Orwell Award Announcement SusanOhanian.Org Home

The Mismanagement of Reading First: Summary of Evidence, Part 2

Susan Notes: This report documents why Reading First is known as the Enron of education policy.

NOTE: This is Part 2 of a two part study
June 2006

Did the Department of Education Promote the Use of DIBELS?

The Reading First legislation called for tests to be used for screening, diagnosis, and progress monitoring, to ensure that students receive the correct level of instruction and remediation as necessary. The only constraint on these assessments in the law is that they be valid and reliable. Yet almost every state has been required by ED to use one measure, DIBELS, developed at the University of Oregon and published by Sopris West. Almost unheard-of before Reading First and rarely proposed for use in first drafts of early state Reading First proposals, DIBELS was the only assessment presented to states at the Secretary's Reading Leadership Academies, and it was then pressed on states by reviewers of the state Reading First applications. How did the USDE persuade states, in many cases, to abandon tests they already used in favor of a relatively new and untried assessment? The following points outline the process:

1. The U.S. Department of Education contracted with the National Institute for Literacy to review a short list of selected assessments, excluding some of those most widely in use, and subsequently promoted the use of DIBELS. Numerous valid, reliable assessments of early reading exist and are in use in many classrooms. Among these are the Developmental Reading Assessment (DRA), Woodcock-Johnson, and many others. Yet in a review of assessments for use under Reading First - one of the first products that the US Department of Education commissioned for Reading First –none of the above-named instruments were included. The review was led by Dr. Edward Kame'enui, then a faculty member at the University of Oregon, as well as Roland Good (University of Oregon), who developed DIBELS; and Deborah Simmons (then at the University of Oregon). Kame'enui, Simmons, and Good were three of the four extenal members of the design team for Voyager Passport, a remedial program that is built around DIBELS.
According to the published report, this initial review of assessments was to be an example review, with the intent that further reviews would consider other instruments. No subsequent reviews took place at a national level, despite requests from publishers of other tests to do so. Yet when new DIBELS subtests were introduced, however, they were adopted without review.

2. The review panel, through numerous comments and multiple reviews, frequently required states to designate a single assessment for statewide use for screening and progress monitoring. This requirement is above and beyond the requirements in the law.

3. Many states, in their initial applications for funding, wrote that districts should be allowed to select their own assessments. Some of these states further specified that these selections should be from the list of assessments reviewed. However, in each case, reviewers questioned the use of any instrument other than DIBELS for progress monitoring. The DIBELS, before Reading First, was in use in few schools, while for example, the DRA was in use in thousands of schools, and in some states, was used statewide (Kentucky is one example.) Yet reviewers repeatedly insisted that the DRA was not technically adequate. Across all applications, despite states' clear desire to maintain their assessment systems, reviewers criticized those applications that did not use DIBELS. According to published news reports, Reading First Director Chris Doherty himself told the Kentucky Reading First team that they had to use DIBELS in order to get funding. The Kentucky State superintendent complained to the Department of Education that the technical assistance consultant sent to help the state revise its proposal was a certified DIBELS trainer (Brownstein & Hicks, 2005a).

According to reviews of state applications, use of DIBELS in addition to other assessments was not sufficient. Schools and districts had to agree to stop using their former assessments – and change entirely to DIBELS. The State of Illinois, for example, had its application approved to use the ISEL. However, the US Department of Education informed Illinois that it had to switch to DIBELS in the middle of the school year, resulting in the state's having no baseline, or comparison, data that year. Illinois Reading First reports and correspondence refer to "abrupt policy directives from the funder" – the U.S. Department of Education. When New Mexico received two consecutive negative monitoring reports, it developed a set of "non-negotiables" in collaboration with the US Department of Education. DIBELS is a requirement on this list. All Reading First schools in New Mexico must attest that they will adhere to this list.
According to a recent report from the Center on Education Policy (CEP), states were "remarkably consistent" in the instruments they selected to assess student progress in reading. CEP's review determined that, at that time, 37 states required use of DIBELS as part or all of school district assessments, and that five additional states included it on a list of assessments from which districts could choose. Since the publication of that report, other states have reported switching to DIBELS.

