Michelle A. Rhee, America’s most famous school reformer, was fully aware of the extent of the problem when she glossed over what appeared to be widespread cheating during her first year as Schools Chancellor in Washington, DC. A long-buried confidential memo from her outside data consultant suggests that the problem was far more serious than kids copying off other kids’ answer sheets. (“191 teachers representing 70 schools”). Twice in just four pages the consultant suggests that Rhee’s own principals, some of whom she had hired, may have been responsible (“Could the erasures in some cases have been done by someone other than the students and the teachers?”).

Rhee has publicly maintained that, if bureaucratic red tape hadn’t gotten in the way, she would have investigated the erasures. For example, in an interview{{1}} conducted for PBS’ “Frontline” before I learned about the confidential memo, Rhee told me, “We kept saying, ‘Okay, we’re going to do this; we just need to have more information.’ And by the time the information was trickling in back and forth, we were about to take the next year’s test. And there was a new superintendent of education that came in at the time. And she said, ‘Okay, well, we’re about to take the next test anyway so let’s just make sure that the proper protocols are in place for next time.’”

At best, that story is misleading.

The rash of “wrong to right” (WTR) erasures was first noticed by the DC official in charge of testing, who, after consulting with the test-maker, asked Rhee to investigate, in November, 2008. Through her data chief, Rhee turned to Dr. Fay G. “Sandy” Sanford for outside analysis.

I have a copy of the memo{{2}} and have confirmed its authenticity with two highly placed and reputable sources. The anonymous source is in DCPS; the other is DC Inspector General Charles Willoughby. A reliable source has confirmed that Rhee and Deputy Chancellor Kaya Henderson discussed the memo in staff gatherings. Sanford came to Washington to present his findings in late January, 2009, after which he wrote his memo.

In response to my request for comment, Rhee issued the following careful statement: “As chancellor I received countless reports, memoranda and presentations. I don’t recall receiving a report from Sandy Sanford regarding erasure data from the DC CAS, but I’m pleased, as has been previously reported, that both inspectors general (DOE and DCPS) reviewed the memo and confirmed my belief that there was no wide spread cheating.” After receiving this statement, I sent her the memo; her spokesman responded by saying that she stood by her earlier statement.

Chancellor Henderson did not respond to my request for a response.

Sanford wanted the memo to be kept confidential. At the top and bottom of each page he wrote “Sensitive Information–Treat as Confidential,” and he urged, “Don’t make hard copies and leave them around.” (The memo.)

The gist of his message: the many ‘wrong to right’ erasures on the students’ answer sheets suggested widespread cheating by adults.

“It is common knowledge in the high-stakes testing community that one of the easiest ways for teachers to artificially inflate student test scores is to erase student wrong responses to multiple choice questions and recode them as correct,” Sanford wrote.

Sanford analyzed the evidence from one school, Aiton, whose scores had jumped by 29 percentiles in reading and 43 percentiles in math and whose staff–from the principal down to the custodians–Rhee had rewarded with $276,265 in bonuses. Answer sheets revealed an average of 5.7 WTR erasures in reading and 6.8 in math, significantly above the district average of 1.7 and 2.3.{{3}}

Sanford, a Marine officer who carved out a post-retirement career in data analysis in California, spelled out the consequences of a cheating scandal. Schools whose rising scores showed they were making “adequate yearly progress” as required by the federal No Child Left Behind Act could “wind up being compromised,” he warned. And what would happen to the hefty bonuses Rhee had already awarded to the principals and teachers at high-achieving schools with equally high erasure rates, Sanford asked? And, Stanford pondered, “What legal options would we have with teachers found guilty of infractions? Could they be fired? Would the teachers’ contract allow it?”{{4}}

While Sanford’s memo doesn’t raise the issue, falsely elevated scores would deny remedial attention to children whose true scores would trigger help. Just how many children could only be determined by an investigation.

Michelle Rhee had to decide whether to investigate aggressively or not. She had publicly promised to make all decisions “in the best interests of children,” and a full-scale investigation would seem to keep that pledge. If cheating were proved, she could fire the offenders and see that students with false scores received the remedial attention they needed. Failing to investigate might be interpreted as a betrayal of children’s interests–if it ever became public knowledge.

*****

The 37-year-old Michelle Rhee had been a surprise choice to lead the schools. After college, she joined Teach for America and taught for three years in a low-income school in Baltimore. After earning a graduate degree in public policy at Harvard, she took{{5}} over a fledgling non-profit that recruits mid-career professionals into teaching, The New Teacher Project. In that role, she eventually ended up supervising 120 employees. As Chancellor, Rhee would be managing a school system with 55,000 students, 11,500 employees and a budget of nearly $200 million.

She surrounded herself with people with no experience running a large urban school system. Her deputy would be her best friend, Kaya Henderson, another former Teach for America corps member who was then Vice President for Strategic Partnerships at TNTP. She would be managing the District’s 11,500 employees.

Her Chief of Data and Accountability would be Erin McGoldrick, whom Rhee had met at Sacramento High School some years earlier and who was an avowed fan of Rhee. A classics major at Notre Dame, McGoldrick also studied public policy at UCLA. Although she was in charge of data analysis at the California Charter Schools Association when Rhee offered her the job, McGoldrick had no experience in Rhee’s ‘data-driven decision making,’ according to several reliable sources.

