The Reading First Controversy
Promise and perils of federal leadership
“Reading First is the most effective federal program in history.” So reads the opening line of a report that Alabama superintendent of education Joseph Morton sent to his congressional delegation last June, in which he recounts how the program has raised reading achievement for poor students in his charge. Morton’s view is shared by leaders in many other states, where thousands of Reading First elementary schools have reported unprecedented progress closing the “literacy gap” among the poor.
And yet a string of ironies plagues Reading First, threatening its future at the height of its promise. The program is the only component of No Child Left Behind to be rated “effective” by the White House Office of Management and Budget, yet the administration has done little to protect it. Secretary of Education Margaret Spellings calls it “the most effective and successful reading initiative in the nation’s history,” yet she removed its two leaders from their jobs. It is the one part of the No Child Left Behind law educators say they like, yet it has been targeted for deep cuts in funding (see Figure 1). Its prescriptive approach would seem easy fodder for Republicans to attack, yet Democrats in Congress have led the assault against it.
A glance at education headlines suggests that Reading First has succumbed to the fallout from “scandal” (see sidebar). The true story, however, is that the program is a victim of its own high standards. It’s a federal program that appears to be working to change the way states and schools teach reading—and change never comes without conflict.
A Questionable Controversy
Complaints from three vendors who felt unfairly shut out of the program led to an investigation and a series of reports by the Department of Education’s Office of the Inspector General citing supposed lapses by Reading First staff and potential conflicts of interest among contractors and panelists reviewing programs.
The program vendors who raised concerns were Robert Slavin, author of the Success for All reading curriculum; Cindy Cupp, author of the Dr. Cupp Readers; and Jady Johnson, executive director of the Reading Recovery Council of North America.
Slavin’s is a spurious case. Early in Reading First’s implementation, he alleged that two curricula of proven efficacy, his own Success for All, as well as a rival, Direct Instruction (DI), were being unfairly shut out of Reading First. More specifically, he charged that, under federal duress, states were pressuring districts to avoid his program or to create guidelines that made its adoption less likely. Yet the inspector general (IG) not only failed to substantiate Slavin’s charges about Success for All, but actually built its “case” around the fact that several peer review panelists—as well as the program’s director, Chris Doherty—had ties to Direct Instruction programs.
Yet even this charge is dubious; it would be impossible to construct any panel of reading experts that didn’t include individuals with professional ties to the existing reading programs, and the IG never alleged actual financial conflicts of interest. Furthermore, DI commands only a modest share of the Reading First market. Nor did the federal review panel have direct oversight over the selection of individual programs, since few states discussed specific core programs in their state applications, and none identified Direct Instruction. The decision about which “research-based” curricula to use was left to the states and districts.
Cupp alleged that the state of Georgia, with federal encouragement, shut her product out of Reading First by prohibiting its use until she got an independent appraisal showing that it was research-based. The IG’s investigation of Georgia exonerated the state and the feds.
As for Reading Recovery, which is a one-on-one intervention program for first graders, its efficacy, cost effectiveness, and research base are all disputed. (The What Works Clearinghouse found some evidence of effectiveness in a recent review; other studies have found its approach to be less efficient than modifications to the program that use more phonics and that deliver instruction to small groups.) What’s most relevant, though, is that it’s a “pull-out” program using a tutoring approach, while Reading First explicitly requires the use of core, classroom-based reading programs. Its use has been discouraged by Reading First directors in most of the states cited in this article.
Doherty was forced to resign after he was “caught” steering states away from programs he felt did not meet the requirements of the law. (A series of vivid e-mails in which he denigrated things he believed were flagrantly against the law did not help his cause.) His deputy, Sandi Jacobs, was reassigned. Congress held hearings and decried the Department of Education’s management of the program. And, most recently, it slashed the program’s budget by more than 60 percent.
Yet it is far from clear that Doherty did anything wrong. The dispute over his actions rather reflects a deeper disagreement over the proper role of the federal government in education, an ambivalence well captured by the Reading First law itself, which embodies a contradiction. Not only does the law require that districts use curricula that rely on “scientifically-based reading research,” a phrase, defined at length that appears 25 times in the statute; it further instructs the Department of Education to assist and hold states accountable in meeting this rigorous requirement. But, as if denying or ignoring the import of what such assistance necessarily entails, it leaves in place older language that prohibits the department from endorsing programs or dictating local decisions about curricula. The IG and some in Congress have interpreted this prohibition as trumping the mandate to guide the states and hold them accountable, but no legal guidance has been issued establishing that this is so.
A New Breed
The goal of Reading First is for all children to read at grade level by the end of 3rd grade. The law instructs the U.S. Department of Education to make sure that states and districts use funds for curricula and practices that are grounded in “scientifically-based reading research,” as defined by the National Reading Panel report commissioned by Congress and published in 2000.