4. DIBELS has been featured at both of the two national Reading First conferences, despite ED's stated "no vendors" policy. ED set up a special registration system to ensure that no vendors could attend these conferences, yet its sessions on DIBELS show a clear preference for this single commercial instrument.

Did the Department of Education Promote
Specific Experts to Provide Professional Development?

One of the most remarkable elements of Reading First is its promotion of a limited number of professional development models for schools across the country. In most initial state Reading First applications, states proposed to use reading experts from their local universities. These were almost invariably rejected. Regardless of the qualifications of their state experts, states were routinely cautioned that "degrees do not equal expertise in SBRR." Reviewer comments uniformly questioned the qualifications of anyone other than a set of "national experts." In practice, our reviews of state Reading First proposals, federal reviews of those proposals, correspondence from state Reading First directors, and invoices for payment, make it clear that states primarily used three professional development providers: Louisa Moats and her LETRS program, published by Sopris West; Edward Kame'enui and Deborah Simmons and their Institutes on Beginning Reading; and Sharon Vaughn and adaptations of her Texas Reading Academies.

LETRS, published by the for-profit Sopris West publishing company, was created by Louisa Moats, a close personal friend and co-author of Reid Lyon. Initial state proposals rarely mentioned LETRS at first, but most did so in their final, funded proposals.

It is not our intention to criticize Louisa Moats or her LETRs training program but rather to question the idea that she was so uniquely qualified that the federal Reading First administrators and reviewers were justified in rejecting other qualified trainers in her favor. Her vita shows a modest number of publications before 2001, all of which are in special education (not reading) and most of which focus on spelling. Yet Moats – a vendor from Sopris West, a for-profit publishing company--headed up the "professional development committee" that helped to plan the Secretary's Reading Leadership Academies, meetings where states learned what Reading First expected of them. Moats spoke at the academies as well. Her LETRS program was pressed on states in preference to other professional development programs and providers. For example, Michigan's Reading First director stated in a news article that she felt "forced" to use Moats and the LETRS program (Brownstein & Hicks, 2005b) – a program that has never been evaluated, and in fact was still being developed and revised during the first few years of Reading First.

The Institutes on Beginning Reading were created by Drs. Deborah Simmons and Edward Kame'enui of the University of Oregon for the State of Oregon during the Reading Excellence Act. Kame'enui and Simmons both participated in planning the Secretary's Reading Leadership Academies and spoke at those meetings. Since both were under contract to RMC Research Corporation to provide technical assistance to states on their applications, they were in an extremely advantageous position to recommend the Institutes. Further, the U.S. Department of Education routinely referred states to the Oregon web site, where these Institutes were advertised alongside the Oregon list and the approved assessment list.

Did the Department of Education Promote the Use of the Three-Tier Model of Instruction?

A key part of Reading First implementation has been the promotion by all three Reading First technical assistance centers of a "three-tier" model of instruction. This model is much more than a general suggestion to provide core instruction (Tier 1), supplemental instruction (Tier 2), and intervention (Tier 3), depending on students' needs. Instead, it prescribes specific grouping strategies, instructional strategies, and assessment strategies. It requires small group remediation and bans one-to-one tutoring. The three-tier model has never been evaluated in comparison to a control group. This makes Reading First a $1 billion-a-year pilot test of the three-tier model, which is pushing out well-established programs such as Success for All and Direct Instruction. SFA and DI have far better evidence of effectiveness, yet conflict with many details of the three-tier model.

1. From the very beginning of the Reading First program – for example, the planning of the Secretary's Reading Leadership Academies, and the initial contracts with RMC research for technical assistance to states in writing their proposals – the U.S. Department of Education selected three individuals from the field of special education to play central roles. All three of these individuals – Edward Kame'enui, Sharon Vaughn, and Joe Torgeson – are heavily identified with the "three-tier model." Kame'enui and Vaughn share a large federal grant to study the three-tier model, on which Torgeson is a major consultant. Vaughn's center at the University of Texas at Austin sells modules and technical assistance on the three-tier model, and the model is described in detail on the University of Oregon's Reading First Web site. At the Secretary's Reading Leadership Academies, this group introduced the requirement – not contained in the Reading First legislation – that states would need not only a "core" reading program, but also a "supplemental" and an "intervention" program. Note that Kame'enui, Simmons, Vaughn, and (at first) Torgeson were all highly paid consultants to Voyager, which has apparently dominated the "supplementary" and "intervention" niches created by the three-tier model.