Rhee selected Jason Kamras, the 2005 National Teacher of the Year and a veteran of seven years in the classroom, to lead what she called her ‘Human Capital Design Team.’ Kamras’ assignments were to design a teacher evaluation system and create a model union contract.

That no one in her inner circle had any experience managing an urban school system did not seem to concern Rhee.

And if inexperience led her astray, Rhee believed that she had a fail-safe system that would steer her back on course, data-driven decision making. “We’re going to be doing parent satisfaction surveys, principal satisfaction surveys, teacher satisfaction surveys, so that we can gauge how good a job we are doing,” she said. There would be no management by hunches or anecdotal evidence–only numbers. “I am a data fiend,” she told me. “Measure everything. Don’t do anything you can’t measure.”

She was determined not to let anything get in her way. “What I am is somebody who is focused on the end result that I think needs to happen,” she told the PBS NewsHour in September, 2007. “If there are rules standing in the way of that, I will question those rules. I will bend those rules.”{{6}}

Rhee said she would be guided by one principle: “I am going to run this district in such a way that is constantly looking out for the best interests of the children.” And she knew that her actions were being watched beyond the District of Columbia. “All the eyes of the country are now on DC,” she said. “I believe that what we are embarking upon is a fight for the lives of children.”

*****

From her first days in Washington, Michelle Rhee had flaunted her inexperience (“I have never run a school district before,” she told her 5,000 teachers at their first meeting.), but here it seems to have hurt her. An experienced educator might well have gone public with the erasures and simply cancelled the results, boldly declaring that, because the interests of children came first, she was ordering retesting, this time with the tightest possible security. Privately, the veteran could have raised the roof, but publicly she would have been a hero.

Getting at the truth would have required bold action. The essential first step: a deep erasure analysis{{7}} to determine whether the erasures showed patterns, because patterns are very strong evidence of collusion. Even with 70 schools involved, that could have been done quietly, but step two–putting people under oath–would have been public. The Mayor and City Council would have to be involved. While that would have been messy, it would have been dramatic evidence that she truly did put the interests of children above those of all adults, including her own.

*****

The model for an effective investigation can be found 640 miles to the south, in Atlanta, Georgia, where an eerily similar situation involving roughly the same number of adults and schools existed. As in Washington, the Atlanta superintendent resisted investigation. As in Washington, an expert was privately asked for his analysis, which was then ignored and kept out of view.{{8}}

Because of aggressive reporting by the Atlanta Journal-Constitution and strong political leadership from two Republican Governors, the situation in Atlanta was investigated from top to bottom. An investigative team led by former Attorney General Mike Bowers and former DeKalb County District Attorney Robert Wilson interviewed more than 2,000 people and reviewed more than 800,000 documents. Because Wilson and Bowers were working with the Georgia Bureau of Investigation, they were able to put people under oath when they questioned them.

Cheating was found to have taken place in 78.6% of the schools investigated. “Superintendent Beverly Hall and her senior staff knew, or should have known, that cheating and other offenses were occurring,” the 813-page report says. Hall, a former National Superintendent of the Year, left the district just before the Governor released the report, which implicated 178 principals and teachers. If convicted, she faces up to 45 years in prison.

According to the July 5, 2011 report, “a culture of fear and conspiracy of silence infected (the Atlanta) school system and kept many teachers from speaking freely about misconduct.”

As Georgia Governor Nathan Deal said when releasing the conclusions, “When test results are falsified and students who have not mastered the necessary material are promoted, our students are harmed, parents lose sight of their child’s true progress, and taxpayers are cheated.”

In an interview in February 2013, Wilson said that he had been following the DCPS story closely. “There’s not a shred of doubt in my mind that adults cheated in Washington,” he said. “The big difference is that nobody in DC wanted to know the truth.”

*****

It’s easy to see how not trying to find out who had done the erasing–burying the problem–was better for Michelle Rhee personally, at least in the short term. She had just handed out over $1.5 million in bonuses in a well-publicized celebration of the test increases{{9}}. She had been praised by presidential candidates Obama and McCain{{10}} in their October debate, and she must have known that she was soon to be on the cover of Time Magazine{{11}}. The public spectacle of an investigation of nearly half of her schools would have tarnished her glowing reputation, especially if the investigators proved that adults cheated–which seems likely given that their jobs depended on raising test scores.

Moreover, a cheating scandal might well have implicated her own “Produce or Else” approach to reform. Early in her first year she met one-on-one with each principal and demanded a written, signed guarantee{{12}} of precisely how many points their DC-CAS scores would increase.

Relying on the DC-CAS{{13}} was not smart policy because it was designed to assess students’ strengths and weaknesses. It did not determine whether students passed or were promoted to the next grade, which meant that many students blew it off.

Putting all her eggs in the DC-CAS basket was a mistake that basic social science warns against. “The more any quantitative social indicator is used for social decision-making, the more subject it will be to corruption pressures and the more apt it will be to distort and corrupt the social processes it is intended to monitor.” That’s Campbell’s Law, formulated in 1976 by esteemed social scientist Donald Campbell (1916-1976) .

Applied to education, it might go this way: “If you base nearly everything–including their jobs–on one test, expect people to cheat.”

