How D.C. Became the Darling of Education Reform

When it comes to education reform, perhaps no city has inspired more controversy and acclaim over the last decade than Washington, D.C. Even today, uttering the name "Michelle Rhee"-the city's first schools chancellor appointed in 2007 after a major shakeup in the district-still evokes heated reactions from local residents. Following the dissolution of the local school board and the centralization of education decision-making within the mayor's office, then-Mayor Adrian Fenty commanded an unusual amount of power to change D.C.'s schools.

Over the past ten years, the policies undergirding the national education reform movement-offering more school choice, weakening teacher union power, and creating new accountability systems (with incentives like pay-for-performance and teacher evaluations based partly on student test scores)-have taken hold in the nation's capital. Some see these moves as encouraging proof that education reform is working. Proponents point to positive benchmarks: District enrollment is growing; D.C. scores on the National Assessment of Educational Progress (NAEP) have improved (in some cases at a much faster rate than students in other large urban districts); and teachers who left the district after receiving low marks on D.C.'s new teacher evaluation system were replaced with higher-scoring teachers who boosted student achievement.

Research suggests that D.C. charter schools have made strides in student learning compared with the city's traditional public schools, and the city's overall test gains cannot be explained by demographic changes alone. In 2016, Jonathan Chait, a liberal writer for New York magazine (whose wife helped craft some of D.C.'s new policies and now works for a local charter school), declared, "The dramatic improvements registered in places like Washington show the revolutionary possibilities of education reform."

For others, these gains have been overstated. Critics point to large racial and socioeconomic achievement gaps, misleading claims made by the school district's public relations department, uncritical press coverage, a precipitous decline in black educators, and funding that has been inequitably distributed to some of the city's most impoverished schools.

"I know that too many of the successes boasted of by schools and by educators like me are little more than polite interpretations of the same data scores," a D.C. charter teacher wrote recently. "Too much of what I see in my school today is exactly what I saw ten years ago." After a decade working in D.C. schools, she is calling it quits.

(Photo: AP/Pablo Martinez Monsivais)

Subsequent D.C. mayors (Vincent Gray, elected in 2011, and Muriel Bowser, elected in 2014) and schools chancellors (Kaya Henderson, appointed in late 2010, and Antwan Wilson, in late 2016) have largely continued to promote the school reforms launched by Fenty and Rhee. Though it's been more than two months since Wilson took over as D.C.'s new schools chancellor, it is unclear how he will steer the District of Columbia Public Schools (DCPS) between these competing narratives of success and hype. A better understanding of D.C. school reform, which was long heralded by the Obama administration as a national model, matters even more now that Donald Trump's administration aims to expand school choice policies across the country-likely beginning with the nation's capital.

THOUGH PEOPLE REMAIN starkly divided over education reform in D.C., the one thing both critics and supporters agree on is that the old way of evaluating teachers had to change. Removing bad teachers from the classroom had been too difficult. Mary Levy, a longtime independent budget analyst for the D.C. schools and a former DCPS parent, says it was well-known that some teachers shouldn't have been there, but they were hard to fire.

"There was peak enrollment in the late 1960s, and after that [the district] just abandoned their gatekeeping test and started hiring anyone who was breathing so long as they had a degree," Levy says. "My older daughter had one of those teachers, and she was unbelievably bad. So the district had an older workforce to whom no standards had been applied, and when enrollment started going down, and there were big layoffs in the 1980s, every elementary teacher with less than ten years in the system lost their jobs, and the older ones got to stay."

"The union contract in D.C. was awful," says Mark Simon, an Economic Policy Institute research associate and a former president of the Montgomery County (Maryland) teachers union. "It was an example of the kind of contract that existed in some school districts where the limitations placed on teachers' time and the specificity of what administrators had to do [for] an evaluation [to] hold weight was so rigid that more often than not, teachers could not be evaluated out of the school system." Simon added, "If a principal did not get the right documents filled out the right way on just the right line, then the whole thing was thrown out by an arbitrator."

An American Prospect review of a 2006 D.C. teacher evaluation handbook corroborates these observations. One byzantine rule stipulated that to terminate an ineffective teacher by the end of the school year, the administrator had to make a decision no later than the first week of January. If the process began with less than 90 days remaining in the school year, "the educator must be granted permission to return to the same site the next school year" as the process continued.