2. The three-tier model has been presented many times to key Reading First groups, most importantly during meetings of the state Reading First directors, for which the agenda was entirely determined by the US Department of Education. The individuals named above participated in multiple contracts, including running the three technical assistance centers tasked with the daily operations of the Reading First program. In this role, they and their staff traveled the country, presenting the three-tier model not as "a" model, but as "the" model that must be used in order to drive reading achievement at schools and in districts. Similarly, the three-tier model was presented at both national Reading First conferences. No alternate models were presented.

3. Subsequent requests for technical assistance have focused on problems implementing features of the three-tier model. After such presentations took place, documents including formal requests for technical assistance show that states and districts believed this model was a requirement under Reading First. In Massachusetts, a district using Success for All was forced to negotiate with Reading First regional coaches to alter the program to better "fit" the three-tier model. In New Mexico, Success for All was dropped from use by Bernalillo schools upon determination by the New Mexico Reading First director that it did not "fit" the three-tier model. In Florida, regional evaluators use the three-tier model as the default – what they are trained to look for – when they visit classrooms.

4. Monitoring instruments used both by federal contractors and states penalize schools for not using the three-tier model. Early in Reading First's implementation, states appeared to concern themselves with adopting a core reading program. However, in subsequent years after instructional materials issues were settled, the major concern was on instruction – through the three-tier model. With the core reading program as the first tier, schools were then to adopt a supplemental program as a second tier, and an intervention program as a third tier.

Monitoring done by both the state and the federal government is supposed to be measured against the state's approved proposal; that is, monitoring is not intended to evaluate whether or not the state has done what the federal government wants them to do, but what the state plan has established will be done. A review of final approved state plans reveals that very few states ever discussed the three-tiered model in their plans (Louisiana, Texas, and Florida were the exceptions). However, the monitoring protocol appears to expect that all states will be using the three-tier model, whether or not it is specified in their application. Schools that felt their program was adequate with only a core reading program received negative findings on their monitoring reports. In Massachusetts, the three-tier model is a component of the state monitoring framework; unless a school is implementing all aspects of the model, it does not meet the established criteria.

The specification of this instructional model is a major component of a state's Reading First implementation. Yet we could not find evidence that any state that had not originally stated that it would use a three-tiered model ever submitted an amendment to its Reading First application requesting permission to use this model. In contrast, states that made relatively minor adjustments in their state plans – Michigan, for example, tweaked its professional development delivery system, and Massachusetts changed the ratio of coaches to schools – were required to submit these changes, in writing, as proposed amendments to their plans. New Mexico, as stated above, received two negative monitoring reports and as part of its action plan, developed in conjunction with the US Department of Education, it agreed to require all RF grantees to use the three-tier model.

5. The RMC Research Corporation bid for, and received, a large contract to provide technical assistance to the 50 states through the three centers named above. In their application, which was accepted by the US Department of Education, they specified that the three-tier model will be the instructional model that will be taught and used across all Reading First sites.

6. Success for All has repeatedly been told by national and state Reading First officials that it should alter its program to fit the three tier model. In a conversation with Reading First Director Chris Doherty, Dr. Robert Slavin discussed the negative implications of the three-tier model for Success for All. Essentially, the three-tier methods for student grouping, the percentage of teaching time (as opposed to individual work), and prescribed methods of teacher student interaction conflict with how SFA operates. Dr. Slavin was told by Mr. Doherty that SFA should change its program to accommodate the three-tier model. Of course, to change the program in those very essential characteristics would be to negate 19 years of research on the program's effectiveness. Mr. Doherty clearly believed that the three-tier model takes precedence over any other instructional model, no matter how well researched.