And the novice Chancellor was basing nearly everything on the DC-CAS.

*****

Associate Superintendent Francisco Millet sat in on some of the meetings with individual principals. “In that 15-minute period she would ask each one of the principals, ‘When it comes to your test scores, what can you guarantee me?’ And she would write it down. And you could cut through the air with a knife, there was so much tension.”{{14}}

Millet had no doubt that Rhee was sending the message that they would be fired if they didn’t achieve those guarantees. “Absolutely. Principals were scared to death that, if their test scores did not go up, they were going to be fired. And they knew that she could do it.”

Millet, who resigned after Rhee’s first year, is convinced that principals passed the message along. “There was this whole atmosphere of uncertainty. And when principals feel threatened that if their scores don’t go up, what do you think they’re going to bring down the next level, to their teachers? They’re going to make their teachers feel extremely intimidated that if they don’t do better this year than they did last year, there are going to be consequences.” That led to changes in teaching. “Everybody felt this urgency to improve test scores, and there was no focus on instruction,” Millet says. “The entire focus was on improving test scores.”{{15}}

Rhee categorically rejected this interpretation in an interview in September 2011 when I asked her if she had created a ‘climate of fear.’ “No! Absolutely not!,” she exclaimed, adding, “Was there a lot of pressure to improve student achievement levels in the district? Absolutely. A hundred percent. There was a lot of pressure to do that. But I think that somehow that making the leap from that to and therefore you added to it, it’s crazy.”

Perhaps inadvertently, the rookie Chancellor seems to have provided principals with two motives to cheat, a carrot–the possibility of large bonuses–and a stick–the threat of being fired.

DC’s procedures for administering tests–established before Rhee’s arrival–provided multiple opportunities for cheating. According to a veteran principal, “The test booklets came into the school a week before they were given, and they were just in a shrink-wrapped package. The booklets weren’t sealed. They were just wide open. You could just flip through the pages and see what was inside of them.”{{16}} From there, the principal said, it would have been easy to tip off teachers.

A number of teachers, including Martha Harris, a veteran of 46 years in DCPS, told us that some teachers received special treatment. “If you were one of the favorites, you were given (the test) by the head of the testing committee, or someone allowed you to put hands on that test ahead of time.” In short, it would have been easy for teachers to make sure their students knew the right answers ahead of time.

After-the-fact cheating–by erasing and changing answers–was even easier. “The tests would stay in the building for almost two weeks after they were given” so students who had missed a test could make it up. “They were in the building for a good month between arriving about a week ahead of time and finally getting shipped out. It would have been fairly easy for people to sit down and look through the booklets and change answers.”{{17}}

The erasures stayed buried for years. The official who had spotted the problem and urged Rhee to investigate has kept her mouth shut. Five months after she had informed Rhee of the widespread erasures, Deborah Gist resigned to become State Superintendent in Rhode Island. Rhee now publicly praises her efforts there.{{18}} Sandy Sanford, who earned roughly $9,000 for his work on the memo, has been paid at least $220,000 by DCPS for various services.{{19}}

When erasures continued in Rhee’s second and third years at slightly diminished rates, she and Henderson contracted for three severely limited investigations, none of which allowed for erasure analysis or an examination of the original answer sheets.

Two were performed by Caveon, a Colorado-based company. The other was performed by Alvarez and Marsal, a firm that usually coaches corporations on how to improve profit margins{{20}}. D.C. officials set limits on investigations, never insisting on the obvious essential step of erasure analysis; they dictated which schools should be investigated and even suggested the questions to be asked. A D.C. official sat in on many of the interviews with staffers. No erasure analysis has ever been performed. Caveon’s president, John Fremer, later told the Washington Post and USA Today that it had performed ‘a security audit’ and not an investigation.

Caveon’s 2009 inquiry turned up no cheating. Its 2010 investigation fingered three adults — from three different schools. One of them, a first-year teacher, confessed that he had stood over some of his pupils and coached them until they penciled in the right answers. His explanation: That was the way testing was conducted at his school. He lost his job.

The situation came close to exploding in March 2011 when USA Today blew the whistle on the erasures. The newspaper’s investigative team {{21}} reported that the odds against the 2008 wrong-to-right erasures having happened by chance in some of the schools were greater than the odds of winning the Powerball.{{22}} Tom Haladyna, a professor emeritus at Arizona State who has spent decades investigating cheating, told the newspaper that the score gains reported at DCPS were implausible, observing that “a slow runner can improve a little in each race he runs, but he’s not going to set a new world record.” And some of those score gains were akin to setting a new world record or “losing a hundred pounds a month on a new diet,” Haladyna said.{{23}}

Even then there was no full investigation. Chancellor Henderson somehow persuaded DC’s Inspector General to investigate the matter without looking into the 2008 erasures. He spent 17 months on the case, during which time he interviewed only 60 people from just one school (even though by then more than 90 schools had been implicated).{{24}} When I asked Mr. Willoughby why he had not looked into the first year or at other schools, all he said was that it was not “a fishing expedition,” adding “We stand by our report.”

