In the end, it is families, educators, and communities who need to set appropriate goals for individual children, not the state and certainly not the national government or organizations.

Lastly, and perhaps most importantly, we should abandon PARCC and purchase an already-developed, nationally-normed standardized test from ACT or any of the reputable testing companies. The purpose of PARCC is to drive changes in educator behavior in ways that are desired by Common Core. But we should not be using tests aligned with a set of standards to coerce schools and educators to change their practice. What we really need from standardized testing is just information about how our students are performing. This can be accomplished at much lower cost by just buying a nationally-normed test off of the shelf. And lower stakes tests that are primarily about information rather than coercion will produce much less harmful narrowing of the curriculum.
Charlie M. Belin and Brian Kisida, “Science Standards, Science Achievement, and Attitudes About Evolution,” Educational Policy, September 21, 2014.

Eric A. Hanushek, “Is the Common Core Just a Distraction?” Education Next, May 9, 2012.

Tom Loveless, “How Well are American Students Learning?” The 2012 Brown Center Report on American Education, The Brookings Institution, February, 2012.

Grover J. “Russ” Whitehurst, “Don’t Forget Curriculum,” Brown Center Letters on Education, The Brookings Institution, October, 2009.

Indiana Allows Greg to Once Again Put Mathews on the Canvass

April 30, 2015

(Guest Post by Matthew Ladner)

Indiana session has ended with both an increase in the size of the tax credit and separately an increase in the voucher program amounts. For those scoring at home:

AR: New special needs voucher

AZ corporate tax credit improvement

AZ expansion of ESA to tribal areas

Indiana- increase in corporate scholarship credit cap

Indiana-increase in voucher amounts

MS New ESA for special needs students

NV New corporate tax credit

TN New ESA for special needs students

Down goes Frazier Mathews!



Teaching Shakespeare

April 29, 2015

Shakespeare can be taught.  Students don’t even have to know how to read to become familiar with the sound and cadence of Shakespeare’s language.  Watch this video of Brian Cox teaching a friend’s toddler to recite a portion of the “To be, or not to be…” soliloquy.

Meaning and understanding comes later, but familiarity almost certainly makes that easier.  And to grasp the meaning, it would help if only we had more teachers follow the example of Geoffrey Tennant in the great Canadian TV series, Slings and Arrows, and declare: “So, let’s get rid of the curriculum and I think we should just f*ck around with some texts.”  I’ve experienced many an English teacher drain  all of the joy and depth from Shakespeare by mechanically having the class take turns reading passages while ticking off what students need to know for the AP exam.

Watch Geoffrey motivate an accountant from the plastics firm to do a better reading of Macbeth’s “Tomorrow” soliloquy:


Get every new post delivered to your Inbox.

Join 2,827 other followers