Has the Department of Education Fulfilled the Reading First Act's Emphasis on Scientifically-Based Reading Research?

The Reading First legislation strongly emphasizes the idea that programs used under Reading First must be "based on scientifically based reading research" (SBRR). The most rigorous definition of SBRR defines it as having evidence indicating that programs themselves have been rigorously evaluated in comparison to control groups:

"Reading First…provides assistance to states and districts in selecting or developing effective instructional materials, programs, learning systems, and strategies that have been proven to teach reading." U.S. Department of Education, 2002

Elsewhere, the legislation defines what kind of research a program must have to be "proven to teach reading:"

• Employs systematic, empirical methods that draw on observation or experiment;

• Involves rigorous data analyses that are adequate to test stated hypotheses;

• Relies on measurements or observational methods that provide valid data…;

• Has been accepted by a peer-reviewed journal or approved by a panel of independent experts through a comparably rigorous, objective, and scientific review."

According to P.L. 107-110, ss. 1201 (20 USC 6361), the purpose of Reading First is "To provide assistance to State educational agencies and local educational agencies in selecting or developing effective instructional materials (including classroom-based materials to assist teachers in implementing the essential components of reading instruction), programs, learning systems, and strategies to implement methods that have been proven to prevent or remediate reading failure within a State."(emphasis added). In 2002, when then-Assistant Secretary of Elementary and Secondary Education Susan B. Neuman invited states to apply for Reading First, she wrote, "Reading First is the academic cornerstone of the No Child Left Behind Act, which recognizes the importance of both improving student achievement and implementing programs and strategies scientifically proven to be effective." Indeed, the Department repeats this message on its web site: "Reading First is a prime example of the No Child Left Behind law's emphasis on programs and teaching methods that have been proven to work." The Department noted:

• The emphasis on scientific research prevents the use of unreliable and untested methods that can actually impede academic progress; makes teaching more effective, productive, and efficient; and can be better generalized and replicated in multiple schools.
"Just like every other aspect of No Child Left Behind, states and local communities maintain control."

• States and local schools have the flexibility to determine how reading programs are selected, as long as the selected program has been scientifically proven to work (emphasis added).

• There is no federally prescribed reading program.

• States are responsible for the quality of the local programs they fund, and for ensuring that these programs rely on scientifically based reading research. http://www.ed.gov/programs/readingfirst/nclb-reading-first.html.
However, apart from repeated rhetoric, it is clear that the US Department of Education has acted to minimize, and in many cases prevent the use of programs that meet these research standards in favor of those without any evidence of effectiveness.

Among the five top-selling basal textbooks (the "Michigan list"), there is only one text examined in a single study that met these standards. In contrast, there are dozens of rigorous experimental-control comparison studies published in peer-reviewed journals showing the effectiveness of Success for All, Direct Instruction, and a few others. With such a clear focus on rigorous research, how could Reading First have ended up virtually mandating programs lacking such evidence?

One answer is that in practice, Reading First administrators applied a much less rigorous definition of "based on scientifically based research." This definition requires only that programs incorporate elements that have been rigorously evaluated. In practice, this lesser standard has meant that programs must emphasize the five elements of effective reading instruction derived by the National Reading Panel (2000): phonemic awareness, phonics, vocabulary, comprehension, and fluency. Since virtually every reading program contains at least some focus on these elements, this sets an extremely low standard.

1. The U.S. Department of Education re-purposed a "Consumer's Guide," developed by the University of Oregon under a previous contract, to assist states in evaluating instructional programs. Although Ray Simon wrote in a letter to Senator Lugar that "states were under no obligation to use the Consumer's Guide as their review criteria, and many chose not to," there is significant evidence to conclude that states who wished to examine any reading program other than the five basals included by the State of Michigan were compelled to use this Consumer's Guide. The Consumer's Guide is specifically required in many reviews of state applications, and an e-mail from Sandi Jacobs, Chris Doherty's assistant, told Missouri officials to "say more about [the Consumer's Guide]. It is one of the main things the reviewers are looking for."
In its review of Reading First, the Center for Education Policy (CEP) found that:

"in their approved applications 42 states indicated they would use [the Consumer's Guide] at the state and/or local levels. Only 2 states specifically mentioned the use of other instruments, both of which were developed by the individual state. The remaining states did not mention a specific instrument…. In order to examine the influence of the peer review process on the use of these two items, CEP compared the original applications with the final versions in a sample of 10 states. In each case, 4 of the 10 states added the Consumer's Guide to their applications after initial review... In all, 8 of 10 are using the Consumer's Guide. It is not clear from this analysis exactly why states added the Consumer's Guide. Revisions could be due to the states' belief that, in retrospect, DIBELS and the Consumer's Guide are the best instruments to use or could be due to pressure from reviewers."

Use of the original form of the Consumer's Guide is itself not problematic. It had two sections. The first section asked the reviewer to determine if there was sufficient scientific evidence on the program that had determined it to be effective. The second section, to be used only if the first section did not find evidence on the program itself, asked reviewers to examine the extent to which the program addressed the five components of reading instruction identified by the National Reading Panel.

According to numerous state proposals, Drs. Simmons and Kame'enui and others from the University of Oregon provided extensive training to many states in use of a version of the Consumer's Guide that removed the first section, thereby removing any examination of the evidence on any instructional program. This follows the method that apparently was used by the University of Oregon when the Oregon review was carried out. Programs were rated only on section two, the extent to which programs addressed the five components of reading instruction. In fact, section one of the Consumer's Guide was used only in one case we were able to document – the coalition review comprised of Washington, Alabama, Montana, and Idaho. Simmons and Kame'enui officially revised the Consumer's Guide in late 2004 to eliminate section one entirely.

2. The Florida Center for Reading Research (FCRR) was also given the task of evaluating programs, initially for use in Florida and then as a resource for other states. FCRR reviewed not only the programs, but also the research on those programs. Originally, its web site contained two charts, one on the five components and one on the research evidence behind each program reviewed. The research evidence chart was removed in mid-2004 and has not re-appeared. Further, as noted earlier, FCRR's web site "approved" the six top basals without review and only reviewed research on other programs.

3. The US Department of Education's promotion of the Oregon list and the version of the Oregon Consumer's Guide that excluded evidence of effectiveness as a criterion sent a clear message to states that they did not need to, and should not, consider evidence of effectiveness when reviewing materials. If the US Department had instead promoted the Washington coalition review process, which examined evidence in a meaningful way, or the original (later suppressed) Florida Center for Reading Research review, it would have had the opposite effect on state actions. In fact, if the US Department of Education had not promoted any one of these reviews, one would expect that states would have adhered to the legislation and guidance which, as excerpted above, are clear about requirements for evidence of effectiveness.

The US Department of Education also ensured that evidence of effectiveness would not be used through its Reading First Expert Review Panel, which reviewed all state applications for funding. Across the applications, most states initially deferred to local control, stating that their districts would review instructional programs according to their correspondence with scientifically based reading research. However, this was not sufficient, according to both the reviewers and ED staff. They had to use Section 2 of the Consumer's Guide, as noted earlier.
The review panels further required that the districts' ratings of programs be reviewed by "national experts." The review panels did not sign off on any state application for funding until the state guaranteed that it would evaluate districts' choices in consultation with "national experts in scientifically based reading instruction." In some cases, those national experts are even named. With few exceptions, the named reviewers were closely connected to the University of Oregon and University of Texas individuals who are authors or consultants to commercial basal and supplementary texts (Kame'enui, Simmons, Good, Vaughn), or they were connected to Louisa Moats of Sopris West, another commercial publisher.

4. The State of Oklahoma attempted to
follow the legislation in writing its state application. Oklahoma submitted four iterations of its proposal. In the first two, Oklahoma wrote that it would require districts to use instructional programs that had at least three years of longitudinal research on their effectiveness, a criterion that is fully in keeping with the legislation. The state was urged by reviewers twice not to do this because it would limit districts to only a few programs. Oklahoma relented on the third application, and in the fourth application – which was funded – Oklahoma agreed to restrict schools to the traditional basals on its state adoption list.
Other states, however, specified lists of instructional programs that were as limited in numbers of qualifying programs as Oklahoma's would have been. California, for example, approved only 2 programs – the two textbooks that had already been approved through their state adoption process. There are no comments by Federal reviewers on the limitations of the California list.