Choosing to bury the problem and minimize investigations{{25}} allowed Rhee to continue with her radical makeover of the low-performing DC public school system. She extended her ‘produce or else’ approach to teachers{{26}} and continued to remove or reward principals based on DC-CAS scores. In 2010, Rhee confidently predicted that, within five years, the D.C. school system would be “the highest performing urban school district in the country and one that has the faith and confidence of the citizens of the city.”

Her policies remained in force even after she left DC in October 2010 to start, as she proclaimed on Oprah, “a revolution on behalf of America’s children.” Through her well-financed “StudentsFirst” lobbying non-profit organization, she began crisscrossing the nation, urging governors and legislators to do what she did in Washington.

She has been remarkably successful. At least 25 states have adopted her ‘produce or else’ test-score based system of evaluating teachers.{{27}}

But politicians (and citizens) in those 25 states might want to take a closer look at what she actually accomplished. Sadly, DC’s schools are worse by almost every conceivable measure.

For teachers, DCPS has become a revolving door. Half of all newly hired teachers (both rookies and experienced teachers) leave within two years; by contrast, the national average is said to be between three and five years.{{28}}

It was a revolving door for principals as well. Rhee appointed 91 principals in her three years as chancellor, 39 of whom no longer held those jobs in August 2010. Some left on their own; others, on one-year contracts, were fired for not producing quickly enough.{{29}} She also fired more than 600 teachers.{{30}}

Child psychiatrists have long known that, to succeed, children need stability. Because many of the District’s children face multiple stresses at home and in their neighborhoods, schools are often that rock. However, in Rhee’s tumultuous reign, thousands of students attended schools where teachers and principals were essentially interchangeable parts, a situation that must have contributed to the instability rather than alleviating it.

The teacher evaluation system that Rhee instituted designates some teachers as ‘highly effective,’ but, despite awarding substantial bonuses and having the highest salary schedule in the region, DCPS is having difficulty retaining these teachers, 44% of whom say they do not feel valued by DCPS.{{31}}

Although Rhee removed about 100 central office personnel in her first year, the central office today is considerably larger, with more administrators per teachers than any district surrounding DC. In fact, the surrounding districts seem to have reduced their central office staff, while DC’s grew.{{32}} The greatest growth in DCPS has been in the number of employees making $100,000 or more per year, from 35 to 99.{{33}}Per pupil expenditures have risen sharply, from $13,830 per student to $17,574, an increase of 27%, compared to 10% inflation in the Washington-Baltimore region.{{34}}

A comparison of pre- and post-Rhee DC-CAS scores shows little or no gain, and most of the scores at 12 of the 14 highest ‘wrong to right’ erasure schools are now lower. Take Aiton Elementary, the school that Sanford wrote about: The year before Rhee arrived, 18% of Aiton students scored proficient in math and 31% in reading. Scores soared to over 60% during the ‘high erasure’ years, but today both reading and math scores are more than 40 percentile points lower.{{35}}

Enrollment declined on Rhee’s watch and has continued under Henderson, as families enrolled their children in charter schools or moved to the suburbs. The year before Rhee arrived, DCPS had 52,191 students. Today it enrolls about 45,000, a loss of roughly 13%.{{36}}

Even students who remained seem to be voting with their feet, because truancy in DC is a “crisis” situation{{37}}, and Washington’s high school graduation rate is the lowest in the nation.{{38}}

Rhee and her admirers point to increases on the National Assessment of Educational Progress, an exam given every two years to a sample of students under the tightest possible security. And while NAEP scores did go up, they rose in roughly the same amount as they had under Rhee’s predecessor, and Washington remains at or near the bottom on that national measure.{{39}}

The most disturbing effect of Rhee’s reform effort is the widened gap in academic performance between low-income and upper-income students, a meaningful statistic in Washington, DC because race and income are highly correlated. On the most recent NAEP test (2011) only about 10% of low income students in grades 4 and 8 scored ‘proficient’ in reading and math. Since 2007, the performance gap has increased by 29% in 8th grade reading, by 44% in 4th grade reading, by 45% in 8th grade math, and by 72% in 4th grade math. Although these numbers are also influenced by changes in high- and low-income populations, the gaps are so extreme that is seems clear that low-income students, most of them African-American, did not fare well during Rhee’s time in Washington.{{40}}

*****

It’s 2013. Is there any point to investigating probable cheating that occurred in 2008, 2009 and 2010? After all, the children who received inflated scores can’t get a ‘do-over,’ and it’s probably too late to claw back bonuses from adults who cheated, even if they could be identified. While erasure analysis would reveal the extent of cheating, what deserves careful scrutiny is the behavior of the leadership when it learned that a significant number of adults were probably cheating, because five years later, Rhee’s former deputy is in charge of public schools, and Rhee continues her efforts to persuade states and districts to adopt her approach to education reform–an approach, the evidence indicates, did little or nothing to improve the public schools in our nation’s capital.

This story is bound to remind old Washington hands of Watergate and Senator Howard Baker’s famous question, “What did the President know and when did he know it?” It has a memo that answers an echo of Baker’s question, “What did Michelle know, and when did she know it?” And the entire sordid story recalls the lesson of Watergate, “It’s not the crime; it’s the coverup.”

That Michelle Rhee named her new organization “StudentsFirst” is beyond ironic.