The report identified five interlocking elements essential to effective reading instruction. Reading programs funded by Reading First must include explicit, systematic instruction in these five elements: 1) the ability to hear and manipulate the individual sounds, known as phonemes, that make up words—referred to as “phonemic awareness” (the word “cat” has three phonemes); 2) phonics, the relationship of phonemes to the letters that represent them in written language; 3) fluency, the ability to read text accurately and quickly; 4) vocabulary development; and 5) comprehension strategies.
In addition, as a condition of funding, Reading First schools must administer timely classroom assessments and must adjust instruction as needed for every child based on the results.
The emphasis on early intervention to prevent failure in the nation’s youngest and most at-risk students marks a major break from the unsuccessful remediation models of the past three decades, which focused on helping older students after they had already fallen behind. Indeed, Reading First’s commonsense demand—do what has been shown to work—amounts to a sea change for teachers, principals, curriculum coordinators, publishers, trainers, state education agencies, colleges of education, professional associations, teacher accreditation agencies—everybody in the field of reading education.
The initiative’s detailed prescription also represents a sea change for federal education programs. Beyond prescribing a particular approach to reading instruction, it also demands scrutiny of all funded activities, every step of the way. The program is a hybrid: it gives formula grants to states, but to receive their share of funds (fixed amounts calculated by a formula tied to the states’ levels of need) states had to submit applications specifying in detail how they would set up competitive grant programs for their districts aimed at helping low-performing, high-poverty schools improve reading instruction in grades K–3. A panel of experts reviewed the state applications and rejected the initial proposals of virtually all states, typically because they did not meet the program’s rigorous criteria. (States were allowed to submit their applications again and again, until finally all 50 state applications met all criteria and were accepted. Several states had to submit their applications more than five times before the panel finally approved their plans.)
But the vigilance of the federal government didn’t stop there. It created and funded (at $1.8 million per year for each) three university-based technical assistance centers to help build the capacity of state staff to serve their districts, and set aside nearly $1 million per year to monitor the states and help keep them on track. As a result, if a state or a funded district fell off the wagon and started using reading programs not based on scientific research, the program’s leaders in Washington knew about it and tried to intervene. That sort of oversight was unprecedented and, most surprisingly, welcomed, at least by the state directors of Reading First who had seen many federal programs come and go, and who began to see this new initiative as something very different and very useful.
As of fall 2007, more than 6,000 elementary schools in 1,700 districts had received Reading First grants, about 10 percent of the public elementary school market. Making sure these districts and schools remained faithful to scientifically based reading research fell primarily to state departments of education (SEAs), agencies with little or no experience leading instructional change. Yet the crafters of the Reading First program viewed the states as key leverage points. Indeed, the most important (and uncertain) premise of Reading First was that it could catalyze and support meaningful change in the SEAs—could help them build agile expert systems that gave high-quality support to schools and districts—and thereby improve reading achievement among the poor, not just in isolated schools and districts as in the past but across entire states. Toward that end, the regulations permitted states to retain 20 percent of their grants (instead of the maximum 5 percent in other programs) for administration.
To find out if this gambit has paid off, I conducted about 200 interviews (each lasting an hour or more) from October 2005 through October 2007, with participants at all levels of Reading First, including directors in 17 states and the federal Bureau of Indian Education (I interviewed most of these more than once); directors and staff at the program’s three regional technical assistance centers; external evaluators of state programs; senior executives at publishing companies whose core reading programs are in many Reading First districts; trainers that the state directors cited as particularly effective; and district personnel whose schools had shown significant gains in reading achievement. The 17 states are Alabama, Arizona, Arkansas, California, Colorado, Florida, Idaho, Louisiana, Mississippi, Montana, Nebraska, New Mexico, New York, North Carolina, Tennessee, Washington, and West Virginia. I chose them on the basis of recommendations from trainers, colleagues, and U.S. Department of Education staff who felt they were committed to the program and afforded diverse viewpoints on its implementation. I used these criteria and the strength of the data to select five case studies.
And, indeed, it appears that the strategy is working, and not just in terms of promising student results (see sidebar). The most enduring achievement of Reading First may be that it has nurtured a group of state leaders who have developed deep expertise in the science of reading instruction and have been able to get steadily better at helping the districts teach more children how to read. In states where Reading First is working, districts look not to their long-standing networks of consultants and colleges for expertise, but to their state administrators. This is a bureaucratic revolution.
Preliminary Results Are Promising
Several studies of Reading First are under way, including some randomized field trials and a few using quasi-experimental methods. Until those results are in, we can only make very broad comparisons between Reading First schools and other schools in each state. This early evidence is pointing in a positive direction.