Scientific review was not necessary if states chose the five top basals. Michigan specified in its response to a public information request by the Success for All Foundation that it had simply "place[d] phone calls to all major publishers requesting examination copies for review." Neither reviewers nor the US Department of Education objected to this procedure that explicitly ignored scientific evidence. In fact, the Florida state web site explicitly excused the six top basals from any need for review.
The U.S. Department of Education's promotion and virtual requirement of the three-tier model is itself a policy in opposition to the use of research on program effectiveness. The three-tier model, which has never been evaluated in even a single study in comparison to a control group, reinforces the idea that research is only relevant when the Reading First leadership wants it to be. There is no need for research on an instructional model's effectiveness – it is enough, according to the US Department of Education, that the model be "based on research," regardless of research on outcomes or effectiveness, and even this minimal requirement was waived if states promised to use any of the top five basal texts.

Have There Been Conflicts of Interest
Among Reading First Leaders and Consultants?

The story of the mismanagement of Reading First is one of extraordinary disregard for commonsense ethical standards regarding conflicts of interest. Reading First contractors are authors of major basal series that have benefited enormously from Reading First funding. Several are consultants to Voyager, which has aggressively exploited the supplemental text niche created by Reading First. In an early 2005 conversation with Reading First director Chris Doherty, we pointed out these conflicts. He was undisturbed, said the individuals had all been vetted by the ED, and pointed out that if you want "the best people in reading," they were bound to be involved with publishers. Yet most of the conflicted individuals were virtually unknown in the reading world before Reading First; most who had any reputation at all had it within special education, not reading. There are many reading experts without such conflicts. The disregard for ethical standards is illustrated by one small but particularly extraordinary fact. On the Oregon Reading First technical assistance center web site under the heading "curriculum," there was for several years a link to the Scott Foresman web site, including information on the Scott Foresman Early Reading Intervention program authored by former center directors Ed Kame'enui and Deborah Simmons. It is unconscionable, of course, for a federally funded technical assistance center to openly promote a product that pays royalties to its directors. This link was mentioned in a press report in August, 2005. Yet for many months thereafter, the link was still there. More importantly, many of the conflicts discussed here have been reported in the press in articles going back to April, 2005. Yet none had been acknowledged or addressed six months later. An October 11, 2005 letter from RF Director Chris Doherty to RMC Director Everett Barnes did ask him to ask contractors in the technical assistance centers with conflicts of interest to recuse themselves from conversations about their programs. Yet by this late date, all states and almost all schools had already received their funding and adopted their programs.

Based on voluminous correspondence among these individuals and between the U.S. Department of Education and state Reading First programs, Reid Lyon, Edward Kame'enui, Deborah Simmons, Sharon Vaughn, and Louisa Moats were in a tremendously influential position. Correspondence attests to the fact that these individuals approved providers of instructional materials or professional development used in many states. For example, the federally approved Alabama proposal required that every reviewer of local proposals had to be approved by "one of the chairpersons of the Secretary's Reading Leadership Academy": Kame'enui, Vaughn, Moats, and Carnine. Three of these four (all but Carnine) have well-documented conflicts of interest. North Carolina's and Washington's approved proposals contain identical language and an identical list.