—————-

[[1]]1. Conducted September 2011 when Rhee appeared at NBC’s “Education Nation.”[[1]]

[[2]] 2. I first got wind of the Sanford memo while working on a documentary about Rhee for the PBS series “Frontline.” I spoke with Sanford, who confirmed that he had indeed written a memo but could not turn it over to me because it was “a work for hire.” Producer Michael Joseloff and I filed Freedom of Information Act requests with DCPS, which denied its existence, and with two Inspectors General, who acknowledged that they had it in their possession but refused to turn it over on grounds that the memo was part of the “deliberative process” within the D.C. government and thus not available to the public.

Not long after I blogged about “The Missing Memo” on January 15, 2013,it arrived in a plain white envelope. A second copy–perhaps from the same leaker–arrived in late March.

The original leaker of the memo added a note: “You’ve made some folks here nervous, but Rhee, Henderson and Kamras will all deny knowing anything about what Erin worked on.”[[2]]

[[3]]3. The differences, measured in standard deviations from the norm, were astonishing. Experts say that three standard deviations is a ‘red flag’ and four or more are prima facie evidence of cheating. At some DCPS schools the WTR erasure rates were five, six and even seven standard deviations from the norm.[[3]]

[[4]]4. Apparently Dr. Sanford expected Rhee to act decisively. In his memo he recommends that, ”as soon as you think it advisable, someone contacts the legal people and asks them to determine what possible actions can be taken against identified offenders.” As far as we could determine, Rhee did not follow this recommendation.[[4]]

[[5]] 5. Her claim to have founded TNTP is disputed by people in a position to know. It began within Teach for America and was the brainchild of TFA founder Wendy Kopp.[[5]]

[[6]]6. Rhee’s biographer, Richard Whitmire, has characterized her as ‘a zealot’ who believes in the absolute rightness of what she is doing. His book is The Bee Eater.

She also declared–proudly–that she was not inclined to look back. Shortly after she fired a principal she had appointed just a few weeks earlier (replacing one she had fired), I asked her if she had any regrets about her actions. “I’m a very unusual person in that, in my entire life, I don’t have any regrets. I’m a person without regret. Now, are there things that I could have done differently? And if I had to rewrite it, you know, I would have, you know, done it with a smarter way or whatnot, yeah. There are absolutely things that I could have done better. But regrets? No.” She said that in December 2007. Would she say that today?[[6]]

[[7]]7. At its simplest level, investigators can determine which kinds of questions had their answers erased, because multiple-choice tests are planned carefully in include some easy questions, some hard questions, and some of a moderate degree of difficulty. Students would be expected to answer the easy questions correctly and make their mistakes on the harder ones. If answer sheets showed that students with lots of erasures consistently got the difficult questions right–because of WTR erasures–but messed up on the easy ones, that is evidence of tampering. If in one classroom almost all the students erased the same answers in a WTR pattern, that is evidence of tampering, probably by one individual. Sophisticated analysis will reveal whether several classrooms display a identical patterns, which would be evidence of coordinated cheating, a so-called “erasure party.”[[7]]

[[8]]8. Professor Andrew Porter of the University of Pennsylvania did the analysis for Hall.[[8]]

[[9]] 9. August 8, 2008. When those results arrived, Rhee, her deputies and the principals who later would receive $10,000 bonuses gathered in her meeting room. One person who was present described the festive atmosphere. “We were euphoric. Michelle gave a rousing speech, and we drank sparkling cider out of plastic Champagne glasses.”[[9]]

[[10]] 10.October 16, 2008. OBAMA: I’ll just make a quick comment about vouchers in D.C. Senator McCain’s absolutely right: The D.C. school system is in terrible shape, and it has been for a very long time. And we’ve got a wonderful new superintendent there who’s working very hard with the young mayor there to try…

MCCAIN: Who supports vouchers.

OBAMA: … who initiated — actually, supports charters.

MCCAIN: She supports vouchers, also.[[10]]

^[[11]] 11. She would be pictured in an empty classroom holding a broom, on the December 8th issue. The photo caused an uproar, implying as it did that the solution was to sweep out the bad teachers.[[11]]

^[[12]] 12. At least one dozen principals described this process, including the written guarantee, in interviews for “Frontline.” Several principals apparently chose not to sign and instead submitted their resignations.[[12]]

^[[13]] 13.The DC-CAS was an assessment tool with roughly 50 multiple choice questions in both reading and math. The scores did not determine whether kids passed or failed. In the parlance of education, it was not a ‘high-stakes’ test for students, and so many–perhaps most–kids did not care a whit about it. Some slept through the testing, and others filled in bubbles randomly.

Their indifference led principals to take extreme measures to try to get students to take the test seriously. Some principals offered cash to students who promised to try hard. One pledged to get a tattoo if his students did well on the DC-CAS. They would get to pick the tattoo; he would pick the spot on his body where it would be permanently engraved. (Even that desperate strategy did not work!)