Simon opposes D.C.'s new system, IMPACT, which ranks teachers as highly effective, effective, developing, minimally effective, or ineffective, arguing that it de-professionalizes teachers. He contrasts IMPACT with the system he helped pioneer in the 1990s as union president for Montgomery County, D.C.'s suburban neighbor. Simon wrote in 2012, "The focus of teacher evaluation in Montgomery County is professional growth-the nurturing of good teaching, not the sorting and ranking of the teacher workforce." He added: "Although an evaluation system must be able to weed out people who never should have entered teaching, that objective only applies to a tiny percentage of the workforce and must not be the system's main purpose. Good teachers are not found through some magical recruitment pipeline. They are made, over time."

Simon says that in 2008 he approached Jason Kamras, the D.C. school official charged with developing a new teacher evaluation system, and suggested that the district craft a system similar to Montgomery County's. "[Kamras] ran it up the food chain, said other people had suggested the same thing, but that the response was that it takes too long, costs too much, we're not interested, we want to use a rubric to hire and fire," says Simon.

There had been some innovative teacher evaluation models at the time-Toledo, Ohio, was experimenting with peer review and others were exploring so-called professional learning communities. Even though Simon was critical of IMPACT, he agreed that policymakers had not been focusing much on improving teacher quality through feedback and evaluation.

"I think the reformers are right that people hadn't been paying enough attention to teacher evaluation, and in a lot of places the systems were pretty pro-forma," says Jesse Rothstein, a University of California, Berkeley public policy and economics professor. "But there were places that were doing it better, and that typically involved things like mentor[ing] teachers and careful classroom observations."

One reason D.C.'s education reforms attracted significant attention across the country was their timing: DCPS started using IMPACT to evaluate teachers during the 2009–2010 school year, just as the education reform organization The New Teacher Project (TNTP) released a report recommending that districts develop evaluation systems that rate teachers "based on their effectiveness in promoting student achievement."

IMPACT and TNTP's report heavily influenced the Obama administration's $4 billion Race to the Top program, which rewarded states that created new evaluation systems based on student test scores. (The administration also used No Child Left Behind waivers to incentivize similar policies.) According to the National Council on Teacher Quality, 35 states and Washington D.C. revamped their teacher evaluation processes to include student achievement as a "significant or the most significant factor" from 2009 to 2013.

By January 2010, 40 states had applied for the first round of competitive Race to the Top grants. The first two winners, Tennessee and Delaware, were awarded grants of $500 million and $100 million, respectively. Tennessee's proposal notably included a teacher evaluation system that looked just like D.C.'s.

(Source: National Council on Teacher Quality)

Since Tennessee won the first and biggest prize for a proposal modeled on IMPACT, D.C.'s program garnered even more notice. There was little research on its actual effectiveness, but many states nevertheless looked to D.C. as a leader to emulate. "All of these states were in the middle of a financial crisis, where their revenue declined dramatically, and to get this grant money they had to pretty quickly come up with new plans," says Matt Di Carlo, a senior research fellow at the Albert Shanker Institute. "I certainly think there is a tendency, an understandable tendency, to look around and see what other people are doing who were successful winning funds."

FOR YEARS, THE D.C. public schools have been known as factious battlegrounds for education reformers of all stripes; new plans and policies would be implemented every few years, only to have new leaders and competing agendas ushered in shortly afterward. The day before Rhee was appointed, The Washington Post traced this trajectory, noting: "The history of D.C. school reform is filled with fix-it plans hailed as silver bullets and would-be saviors who are celebrated before being banished. ... Isolated gains achieved under one reform theory were tossed aside, lost or forgotten in the next. Some reforms that did have an impact went awry, accelerating inequality, distrust and decline."

In 1989, a coalition of more than 60 business and community leaders published a report calling for sweeping changes to D.C. education, including closing and rehabilitating schools, lengthening the school day, and drafting new curriculum standards. "There have been countless studies, task forces, and five-year plans for the District's schools, but few come close to the size and scope of this effort," the Post reported at the time. The coalition spent six months and $500,000 on the effort, yet like those that came before it, their recommendations bore little fruit.

By 1996, the D.C. Financial Responsibility and Management Assistance Authority issued another report declaring the city's public schools to be in crisis, and called for urgent changes. By 2004, the Council of the Great City Schools, a national nonprofit, published its own report, noting that D.C. remained one of the lowest-performing urban school districts in the nation. They recommended a series of reforms that had been floated over the past five decades-new accountability systems for student achievement, more standardized curricula and instruction, and incentives to attract high-quality teachers to work in the most challenging schools.