1. Edward Kame'enui and Deborah Simmons. Edward Kame'enui and Deborah Simmons were until 2005 the directors of the Reading First technical assistance center at the University of Oregon. They have a particularly long list of conflicts of interest. Both are authors of the Scott Foresman Early Reading Intervention program widely adopted under Reading First. Kame'enui has now joined the U.S. Department of Education as Commissioner of Research on Special Education, and he filed a financial disclosure form on taking his new job. It showed that he was earning $100,000 to $150,000 per year from this textbook. In addition, Kame'enui and Simmons are authors of Reading Street, the 2007 Scott Foresman basal series. They were both members of the "design team" for Voyager, the popular commercial "tier 2" supplementary program under Reading First. Kame'enui only resigned his Voyager consultancy in 2005, when he accepted his new job. Both Kame'enui and Simmons have long careers involved with the Reading Mastery program. Despite these many involvements with commercial publishers, Kame'enui and Simmons led the Oregon review panel, and served on those of several other states. They are the authors of the Consumer's Guide virtually required in most states to evaluate Reading First reading programs. They served as consultants to a large number of states on selecting instructional programs and materials, and recommended current and former colleagues to serve in other influential positions in Reading First at the state level, such as technical advisors to states during the application process for Reading First.
The fact that Kame'enui and Simmons were Scott Foresman authors is particularly important in light of the evolution of the "Oregon List" of basal textbooks that states and districts across the country used to determine which programs would be implemented under Reading First funding. As noted earlier, the original "Oregon List" of 2002 listed Success for All fifth, ahead of Scott Foresman. Scott Foresman was felt by many in the Reading First world to have too little emphasis on phonics, and the Florida web site recommended using Scott Foresman only with a phonics supplement.

In 2004, the Oregon center, led by Kame'enui and Simmons, revised its list of recommended basals. The only change was to move Success for All from fifth to seventh, and to move Scott Foresman from seventh to fifth. This change now made it possible for states to require the "top five" or "top six" basals on the Oregon list and thereby include all of the five favored basals, including Scott Foresman. Recall that the publisher of Scott Foresman, Pearson, was already paying Kame'enui and Simmons $100,000 to $150,000 per year in royalties, and had made them authors of America's top-selling reading text, ensuring them personal royalties of many hundreds of thousands of dollars per year.
The low ethical standards of the Oregon group is also illustrated by the fact that until recently, the Oregon Reading First Technical Assistance Center had on its web site a link to the Scott Foresman web site, conveniently set up so that the Kame'enui and Simmons Early Reading Intervention could be "added to your shopping cart."

As noted earlier, Kame'enui has now taken a position at the U.S. Department of Education and Simmons has moved to Texas A & M University, but both continue to be very influential in Reading First and continue to profit from sales of programs promoted by Reading First.

2. Sharon Vaughn. Until June 2005, Sharon Vaughn was the director of the University of Texas Reading First technical assistance center. She was the chair of the design team for Voyager, a major beneficiary of Reading First policies, and remains involved with Randy Best, Voyager's founder. External employment forms from her university reveal that for several years she has spent 10 percent of her time as a Voyager consultant. She is also an author of Reading Street, the 2007 Scott Foresman basal series, and is a frequent consultant for Pearson, the publisher of Scott Foresman. Scott Foresman is another major beneficiary of Reading First policies. Vaughn is a major proponent of the three-tier model and has been particularly outspoken in promoting this model within Reading First. Like Kame'enui and Simmons, despite her apparent conflicts of interest she has played a central role in organizing national Reading First conferences, serving as a consultant to many states, and recommending others as consultants.

3. Roland Good. Roland Good is a professor at the University of Oregon and is the main developer of DIBELS. Although DIBELS is available free on the web, most schools purchase it from the for-profit Sopris West company, which acquired it in 2003. Voyager Passport is built around DIBELS, and the virtual requirement that RF schools use DIBELS has therefore been a major selling point for Voyager. He has also been a member of the Voyager design team. Good served on the Oregon Reading First review team and on the panel that reviewed early reading assessments (and approved DIBELS but not other widely used measures). He has served on several other state review teams and as a consultant to several states.

4. Louisa Moats. Louisa Moats is an employee of Sopris West, a for-profit publishing company that publishes her LETRS training program as well as DIBELS and programs often used as supplementary texts under Reading First, especially one called Read Well. Although Reading First has been careful to keep vendors out of all national Reading First events, Moats, a vendor herself, has been central to planning and execution of Reading First trainings. Michigan listed Moats as the only person capable of providing the specific expertise required by the U.S. Department of Education, and offered her a sole-source contract. State officials typically preferred trainers from their own states but were usually forced to choose either LETRS, or training from Kame'enui and Simmons or the Texas Reading Academy.