The 2008 version had 45 multiple choice questions in reading and 51 in math. Like most standardized tests, some questions were relatively easy, a few very difficult. Designed to assess student strengths and weaknesses, the DC-CAS results would reveal which students needed remedial help in multiplying with fractions or identifying adverbs, for example. The test has four performance levels: below basic, basic, proficient and advanced. Teachers would get that information and be able provide the individual attention–if the system worked as designed.[[13]]

^[[14]] 14. Interviewed at his home in Dallas, Texas, where he moved after resigning.[[14]]

^[[15]] 15. I spent enough time in classrooms during Rhee’s first year to support his observation. We saw hours and hours of drill on basic concepts, and we heard parental complaints like this one: “There’s no attention to the actual depth of instruction going on in the classroom. What we’ve seen is the quick-fix solutions of choosing to work intensively with kids who you knew could boost their test scores if they just got a few extra hours of instruction on particular glitches in their own test results.” She said that her daughter spent nearly three weeks of school on practice testing.

Did she mean that her daughter spent the day filling in bubbles on sample answer sheets? “Exactly,” she answered. “I’m not joking. That’s what you do. You take tests. You take practice tests.”

Martha Harris, the 46-year veteran, agreed. “That’s all we did,” she said. “There came a point in time where we almost abandoned teaching. After Christmas, there was nothing but test prep.” The DC-CAS is given in April.[[15]]

^[[16]] 16. The interview was conducted in shadow to shield the principal’s identity. We did not use it on “Frontline.”[[16]]

^[[17]] 17. The Frontline broadcast of January 8, 2013 brought one individual’s complaint about cheating into focus. Adell Cothorne, the former principal of Noyes Education Campus, went public with her story of an apparent ‘erasure party.’. “I walked into the room and I saw three staff members,” Cothorne told Frontline. “There were test books everywhere, over 200 test books spread out on desks, spread out on tables. One staff member was sitting at a desk and had an eraser. And then there were two other staff members at a round table and they had test books out in front of them. And one staff member said to me, in a light-hearted sort of way, ‘Oh, Principal, I can’t believe this kid drew a spider on the test and I have to erase it.’”
Cothorne was in her first year at Noyes, appointed because Rhee had promoted her predecessor, Wayne Ryan, to her central office. Under Ryan, Noyes had achieved remarkable growth in DC-CAS scores, and Rhee had (literally) made him her poster child for recruitment. In 2007, for example, only 44.14% of Noyes’ students had scored at a proficient level in reading, but under Ryan’s leadership that number nearly doubled, to 84.21%, in just two years. Math scores had also nearly doubled, from 34.24% to 62.79.

(What Cothorne did not know–could not have known–was that a lot of answers had been changed from ‘wrong’ to right’ on DC-CAS answer sheets from Noyes. At Noyes 75% of the classrooms had been flagged for high erasure rates.)

Education enabled Adell Cothorne to rise from the unpromising circumstances–child of a teenage mother, raised in poverty. She rose to become an Assistant Principal in Montgomery County, Maryland, a wealthy suburb of the Capital with excellent ‘Blue Ribbon’ public schools. It’s one of the top-ranked school districts in the nation.

Her idealism burned brightly, and so she applied for a job in Washington, largely, she told Frontline, because of her admiration for Michelle Rhee.

“I still have the Time Magazine with Michelle Rhee on the cover,” Cothorne told Frontline. “I had been following her for a while, and I admired what I saw on the media and the news. And so to have the opportunity to dialogue and sit across from her and then have her say to me, you know, ‘It’s not a matter of when you’re coming to D.C., but where I’m going to put you,’ that was absolute confirmation for me. And I was over the top.”

Rhee installed Cothorne as principal of Noyes Education Campus, a ‘Blue Ribbon’ school, and in October she met with the Chancellor, one-on-one. At that meeting she promised Rhee a 6% gain in math and a 7% gain in reading.

Even as she was making that commitment, Cothorne knew she had a problem. What she had already seen in her new school did not jibe with the test scores that had been recorded, she told Frontline: “As any good administrator should, I visited classrooms and just made my presence known, (and) noticed a disconnect for myself and what was going on in the classroom. The level of instruction, because I’ve worked at Blue Ribbon Schools before, so the level of instruction that I know is needed for a Blue Ribbon School, I was not seeing on a daily ongoing basis. … There’s these huge disconnects. They’re struggling academically. Yet the data that I have been given is showing great gains. But what I see with my own eyes on a daily basis is not a true picture of great gains.”

The apparent ‘erasure party’ she had walked in on galvanized Cothorne. She arranged for tightened security. She changed the locks and had four additional people to monitor classrooms during the tests.

With heightened security, Noyes’ DC-CAS scores dropped 52 points in reading (from 84.21% to 32.40) and 34 points in math (from 62.79% to 28.17% in math). That meant that in 2010-2011 Noyes performed significantly below its 2007, pre-Rhee, level.

Those numbers, she told Frontline, were the true test scores. “Those were what the students in that school actually were able to produce,” she said.

Adell Cothorne, who was completing work on her doctorate, resigned her principalship and gave up an annual salary of roughly $130,000. She has since opened up a bakery, “Cooks ‘n Cakes,” in Ellicott City, Maryland. Her whistleblower suit against DCPS was dismissed in January 2013. [[17]]

^[[18]] 18. In her Students First rating system, Rhee gave the highest marks to Rhode Island, Louisiana and the District of Columbia. She was interviewed on ‘Morning Joe’ on January 7, 2013 and publicly praised Gist, the only person she mentioned by name in the interview.[[18]]

^[[19]] 19. From documents released by DCPS in response to FOIA requests.[[19]]

[[20]] 20. A&M was technically hired by OSSE, not DCPS, but that is a distinction without much of a difference because OSSE’s leader was in synch with DCPS. A&M conducted interviews in those 60 classrooms and, in June 2012, reported it had found “definitive test tampering” in only two of them.

