A Pot and Kettle Memoir: Obama Schools Chief Slams “Lies” about Poor Children’s Progress

by Caleb S. Rossiter

December 6, 2018

In his recent memoir, How Schools Work, Arne Duncan at first seems to be joining the critics of his signature “school reform” of making administrators and teachers “accountable” for the standardized test scores, grades, and graduation rates of low-income students.  The first chapter is titled “Lies, Lies Everywhere,” and it starts with these words: “Education runs on lies.”  Duncan calls out “the big lies…the ones that schools tell to every level of government about how great their students are doing.”  The truth, he says, is that “far too few of our kids are actually prepared.”      

He’s right about that.  The College Board says that half of high school graduates are not “college-ready” in math.  The figure for African-Americans is even worse: after decades of accountability 77 percent are not ready.  Such students are likely to fail their math placement test, and have to take and pass a non-credit but full-cost high school review course.  Only then can they can move on to the basic statistics course a citizen needs to make sense of public policy.  As I’ve seen as both a high school teacher and a college professor, this is the path of dropping out, not graduating.  

Duncan says he first encountered school lies 30 years ago, when during college he tutored at his mother’s after-school program in a poor black neighborhood in Southside Chicago.  Duncan, who is white, also lived on the Southside, near his father’s job as a professor at the elite University of Chicago.  His tutee was a black high school basketball star who assumed that his “B” average guaranteed a college scholarship.  Duncan soon realized that the boy’s pathetic academic level meant he had no hope of even getting into college. 

The memoir makes it clear that schools are still at it, hiding from poor parents their children’s low effort, achievement, and readiness for college or work, which will keep them trapped in the underclass.  That’s a depressing conclusion coming from someone who presided over a generation of accountability policies as head of the Chicago schools and then as President Obama’s secretary of education.

So has Duncan come around to the views of reformer-turned-critic Diane Ravitch, who has argued in a series of books that 30 years of accountability has led to great turmoil but no systematic improvement in student outcomes in the nation’s many segregated, high-poverty schools?  Told under 2001’s bipartisan “No Child Left Behind” law that their jobs depend upon progress on standardized tests and graduation rates, administrators force teachers to distort the curriculum to bump up the scores of students who pre-tests show are close to the all-important, but actually quite minimal, “proficient” score, and enter phony passing grades that lead to fraudulent diplomas. 

Ravitch’s deeper critique of accountability is that making schools a business where the product is scores and rates undercuts the historical and proven purpose of American schooling: to prepare students to be productive citizens of a democracy and, for low-income students, to provide a bridge to the middle class.  Once the “data-driven” business model is adopted, particularly in a segregated system where outcomes vary by neighborhood, education is no longer based on children’s needs and responsibilities.  That’s because scores and rates are used to measure achievement levels of schools and teachers, rather than students.

The pressures are particularly intense in the school district closest to the federal bureaucracy that Duncan ran.  In 2017 National Public Radio reporter Kate McGee caught the District of Columbia’s high-poverty, all-black Ballou high school awarding half of its diplomas to students who rarely attended, let alone achieved basic competency.  A system-wide investigation found that one third of DCPS graduates in 2017 were actually ineligible. 

Graduation rates had risen sharply in DCPS, but only because of hard-to-fail “credit recovery” courses invented by reform superwoman Michelle Rhee, and administrators doctoring attendance records and pressuring teachers to record phony passing grades.  The much-celebrated 100 percent college acceptance for Ballou graduates turned out to be at an open enrollment community college that few of them ended up attending.  Like most of America’s children in high-poverty schools, they’ve been lied to about their college and career readiness.      

Sadly, Duncan’s apparent reversal turns out to be another case of what Columbia Teachers College professor Aaron Pallas has long called “Duncan vs. Duncan.”  Like many prominent reformers, Duncan bemoans the continuing low achievement of poor children, yet argues that only more of the same will change the reality.  As with his “Race to the Top” program, which was sold as an alternative to “No Child” yet made federal funds contingent on districts financially rewarding or punishing teachers for their students’ test scores, Duncan wants to double down on accountability.  The “big lie” he really has in mind is that schools say they are pushing accountability, but actually haven’t implemented it stringently enough! 