Unlike other places, elected D.C. officials must compete with federal leaders for authority over the city's public schools. Congress can overturn laws passed by the D.C. City Council, and the District's delegate to the House cannot vote on legislation. The introduction of an elected school board in 1968 and the passage of the Home Rule Act in 1973 were attempts to increase local political representation, but the school board and council lacked independent taxing authority. It was no small sacrifice for residents when city leaders voted to dissolve the school board in 2007-dismantling one of the city's only elected bodies. But local officials felt drastic action was needed given DCPS's poor outcomes.

Rhee's tenure as chancellor was controversial, both locally and across the country. In addition to pushing forward a new teacher evaluation system, she fired hundreds of teachers, replaced principals, and closed schools. Her brash style of leadership frustrated even those who backed her policy ideas. Following Rhee's resignation in 2010 after Adrian Fenty lost the Democratic mayoral primary, the new schools chancellor, Kaya Henderson, continued to promote her predecessor's policies-albeit in a less polarizing way.

MEANWHILE, D.C.'S REFORMS continued to attract glowing praise. In 2013, The Washington Post editorial board concluded that there was "unassailable" evidence that the city's reforms, based on "high standards, rigorous evaluation of teachers, an investment in pre-kindergarten and school choice" worked. In 2014, Education Secretary Arne Duncan said D.C. was "by every measure the fastest-improving big city school district in the nation." New America called D.C.'s teacher evaluation "as rigorous and comprehensive as teacher evaluation systems get."

All the talk of success and failure led Steven Glazerman, a Mathematica Policy Research fellow, to coin a new phrase-"misNAEPery"-which describes how leaders and pundits wrongly attribute the rise and fall in National Assessment of Educational Progress scores to the success or failure of specific education policies. "D.C. [NAEP] scores [rose] faster than other cities-that part is basically true, but if you want to say it's because of school reform, that's a harder case to make," says Glazerman. Alan Ginsburg, a retired 40-year veteran of the U.S. Department of Education, published a report in 2011 that found that D.C. NAEP scores were already steadily improving before Michelle Rhee took over in 2007, and that "the rates of D.C. score gains under Rhee were no better than the rates achieved under [the prior two superintendents]."

Another thorny issue is demographics: Some critics charge that any documented learning gains can be attributed to the increase in white, affluent students who now enroll in DCPS. Yet when controlling for demographics, about two-thirds of the city's ten-year gains in math persist for fourth-grade and eighth-grade students. However, controlling for demographics does make the ten-year reading gains for eighth graders almost entirely disappear. In late February, Levy, the independent D.C. budget analyst, went before the city council to testify about the district's low academic performance. She noted that the lowest achieving groups are black males, at-risk students, and special education students. Achievement gaps between white and black or Hispanic students have narrowed somewhat since 2003, but white proficiency rates still run about 65 percentage points above black proficiency rates, and 53 to 61 percentage points above Hispanic rates. Socioeconomic gaps have widened.

"We have an ever-worsening achievement gap in this city, that has been spun into the D.C. miracle," says Elizabeth Davis, president of the Washington Teachers Union.

"Were a teacher to perform in this manner for their students, they'd have long since lost their jobs."

Critics have raised other concerns about the way D.C school reform has been cast as an example of "clear progress." School funding advocates have criticized DCPS for inequitably distributing financial resources to the neediest schools, and last September, the Washington City Paper published a cover story on Kaya Henderson's failure to deliver on her five-year strategic plan. A new report from the UCLA Civil Rights Project explores the city's heavily segregated schools.

But if there's one reform that supporters of D.C.'s school policies point to as evidence of success, it's IMPACT. In 2013, Thomas Dee, a professor of education at Stanford University and James Wyckoff, an education policy professor at the University of Virginia, published a working paper suggesting that D.C.'s teacher evaluation system induced teachers with low evaluation scores to voluntarily leave DCPS, and improved the performance of teachers who stayed. (This paper was published in a peer-reviewed journal in 2015.) In 2016, the researchers published another working paper (also later peer-reviewed) that found DCPS teacher turnover between 2011 and 2013 led to a net positive effect on student test scores-suggesting that turnover is not necessarily bad if low-performing teachers can be replaced with higher-quality ones.

These were encouraging results, but DCPS officials went on to exaggerate the findings. School administrators falsely said the research showed teachers and students improved because of IMPACT, and that IMPACT caused low-performing teachers to leave. The researchers had repeatedly emphasized that their work was not an evaluation of IMPACT, per se.