For example, in Alabama’s 94 Reading First schools, the percentage of all 4th graders deemed to be proficient (stanine 5 and above) on the Stanford Achievement Test rose 12.7 points, from 40.1 percent in 2003 to 52.8 percent in 2007, more than twice as fast as the gain for students at other schools (which rose 5.5 points, from 64.2 percent to 69.7 percent in the same period). The percent proficient jumped 13 points for minority students in Reading First schools and 6.7 points for minorities at other schools. Reading First schools were 82 percent minority (mainly black) in 2007, more than twice the percentage in other schools (36 percent).
In Washington state, the percentage of 4th-grade students meeting or exceeding the grade-level standard on the Washington Assessment of Student Learning in the first group of 51 schools increased 23.7 points (from 39.7 percent in 2003 to 63.4 percent in 2007), while the gain statewide during that period was just 9.7 points (from 66.7 to 76.4 percent). The state’s major at-risk subgroups—blacks, Hispanics, and Native Americans—all posted gains exceeding 20 points.
In Arizona’s first group of 72 schools to win grants (comprising 1,800 educators and 26,000 students), the percentage of students meeting or exceeding the standard on Arizona’s state AIMS (Arizona’s Instrument to Measure Standard) reading test rose from 45 percent in spring of 2004 to 59 percent in spring of 2007, nearly triple the gain for the state over the same period (62 to 67 percent). The percentage of 3rd graders meeting oral reading fluency benchmarks on the widely used Dynamic Indicators of Basic Early Literacy Skills (DIBELS) increased from 28 percent in fall of 2003 to 52 percent in spring of 2006, more than triple the increase for students at a comparison group of non–Reading First schools over the same period (31 to 38 percent). The percentage of English language learners meeting 3rd-grade DIBELS oral reading fluency benchmarks has more than doubled, from 19 percent in fall 2003 to 41 percent in spring 2007.
“Alabama is light years ahead of everyone else in closing the achievement gap,” Sandi Jacobs, Reading First’s former assistant director in Washington, said to me in October of 2005, two years before the state posted the biggest two-year increase in 4th-grade reading scores ever recorded on the National Assessment of Educational Progress.
Until she retired in January, Alabama’s Katherine Mitchell was the highest-ranking state administrator in the country whose sole job was to improve reading achievement. No state has more zealously followed the tenets of Reading First or taken greater advantage of the program’s flexibilities. Alabama, which launched its own reading initiative in 1998, has been allowed to amend its Reading First plan several times and redirect funds to meet a succession of emerging needs identified by the data and Mitchell’s coaches in the field.
“We don’t ever think a school will know how to do something unless we show them how to do it,” Mitchell says. “What varied was how fast the schools could learn. But we had to teach them all just about everything.”
Like some other states, Alabama requires schools to choose reading curricula from a short list of programs that have been certified by the state as based on scientific reading research. Unlike other states, Alabama pays program vendors to give extra training on their products. It is the only state to fund not just reading coaches but also principal coaches, who train principals to be better instructional leaders and who drive accountability to the district level by ensuring that schools get support from superintendents and central-office staff.
Known as a whole-language state whose districts protected their independence, Arizona began to change rapidly when Jaime Molera was appointed state superintendent of instruction in May 2001 and Reading First was signed into law as part of No Child Left Behind eight months later. Molera, his successor Tom Horne, and their Reading First directors, Marie Mancuso and her successor, Kathryn Hrabluk, used the law to leverage big changes in state reading policy that should endure no matter what happens in Washington.
“Before Reading First we had several fragmented statutes that were a reflection of the historic reading wars,” Mancuso says. “The law really gave our state the foundation we needed not only to revise them into one coherent piece of legislation, but to change state board policy, revise our early reading standards to reflect scientifically based reading research and, most important, provide guidance to our districts and schools about how to provide a system centered around a prevention model for early reading achievement. The Reading First award gave us the resources not just in funds but in technical assistance to really implement our plan the way we had written it.”
Like Alabama, Arizona committed extra funds, including $1 million for training, to extend the program’s approach to all its elementary schools, and has worked with districts to redeploy federal Title I and Title II funds to pay for reading coaches, training, and research-based curricula.
Arkansas is among the few states that do not require Reading First schools to adopt a commercially published core reading curriculum. Instead, Reading First director Connie Choate has used her state department of education staff to develop protocols, write materials, and train educators to provide the functional equivalent and closely track their results.
Schools are required to teach phonics and phonemic awareness systematically and explicitly and to use the DIBELS (Dynamic Indicators of Basic Early Literacy Skills) to monitor students at least monthly and adjust instruction as needed. Each school-based Reading First coach receives at least two weeks’ training by state staff over the summer, including on the job while teaching summer school (an idea inspired by Alabama), and 30 days during the school year.