In a recent interview in the Atlantic, which is conveniently owned by the sponsor of his foundation, the widow of Apple mogul Steve Jobs, Duncan tells us that, “whenever you have people who say they have it all figured out, I think they probably don’t.”  Unfortunately he isn’t talking about the certainty with which he promotes accountability, but rather the attitudes of his opponents on school reform’s Common Core tests, “white suburban moms who — all of a sudden — their child isn’t as brilliant as they thought they were, and their school isn’t quite as good as they thought they were.”  He gives no credence to parents’ concerns that his department forced on the states another time-consuming, curriculum-distorting, untested fad whose primary function is to provide fodder for misleading claims of school progress.  One-third of states have rejected, repealed, or delayed Common Core and its tests.  

* * *

The Magic Teacher as Savior

Like the Iroquois nations of New York state, school reformers have a founding myth.  In their youth they have phenomenal success working with disadvantaged students, and this drives them to bring their insights to the policy arena, so many more students can be saved.  Former DC schools chief Rhee’s myth was that when she was an untrained new teacher in a blighted Baltimore neighborhood her third-graders’ average score on a national test zoomed from the expected 13th percentile to the amazing 90th.  By chance, an unrelated research project on her school later discovered that there was no class there that did anything of the sort.

Duncan’s myth is similarly inspiring and uncorroborated.  After playing professional basketball for four years after college, he came back to the Southside to work for a wealthy friend who made an “I have a dream” pledge to the 40 sixth-graders at the high-poverty Shakespeare public school.  These pledges, much in vogue in the 1990’s, provided after-school tutoring to students and offered to pay all college expenses if they graduated. 

Duncan had always felt that if he’d gotten to his basketball star six years earlier, he could’ve helped him gain college-ready academic and study skills.  Based on his mom’s afterschool program, he wanted to “demonstrate that with love, support, and high expectations, any kid could succeed.”  Now he had a chance to prove it.  And by golly, he did!

Despite having no teaching experience or training, Duncan ran a program that saved these children from their appointed fate.  He writes that there were virtually no expectations, curriculum, or assignments at Shakepeare, and that many of the sixth-graders “sadly…came to us at a third or fourth-grade level.”  (Actually that level of achievement in sixth grade would be a good thing in most high-poverty schools.) 

But after Duncan’s team made up their own assignments and followed the children throughout the city as they moved to many different schools, “by 8th grade most were at grade level or beyond…(and) the vast majority of them did graduate (on time) in 1998,” compared to the city’s 43 percent graduation rate.  Amazing.   Inspiring.  Or, as a colleague with 35 years of teaching in DC’s high-poverty schools used to whisper under her breath when Rhee’s instructional chief Jason Kamras would foist yet another initiative on us because “the research shows” it would raise test scores, “Bulls---.”  (In the midst of the Ballou scandal, the Richmond, Virginia, schools hired Kamras as superintendent, citing the “progress” in DC.)

The Department of Education’s What Works Clearing-House found no statistical evidence that these “Dream” programs work, so we have reason to be initially skeptical about the Duncan magic.  (ies.ed.gov/ncee/wwc/Docs/InterventionReports/wwc_ihaveadream_031009.pdf)   Even tracking 40 poor, peripatetic families for six years is a dubious claim; a resulting burst in grade level tests and even graduation by weak students from challenged and alienated families, let alone attendance and completion of college, which was the stated goal, is even more suspect. 

Duncan could settle the matter by listing the 40 children with pseudonyms along with their test scores and schools from the 6th and 8th grades and on into high school, the year of their high school diploma, and their college attendance and graduation data.  Researchers could then dig through these data and see if they say what he claims.  He could also ask the 40 students, who are now in their late 30’s, if they would agree to be identified so that researchers could verify the data.  Until then, it’s just another founding myth.   

Humorously, Duncan’s sister and “Dream” co-tutor Sarah seems to be taking the family magic to even greater heights.  She now runs the Network for College Success in Chicago, which advises principals on how to structure their schools.  Its website claims that: “Our work with school leaders has resulted in significant improvements in student achievement.”  Simply having your principal talk with a Duncan brings the boom!

Critics of school reform use the acronym for the National Assessment of Educational Progress to describe the touting of a few increased test results from the many subgroups and grades tested while ignoring the demographic shifts that usually lie behind them: misNAEPery.  Duncan slams administrators for such trickery, but he too picks a smattering of hard-to-confirm figures and hard-to-replicate cases to claim that “education reform is working.”  This is one of the two fundamentally misleading claims in his book. 