"DCPS has one of the best publicity operations I have ever seen," says Levy. "I think, unfortunately, they go beyond spin, and into some areas of half-truths."

DCPS was not alone in spinning the IMPACT studies. Supporters of VAM, a controversial statistical tool that uses student test scores to come up with estimates of teacher effectiveness, tried to frame the positive IMPACT studies as proof of VAM's merit. "People looked at the study and concluded it must be the VAM-based firing that did it, and that's not supported by the evidence," says Jesse Rothstein, who has raised concerns about using VAM in teacher evaluations.

(Photo: AP/Manuel Balce Ceneta)

The real issue with attributing the researchers' results to IMPACT is that there's no proof that other new teacher evaluation systems wouldn't have also worked. Dee and Wyckoff also caution that despite the positive results of their research, IMPACT might not work as a national model, given that D.C. is a particularly attractive location to live in (thus it has an unusually robust labor pool). The high salaries and bonuses DCPS teachers earn would likewise be difficult for many struggling school districts to adopt.

In an interview with The American Prospect, Dee adds that the leadership in D.C. was very strong and thoughtful, and that a system like IMPACT might not thrive under different political conditions. "When I present the IMPACT work, I say, yes, it does seem extremely promising but I worry it won't be a proof point," says Dee. "You had certain planets in alignment politically, and capable, entrepreneurial leadership."

Indeed, one factor that worked in DCPS's favor was that the 4,000-member Washington Teachers Union was significantly weakened, and unable to successfully fight against using test scores to evaluate teachers. The WTU has been under siege since the Rhee years, and teachers have been working under a contract that expired in 2012.

According to Davis, the union president, DCPS educators still strongly oppose the new evaluation system. "IMPACT does little to seed improvement in practice," Davis says. "Our professionals don't believe teaching every year should be a scene out of The Hunger Games, fighting for survival against what could best be considered arbitrary standards." She adds, "WTU teachers believe that educators should have an evaluation system that focuses on supporting and assisting those who work in the classroom and holds the whole system accountable, not one that obsesses on points, ratings, and consequences solely for teachers."

David Grosso, a city councilmember and the chairman of the education committee, tells me that while he respects the teachers union, when they "testify or complain or say things are awful, it's hard to believe" based on his personal conversations with educators. "Nine out of ten teachers I speak to are pretty happy and feel like they're making a difference," he says. "The fact of the matter is, if you're a teacher in the District of Columbia, you have the support that you need and when you are successful, you will get paid a lot of money and be treated with a lot of respect, and that's just a reality."

For what it's worth, schools located in the poorest areas of the city have the smallest percentage of teachers rated "highly effective" under IMPACT. Teacher turnover districtwide also remains very high. Levy, the budget analyst, finds almost half of all newly hired teachers, whether experienced or new to the profession, leave the classroom within two years; and 75 percent leave within five years. There is similar turnover among principals: Levy finds most schools have had two or three principals in the last five years.

ONE REASON IT'S become so easy for advocates to spin the city's school reforms is that despite DCPS's claims of being "data-driven," comprehensive, accessible data actually remains hard to come by. As a result, it is hard for researchers to get a sense of how specific policies are working, and for the public to hold school leaders accountable.

When D.C. passed its 2007 education reform law, one provision required the mayor to produce annual evaluations on new school reforms, such as academic achievement and personnel policies. The law also allowed the mayor to skip the annual reports and produce a five-year independent evaluation by September 2012. Fenty opted for the latter-but his two proposed evaluators, Frederick Hess of the conservative American Enterprise Institute, and Kenneth Wong of Brown University, had both supported the DC mayoral school-takeover plan. Then-councilmember Vincent Gray objected to Fenty's picks, arguing that they involved conflicts of interest.

Gray also objected to the mayor's desire to have the $750,000 evaluation paid for by an entity known as the D.C. Public Education Fund, a private organization launched and run by a former Fenty aide, which solicits private-sector donations to support education reform. Gray believed that the evaluation should be publicly funded. Yet three years later, when Gray himself ran for mayor, his tune on rigorous evaluations changed. "Adrian Fenty refused to carry out the evaluation, and when Gray ran against Fenty, he also lost interest," says Levy. "Gray's attitude changed a lot when he became mayor."