In one classroom at Martin Luther King Elementary, two students reported that their teacher pointed out correct answers; another student in the same classroom reported that the proctor had read answers aloud, raising his/her voice to indicate the correct answer. But, A&M said, the latter charge was “not corroborated” by either of the other two students interviewed from that classroom. A&M apparently interviewed only three students in the class.

At another school, the Langdon Education Campus, a K-8 school, two students reported that the adult administering the test provided assistance to them and other students during the test. Minor infractions — missing security binders and the like — were found at 15 other schools.

Were the adults fired? The report issued by OSSE said that, at both schools, “personnel action for implicated staff members” was pending.

Among the 13 schools where A&M said it found no evidence of any violations was J.O. Wilson Elementary, where in one preceding year every classroom that gave tests had been flagged for abnormally high erasure rates. On the 2011 test, one Wilson classroom registered an average of 9 erasures per student, compared to a district average of 1 erasure per student on that same test. A&M did not explain why it found nothing amiss at Wilson, at least not in the summary made public by OSSE.

A&M was not given access to test answer sheets and did not perform an electronic analysis.[[20]]

[[21]] 21. Reported by Jack Gillum and Marisol Bello under the direction of Linda Mathews[[21]]

^[[22]] 22. What USA Today reported in March 2011 was truly staggering. Between 2008 and 2010, nearly one-quarter of the nearly 3,000 classrooms tested on the DC CAS showed high rates of wrong-to-right erasures. Such erasures persisted at schools like J.O. Wilson, a primary school in the city’s gentrifying H Street corridor, in all of the classrooms tested. At Noyes Education Campus, nearly every classroom had been flagged — the same year Ryan and DCPS officials were filling out their federal Blue Ribbon School application. All told, those erasures were the kinds of anomalies that Rhee’s hired consultant, CTB/McGraw-Hill and the District’s own officials said could be signs of malfeasance.

Just how significant were the erasures? The rates at which students’ answers were erased and changed to the correct answer were enough to make a statistician squirm. In spreadsheets sent to DCPS, classroom after classroom showed remarkable changes — leaving a smudged erasure mark behind while changing the answer to the correct choice wholesale. The odds that these changes were happening–over and over again–by chance, statisticians told USA Today, were greater than the odds of winning the Powerball.

The mathematics weren’t enough to convince Victor Reinoso, DC’s deputy mayor for education. He–and later Rhee–interpreted McGraw-Hill’s memo on the subject to mean that erasures alone were not enough to draw conclusions about cheating–and therefore there was no reason to investigate. The logic is remarkably convoluted: McGraw-Hill said that the data alone didn’t prove cheating but could be used for ‘follow-up investigation.’ However, following up was apparently the last thing anyone in power wanted to do.

(Some public charter schools — which were also subject to DC testing protocols — responded to erasures by launching inquiries to find out what had happened. DCPS responded by tightening test security for future tests and hoped it wouldn’t happen again.)

The newspaper printed the back-and-forth correspondence between Rhee and the testing supervisor, with Rhee asking first for more time and then for more information. However, USA Today’s reporters did not know of Sandy Sanford’s confidential memo. Rhee, by then leading her new organization, StudentsFirst, initially dismissed USA Today’s reporting as racist because, she said, it implied that poor Black children could not achieve. She subsequently apologized for the comments but did not call for an investigation.[[22]]

^[[23]] 23. Cited by Linda Mathews, in a personal communication [[23]]

^[[24]] 24. The Inspector General’s investigation is remarkable for what it did not investigate. Inspector General Charles Willoughby chose not to investigate Rhee’s first year, the year with the most erasures. He chose not to investigate Aiton, the school Sanford had singled out for special attention because of its high wrong to right erasures. He did not examine the test answer sheets or perform an electronic analysis. And he did not investigate J.O Wilson – a school with excessive WTR erasures in 100% of its classrooms – simply because the School Chancellor had assured him that it was a good school!

Although more than half of DC’s schools had been implicated, he focused only on Noyes Education Campus, the school that USA Today had made the centerpiece of its investigation. Over the course of the next 17 months, his team interviewed just 60 administrators, teachers, parents and teachers (Atlanta investigators interviewed over 2,000), all from Noyes Education Campus. Rather than seek outside experts (as Atlanta investigators had), he relied on information from Caveon and Alvarez & Marsal, two firms in the employ of DCPS. He did not ask to perform erasure analysis but relied on interviews–sometimes conducted over the phone.

Without the power to put people under oath, he told City Council member Kenyan McDuffie that asked them if they had cheated. If they said they hadn’t, that was the end of it, because, he explained, he “wasn’t conducting a fishing expedition.” Test monitors sent by the central office to patrol Noyes for the 2010 test told Willoughby that they had been barred from entering classrooms. School officials denied that charge–and Willoughby believed them, not the monitors.

Willoughby was called to testify on February 21, 2013, by McDuffie (who remains the only elected official in DC to show official interest in the erasures). McDuffie asked the IG why he had not looked at other high-erasure schools. “Because we didn’t find evidence of a conspiracy to cheat at Noyes,” he replied. Was it prudent to take the word of two firms that were paid by DCPS instead of seeking an outside, independent opinion, McDuffie asked. “Yes,” Willoughby replied.