Even after the Ballou scandal showed that the emperor’s clothes were at least tattered and torn, Duncan applauds DCPS as a success story for his “grand bargain” of using “more resources for more accountability” so that good teachers can replace bad ones.  Yet DC politicians and school leaders practice heavy misMAEPery, calling DCPS “the fastest-improving urban school district in America” no matter how many times it is demonstrated to them that its fourth-grade NAEP scores went up because of middle and upper-class families streaming into largely-white elementary schools, and the expansion of all-black charter schools.  Its core challenge, improving outcomes in life for poor, black students, is as far from being met than ever.  

Of course, it would take full access to school and demographic data to untangle each of the many and constant assertions of dramatic success.  When thoroughly investigated though, like the “Texas test miracle,” graduation rates in Houston and DC, and the test-erasure cheating scandals of Atlanta, DC, and many other school districts, “unbelievable” results usually turn out to be exactly that.

Statisticians leery of misNAEPery have recently started to use complex models and assumptions that mimic experiments to see if there is a positive effect in adult life from having a certain type of teacher.  One genre of these studies identifies “high-value-added” teachers from students’ test scores and measures their students’ success later in life.  Another genre highlights “racial” bias in teachers by asking them how far they think students will get in college, and then comparing their predictions with the actual outcomes. 

The most powerful case for magic teachers in Duncan’s memoir is from the first genre: “Replacing a bad value-added teacher with a merely average one would increase the lifetime earnings of a classroom by $250,000.”  Wow!  But let’s take a closer look at the correlation that Harvard’s Raj Chetty found between higher-score teachers and students’ projected earnings, based on salary at age 28.  Divide the pot by 28 kids, 40 years of work, 48 weeks per working year, and 5 working days per week and this translates to less than one dollar a day. 

So, we can tell kids that when they grow up they should buy their morning coffee at McDonald’s instead of Duncan Donuts (let alone Starbucks!), or we can upend the entire school system.  Tough choice!  Chetty and some reformers have implied that the effect is cumulative, meaning more score-raising teachers results in more income, but this claim is not justified by the research design.  Duncan could have reported the 1.3 percent salary increase associated with the higher-value teacher, but $250,000 sounds so much better.  This is a classic example from large sample analysis of finding statistical significance (yes, there is a difference) but not policy significance (but it’s not big enough to justify “accountability”).  

Holly Kuzmich, director of the George W. Bush Institute, has claimed that a large-sample study in 2016 of the second genre by American’s Seth Gershenson and Johns Hopkins’ Nicholas Papageorge “found that white teachers on average had significantly lower expectations for black students than they did for white students” and that these expectations “made a real difference in their college attainment.”  But her reading of the study lies between misleading and not true. 

As might be expected, because the black college graduation rate in America is far lower than the white (and in this sample, 30 percent versus 50 percent), all ethnicities of high school teachers correctly predicted that the rate for black students would be lower than that for white ones.  Interestingly, when the researchers statistically controlled for the effect of student, family, and school characteristics, this “best model” found no differences in prediction rates based on race.  This contradicts the notion that teachers judge students based on race, rather than on the complex web of realities associated with race.

In fact, the bias found by the researchers turned out to be mostly a measure of realism.  All ethnicities of teachers in their sample over-predicted college graduation for all ethnicities of students by about 20 percentage points.  For black students, though, non-black teachers and black English teachers were somewhat closer to reality than black math teachers, who wildly over-predicted college graduation for their most dismal achievers.  The most remarkable finding in the study was that having a teacher who, realistically or not, predicted you would finish college had a staggering 15 percentage point impact on actual college graduation rates.  The effect was found only for white students, however.  In any event, the statistical model accounts for too little of the overall variation in results for what the researchers consider a “novel, arguably causal” effect to be considered more than suggestive.

This entire quest for magic teachers as saviors, the Rhee-Duncan youthful myths writ large as national policy, is an indication that we have been reduced to rearranging deck chairs on an educational Titanic.  The failure of accountability to improve high-poverty outcomes shows that it should not have dismissed segregation and poverty as irrelevant excuses.  The quasi-religious belief in magic teachers has the effect of discouraging true reform, both in our schools and in our poverty policies. 