Choate’s staff did a task analysis of student learning expectations in the state’s English Language Arts standards (recently strengthened) and wrote 320 mini-lessons for 2nd and 3rd grade. Special education teachers were given a six-tiered matrix of intervention procedures for students who need extra instruction and practice. The statewide literacy-training program was strengthened to include instruction on fluency-building exercises, interventions, DIBELS, and how to teach vocabulary.
“As an agency we’re moving from monitoring to technical assistance—that’s a huge transformation for us,” Choate says.
Bureau of Indian Education (BIE)
Spread across 23 states, BIE schools pose enormous challenges. Under the federal Title system, 100 percent of their students are considered economically disadvantaged and at risk for educational failure. The percentage identified as learning-disabled is well above the national average. A high percentage of both students and teachers speak English as their second language. The entity responsible for them—like the schools themselves—suffers from high staff turnover and does not function well. Compared to most SEAs, BIE has minimal resources. Its Reading First director, Lynann Barbero, has a staff of two to cover 25 Reading First schools scattered across 11 states.
But Barbero also has two advantages. First, while state laws often prohibit SEAs from encroaching on district autonomy, because Barbero’s office in effect is the district, she has been able to intervene directly in schools where the core curriculum was not working. Second, because of her background in special education, she knows which curricula work best for students who are far behind their peers or who need extra practice.
“Most core reading programs are not designed for kids way below grade level. It’s crazy to have kids sitting in an ineffective program for 90 minutes before giving them what they really need,” Barbero says. “We’ve totally stopped that. Kids identified with intensive needs are put into a replacement core program right away.”
The new director of BIE, Kevin Skenandore, has decided to use $2.3 million from a congressional education-enhancement appropriation to expand Reading First to all Bureau schools serving K–3 students.
“There is no question for the Bureau that this program is producing results,” says Barbero. “Our data are as clear as day. We can now say, ‘If you follow the Reading First guidelines, you’ll get results. If you don’t, you won’t.’”
When Lexie Domaradzki took the job as the state of Washington’s Reading First administrator in 2003—she has since been promoted to lead administrator for K–12 Reading—she was determined to shatter the myth that poor children couldn’t learn to read as well as anybody else. Four years later, she takes it as almost a personal failure that the 4th graders in her Reading First schools, more than 3,700 students, have cut the achievement gap with the state by more than half but haven’t eliminated it.
Like the program’s national director, Chris Doherty, Domaradzki had worked prior to Reading First with large groups of schools (as an area manager covering five states for the Success for All Foundation). Like Jacobs, she had been an elementary-school teacher. She also has a master’s degree in organizational development, which she says taught her about how systems can be changed.
“Like Sandi and Chris, we set high expectations and have worked to meet them with a mix of accountability and support,” she says. “Organizational culture and results both have improved dramatically.” So much so that Washington’s statewide K–12 reading plan is now based on Reading First.
Reading First is controversial because it is prescriptive. Simply put, it requires states and districts to follow the dictates of reason and science when spending taxpayers’ money on education and holds them accountable if they fail to do so. Navigating among conservatives who oppose intrusive government, liberals who oppose President Bush, educators who guard their independence, and commercial interests who guard their market share, the law’s framers and program leadership sought to leverage the power of the federal government to attack a complex pedagogical problem that the federal government was never designed to solve: illiteracy caused by faulty teaching.
Ironically, the things federal officials were faulted for are the same things that state Reading First directors say they have tried to do themselves and that (they say) have made the program so successful. Doherty and Jacobs selected the best available experts to review grant proposals, evaluate curricula and assessments, and provide technical assistance; they steered educators away from approaches of doubtful efficacy and insisted that they choose among things more likely to work; they closely monitored implementation and pressed states to intervene when districts failed to execute their plans for raising student achievement and to cut off funds when noncompliance was not corrected. Each of these steps was essential to make the program work, the state directors say. Each required strong leadership. And each contributed to the backlash that, to the directors’ dismay, prompted Secretary Spellings (for reasons and with consequences still unclear) to throw the two staffers responsible for her “most effective” success story under the bus.
Says Rick Nelson, a retired teacher and former president of the American Federation of Teachers local in Fairfax County, Virginia: “State standards and No Child Left Behind gave teachers a knife and said: serve those peas faster, longer! Reading First gives 10 percent of teachers a serving spoon.” He might have added: but if you work in Washington, best not get caught telling anyone how to hold the spoon.
Shepard Barbash is former bureau chief in Mexico City for the Houston Chronicle and a freelance writer whose work has appeared in The New York Times, Wall Street Journal, Washington Post, Smithsonian Magazine, and other publications.