Levy thinks that incentives for oversight worsened after the switch to mayoral control. Before the change, the city council would sketch out the school district's finances, but the body could not control how those funds were actually spent. This dynamic frustrated councilmembers who were often blamed for the public schools' struggles, but had few tools to address the problems. This issue led the council to enact tougher oversight measures. "The public would come down and say, 'You need to give us more money,' and the city council wanted to justify not coming up with all of it," Levy explains.

But after the move to mayoral control, DCPS failures were no longer pinned on the city council. "Now the mayor comes up with a budget number for the school system and that's pretty much it," says Levy, who thinks the city council is not interested in rocking the boat. "They too have gotten all this good publicity," Levy says, in regards to the supposed successful turnaround of DCPS.

D.C. finally produced a publicly funded independent evaluation of its school reforms in 2015.

The National Research Council, an organization chartered by Congress, conducted the review and found some promising evidence of improvements, but the evaluators identified many persistent disparities, and noted a lack of comprehensive, accessible data. They said they were often unable to obtain important information for their research effort, and recommended the creation of a data warehouse for ongoing, independent studies.

After the NRC issued its report, a group of education advocates and public policy researchers gathered in 2016 to discuss creating an independent think tank to evaluate D.C. education policies. Inspired by the University of Chicago Consortium on School Research, which has access to a broad range of Chicago Public School data, the D.C. group envisioned their think tank serving a similar function as the Congressional Budget Office.

Mathematica's Glazerman agrees it has been difficult at times to obtain DCPS information to conduct research. "The researchers want to do research, they want access to data, and the people who control the data don't want to give it up, except under tightly controlled circumstances," he says. "Researchers need independence and access to data, and they shouldn't have to worry about whether the agency is going to look good-both in whether they undertake the study, and how they report results from their study."

He thinks the idea of a publicly funded research organization akin to the CBO is a good one, but that it could be a heavy lift to get off the ground. It would take real leadership, and right now, the mayor and the city council have few incentives to poke holes in the narrative that D.C. school reform has been a tremendous success.

"We met for about six months and put together a proposal," says Mark Simon, who was involved in the 2016 effort. "Initially we got good, positive encouragement from David Grosso, and he basically promised to put money in the budget, but when we got to the actual budget hearings we were iced out."

The Prospect asked Grosso why he withdrew his support for the independent research organization. "I hadn't heard that much about it, but I do support the idea for third-party analysis and review of what we're doing in DCPS, but I was not convinced that what they were offering at the time was the best approach," he said. "It seemed like it was a purely academic thing. There was a desire to do something similar to what was done in Chicago and, in the end, I decided I did not want to do that. I thought it would confuse governance in the city more than it would help."

THIS PAST FEBRUARY, DCPS's new schools chancellor, Antwan Wilson, took over. Prior to coming to D.C., he spent two years as the superintendent of the Oakland Unified School District in California and worked as a public school administrator in Denver. He also participated in a superintendent training academy funded by the Eli and Edythe Broad Foundation, which finances education reform efforts. "The candidate [Mayor Bowser] has selected appears by résumé and reputation to have the same kind of forward-thinking passion for excellence that has helped make D.C. schools the fastest-improving urban school district in the country," The Washington Post editorial board said in November.

Wilson declined the Prospect's request for an interview through DCPS press secretary Michelle Lerner. Lerner is a former communications manager for several reform-driven organizations, including the Fordham Institute and the advocacy group American Federation for Children, formerly chaired by Betsy DeVos, now the U.S. secretary of education.

Looking to the future, Councilmember Grosso says D.C. will need to invest more heavily in wrap-around services for poor students, including basic health care, housing, and resources for coping with trauma. He says that he's spoken with Antwon Wilson and that the new chancellor "absolutely understands" this.

The bipartisan political forces that shepherded D.C.'s education policies may shift in the coming years, as the election of Trump and the ascendance of the controversial DeVos threaten to fracture some of the Obama-era coalitions. New leadership, both in the district and the mayor's office, could also portend greater changes for D.C. public education.

Though Glazerman is skeptical that a publicly funded research agency committed to robust, independent evaluations will be created, it is possible that Wilson may be more open to the idea, since his outsider status might shield him from the fallout from any negative findings-at least at the outset. Mary Levy also thinks the independent think tank idea could resurface, citing the new influx of upper-middle-class families who send their children to D.C. public schools.

"They don't take 'no' for an answer," she says. "These are city parents behaving like persistent suburban parents. So in the future, this idea may grow."

CORRECTION: The original version of this article failed to properly identify Professors Dee and Wyckoff, and to clarify that their papers were later peer-reviewed. The article has been corrected.