Teacher union boss Al Shanker made much the same point in 1965 when he sarcastically summarized how school critics were willfully misinterpreting the “Pygmalion in the Classroom” study (cited by Gershenson and Papageorge as a precursor), which claimed IQ could be boosted for students just entering elementary school merely by telling teachers that they were going to “bloom” that year:  

If thousands upon thousands of children are not learning to read, write, speak and compute, it is not because of overcrowded classrooms, the effects of poverty and social conditions, poorly developed educational programs and materials and inadequately trained teachers. No, the children are not learning because the teachers don’t expect them to learn.

* * *

Making Segregated Schools Great Again

Duncan’s second misleading claim is the well-worn assertion that, with proper school reform, ZIP code doesn’t matter: “In America, there’s no reason that a kid born in one place should not have the same chances and opportunities as a kid born in another.”  Anyone who disagrees, who sees from experience lots of obvious reasons why, is accused of practicing George W. Bush’s “soft bigotry of low expectations.” 

School reformers have always been schizophrenic in their claims.  They trumpet their belief that great teachers working hard under their regime can eliminate the “achievement gap” even in segregated, high-poverty schools, even as they call for “wrap-around” social programs that imply they recognize that ZIP code is pretty much all that matters.  Duncan was a founder of both wings of the Democratic Party’s school caucus in the 2008 campaign: the accountability types and the poverty types. https://www.educationnext.org/poor-schools-or-poor-kids/

Jobs programs and addiction treatment for young parents; nutrition instruction for pregnant women and young parents; lead removal; safe housing; visiting nurses who help parents learn to stimulate their babies’ cognitive development with the talk and play common in middle-class homes; pre-school programs like Head Start and Early Start; music lessons; class trips to Europe; expensive AP and college board tutors….  If magic teachers can do the trick, if meth babies, crack babies, abandoned babies, and babies in silent households aren’t handicapped by their experiences, why waste money on all this stuff?  

Because poverty is often destiny.  All teachers know kids who have escaped its tentacles, but they know more who haven’t.  On average, children are physically, emotionally, and mentally damaged by poverty, and socially and educationally disengaged from the middle-class methods of successful schooling by their particular culture of poverty. 

Our history of slavery and segregation for black Americans, discrimination and violence against immigrants, and poverty and subsequent alienation have built significant barriers against child development.  Because most poor children are packed into schools with other poor children, with the packing especially intense for black and Hispanic children, the medical, psychological, and developmental effects of poverty overwhelm both academic standards and support services.  Even worse, the culture of poverty, with its attendant alienation, pessimism, even nihilism, can become the dominant school culture, building a “massive resistance” to studying and achieving that Virginia’s opponents of integration who coined the term would have envied.  This reality must be accounted for in our planning if we wish to educate successfully.  Yet the school culture of poverty is never raised, not once, in Duncan’s book. 

Why?  Because the most logical and promising framework for addressing the school culture of poverty, which is ethnic and economic integration, has been successfully resisted by the white and the wealthy ever since the ethnic part (which usually incorporates the economic part anyway) became legal and even mandated in 1954.  The reformers present plans to make segregated schools great again because they know that the American middle-class, let alone the wealthy reformers themselves, will never let the culture of poverty into their kids’ schools.  Indeed, key players who forced accountability on high-poverty public schools have often sent their own children to exclusive private schools or elite public schools that are not affected by it. 

One of Rhee’s children attended Harpeth Hall prep school in suburban Nashville (where Rhee’s ex-husband, also an aggressive reformer, was ironically serving as Tennessee’s top public school official).  Both of them previously attended the Oyster School, a storied bilingual “neighborhood” public school in wealthy, “west of the park” Washington that also accepts “out-of-boundary” Spanish-speakers, many from upper-class families.  The DC public schools are 60 percent black; Oyster is six percent black. 

“No Child” Bush’s twins attended the Preston Hollow public elementary school in wealthy North Dallas, which was cited by a federal judge in 2006 for “segregating” its African-American and Hispanic transfer students from the white neighborhood students.  They then attended the private Hockaday School and, when the Bush became governor, St. Andrew’s Episcopal in Austin before moving on to Austin High.  Duncan’s children attend the private University of Chicago Lab School as he did, and Obama’s went to Lab and then to the Sidwell Friends prep school in Washington. 

Starting in the 1970’s the resistance of white parents to integration was expressed in the South through white academies and in the North through white flight to segregated residential areas and violent anti-bussing demonstrations.  The resistance of middle and upper-class parents to economic integration was just as intense.  Black parents who were finally allowed by affirmative action to compete and succeed moved away from struggling black areas, making the poverty there even more concentrated.  The growing Hispanic population and poor rural whites were also often residentially segregated.