Asked by McDuffie if he had tried to find an explanation for the pronounced test score drops when security was tightened, Willoughby replied, “We were told that it was caused by an influx of new students.” His 17-month investigation resulted in a 14-page report, which he released August 8, 2012. (The Atlanta report runs 413 pages.) He found no evidence of widespread cheating at Noyes but cited some security concerns and noted that one teacher had been dismissed for coaching students on a test. The IG’s essential message: except for that one teacher, all was well.[[24]]

^[[25]] 25. Henderson praised the IG’s report as exculpatory, and the Washington Post’s editorial page chimed its agreement.

Henderson and Rhee maintain that the investigations were thorough. After our Frontline in January 2013, Henderson issued this statement: “Since 2009, there have been multiple investigations looking into these allegations, including one by the DC Office of the Inspector General and one by the US Department of Education Inspector General. There was an instance of cheating at Noyes and the individual who was found to be guilty was terminated. All of the investigations have concluded in the same way that there is no widespread cheating at DC Public Schools.”

Rhee’s (and Henderson’s defense) against widespread charges of cheating is to cite these five investigations and claim that they turned up no credible evidence. That’s what Rhee did on February 8, 2013, at a Politico breakfast. She told the audience that the five investigations “found that there was some cheating, but that it was isolated to only a few schools. For some people, it puts it to bed. For others, it’ll never be put to bed — it is what is,” Rhee said. The cheaters, Rhee says or implies, were rogue teachers, never principals.

However, none of the five investigations meet the criteria for a serious in-depth investigation. In most cases, DCPS decreed which schools and which classrooms within those schools would be scrutinized. The investigating companies were not given access to answer sheets or the electronic tapes of the answer sheets.

And the all-important answer sheets from 2008 still have not been examined.[[25]]

^[[26]] 26. She imposed a system known as IMPACT that evaluated teachers based on three observations (two by her own staff of ‘master educators’) and DC-CAS scores. The scores counted for 50% of a teacher’s rating. IMPACT used a 1-4 scale, and teachers receiving a 1 were dismissed, regardless of seniority or tenure [[26]]

^[[27]] 27. Various news reports and Rhee’s own website have given this number. The federal government’s “Race to the Top” program also requires states to use student test scores in evaluating teachers.[[27]]

^[[28]] 28. Highly regarded data analyst Mary Levy has done thorough analysis of virtually every aspect of DCPS, including teacher turnover. She is often called upon by the City Council to provide expert testimony, and reporters have also learned they can trust her analysis.[[28]]

^[[29]] 29. Jay Mathews, the Washington Post, August 23, 2010.[[29]]

^[[30]] 30. http://www.nytimes.com/2010/07/24/education/24teachers.html?adxnnl=1&adxnnlx=1363741389-HfYwIQc6OTSizbC97ZZpTg. Tamar Lewin reports that 682 teachers were fired during Rhee’s time in Washington. Her source is a DCPS spokesperson.[[30]]

^[[31]] 31. Data provided by Mary Levy.[[31]]

^[[32]] 32. I called or wrote to every surrounding school district in late 2011. All said their central office staff had been reduced. According to Levy, Rhee’s central office reduction numbers involved some sleight of hand, because she moved some bureaucrats over to the central office of the State Superintendent (OSSE), which means that she did not actually cut the number of central office personnel as much as reported but instead moved their desks.[[32]]

^[[33]] 33. Mary Levy [[33]]

^[[34]] 34. Mary Levy [[34]]

^[[35]] 35. On the 2011 DC-CAS (the most recent) 21% of Aiton’s students were proficient in reading, and just 17% in math. The DCPS website has school by school comparisons as well as year by year. http://nclb.osse.dc.gov/ [[35]]

^[[36]] 36. DCPS website data [[36]]

^[[37]] 37. Washington Post, November 9, 2012. “D.C. Schools Chancellor Kaya Henderson said Thursday that the school system’s high truancy rates amount to an educational “crisis,” as D.C. officials disclosed that more than 40 percent of the students at Ballou, Anacostia, Spingarn and Roosevelt high schools missed at least a month of school last year because of unexcused absences.”[[37]]

^[[38]] 38. 59%. DC and Nevada have gone back and forth for the dubious distinction of being at the bottom. http://nces.ed.gov/programs/coe/indicator_scr.asp [[38]]

^[[39]] 39. NAEP data can be difficult to interpret in the case of Washington, which is–for statistical purposes–both a city and a state. NAEP has two separate data tracks, its measurement of state-by-state performance and its Trial Urban District Assessment, or TUDA. http://nces.ed.gov/nationsreportcard/pubs/dst2011/2012455.asp.

Washington has improved, as Rhee’s supporters say, but, as her detractors point out, the District was so low that the increases are not a reason for great celebration and they are fairly consistent with the gains made under Rhee’s predecessor, Clifford Janey.[[39]]

^[[40]] 40. Mary Levy and Tom Loveless, personal communication. Dr. Loveless has some qualms about this interpretation as a general rule, but the huge disparities in this case indicate that something significant has taken place. [[40]]