From the 1940’s on, black parents and the civil rights movement had instinctively homed in on what educational research was starting to find: integration of schools can increase the chances that the school culture will increase hard work, achievement, and career and college opportunity for poor and minority children.  It appears, though, that the best results for all children come when these children take up only a minority of the seats – enough to be comfortable and not isolated, but not enough to dominate the academic culture.  (See  nber.org/papers/w16664.pdf,  a comprehensive review of court-ordered desegregation that found dramatic improvements for blacks in years of schooling and college.)

Research by psychologists Kenneth and Mamie Clark, which was used by the Supreme Court to argue in 1954 that separate was inherently unequal for the group being discriminated against, provided the intellectual underpinnings for advocates of integration.  When Martin Luther King Jr. recounted in his Letter from the Birmingham Jail telling his daughter that “Funtown is closed to colored children” and then seeing “the ominous clouds of inferiority beginning to form in her little mental sky,” he was simply offering his own version of the Clarks’ experiment.  A sample of 253 Negro children, in the parlance of the day, preferred a white over a brown doll by a two to one ratio, and chose the brown doll as the one that “looks bad” by a four to one ratio.   

The experiment was suggestive but in no way definitive about black children’s problems with self-image.  Interestingly, one result that the Supreme Court seems to have overlooked actually undercut the NAACP’s argument about segregated education.  The ratios for the Northern half of the children, who attended integrated schools, were far higher than the ratios for the Southern half, living under segregation!  However, the Clarks’ work, which ranged far beyond the simple doll study, did presage the core problem of black, Hispanic, and poor school populations today: awareness of lower societal status and anger or depression about it creep into enough students that a school culture of minimal effort and even resistance develops, particularly after students learn in the elementary years that even with low effort and achievement you’ll be passed on to the next grade.

One of my administrators at a high-poverty high school once correctly reminded me that we were not really teachers of our subjects, but of “middle-class social skills.”  The theory behind integration is that middle-class parents provide not just the initiative and political clout needed to push school officials for proper funding and attention to problems, but the critical mass of students whose middle-class approach to school sets the tone for the entire student body.  

Such a theory is anathema in American politics, of course, because it sounds like we are blaming the victim.  Average SAT scores rise by a standard deviation (100 points) as family income goes from low to high, but at each income level, African-American students also score a standard deviation below whites. (https://randomcriticalanalysis.wordpress.com/2015/11/25/no-the-sat-doesnt-just-measure-income/)  The late Peggy Cooper Cafritz, who founded the public Duke Ellington High School for the Arts and served as president of the Washington, DC, school board, said her goal was for average SAT scores in DCPS to rise to match those of white and wealthy Fairfax County, Virginia.  I would try to explain to her why that couldn’t happen, and of course it didn’t.  (Peggy knew that, of course, but was playing cheerleader, hoping to rally her teachers and students to greater effort.)

So why are scores for the poor and black systematically lower than for the wealthy and white?  If you say the obvious – that it’s the result of slavery, segregation, and poverty – you’re implying that government should do something dramatic about discrimination and poverty, like integrate the schools and subsidize employment.  Politically we can’t have those conversations, so it must all be the teachers’ fault!

Indeed, Shanker’s fears after the Pygmalion study have been realized: one of the dogmas of school reform is now that “implicit bias” on the part of teachers is the reason poor children do poorly compared to wealthier ones, and black and Hispanic children compared to whites and Asian-Americans.  The insights from studies like Gershenson and Papageorge’s have been transformed into attacks on teachers, who are now bombarded with trendy “anti-bias” training to counter their damaging prejudices that somehow created the SAT gaps. 

The notion that teachers cause, rather than respond to, deep societal patterns based on history, class, and culture is patently absurd, but that doesn’t stop school reformers from giving it a shot.  After decades of failure, Duncan and the accountability crowd keep on claiming that magic teachers are the solution to poverty and racism, and count on the compliant media to report their latest misNAEPery as proof, year after year.  This narrative distracts us from the dramatic changes we need in schools and in society.  Duncan’s memoir unwittingly shows us why it’s time to jettison federal control and its accountability model, and design schools to meet the right of students to pursue a rewarding life rather than the need of politicians to claim that “test scores are up!”

* * *