5. Reid Lyon. Reid Lyon was a primary architect of Reading First. He was a division chief at the National Institute of Child Health and Human Development (NICHD). Through his involvement with then-Governor George Bush in the Texas Reading Initiative, he developed a personal relationship with President Bush, and was on loan to the White House for a period of time when Reading First was being rolled out. This relationship gave him enormous informal influence on Reading First, and many of the leaders of Reading First were personal friends and co-authors of his (especially Louisa Moats and Sharon Vaughn). Lyon helped plan the national Reading First annual meetings, has consulted with several states on Reading First, and has made speeches all over the country on the program.

Unlike the others discussed in this section, Lyon does not receive royalties from textbook publishers. The only form of conflict of interest we know about relates to his departure from government in 2005 to join Randy Best, the entrepreneur who created Voyager and then sold it for $380 million in 2005. (A Dallas newspaper estimated the value of Voyager in 2000, before Reading First, at $5 million.) Lyon has been identified in press reports as having praised Voyager long before joining Best, and when Reading First funding for New York City was at risk because they insisted on using a reading approach weak in phonics, Lyon was reported to have negotiated a deal in which they could receive their funding if they also adopted Voyager Passport districtwide as a supplementary text.

6. Rod Paige. Rod Paige, Secretary of Education during the establishment of Reading First, also joined Voyager founder Randy Best's company.

It is important to point out that the Department of Education had to be aware of at least some of these conflicts, which clearly violate the Department's own ethics policies for contractors. If, as seems apparent, the Department did know about these conflicts and went forward anyway, it must bear much of the responsibility for the ethical problems that appear throughout administration of Reading First.


The documents we have assembled on the Reading First program provides overwhelming evidence that the leaders and major contractors to Reading First had a clear idea of what they wanted the program to support, and they have used every means at their disposal to manipulate states, districts, and schools into adhering to this plan. The main elements of the plan are as follows:

Core Basal Programs: Macmillan, Scott Foresman, Harcourt Trophies, Houghton-Mifflin, Open Court, Reading Mastery

Assessment: DIBELS

Professional Development:
Louisa Moats (Sopris West)
Kame'enui-Simmons (Oregon)
Texas Reading Academy

Instructional Model: Three-Tier Model

Deviations from the above program elements do occur, but states, districts, and schools generally have had to fight for them and continue to this day to receive pressure to drop them. In contrast, states, districts, and schools that propose to use any of the above elements are not criticized for doing so, and federal funds continue to be used to develop and support extra training and materials designed by the Reading First technical assistance centers to support only these products.

Despite the clear emphasis in the legislation, research has played little or no effective role in establishing or administering the elements of Reading First. Reviewers and monitors do criticize schools for using programs "not consistent with scientifically-based reading research (SBRR)," but the practical definition of consistency with SBRR is whatever the Reading First leaders want it to be. Programs with strong and widely acknowledged evidence of effectiveness have been routinely shunted aside in favor of traditional basal texts that lack such evidence.

The levels of conflict of interest among the Reading First contractors are extraordinary. We do not have documentary evidence showing that particular consultants actively promoted materials from which they derive income, but it is at least a remarkable coincidence that the basal and supplementary textbooks, training programs, and assessments that pay substantial royalties or consulting fees to key RF contractors have received the bulk of Reading First funding, and that the two highest-placed government leaders of Reading First, Reid Lyon and Rod Paige, have joined a company that profited massively from Reading First.

In Reading First, Congress intended to provide substantial resources to see that research-proven reading programs would go to millions of children in low-achieving, high-poverty schools. Instead, they ended up providing a windfall to the same commercial textbook publishers that have always used marketing, not research, to promote their wares.

The cynical manipulation of this well-intentioned program continues to cheat millions of vulnerable children now, and by making a mockery of the concept of evidence-based reform, it cheats millions of children in the future who could have had access to proven programs that would teach them to read.

Adams, G.L., & Engelmann, S. (1996). Research on Direct Instruction: 25 years beyond DISTAR. Seattle, WA: Educational Achievement Systems.

Borman, G.D., Hewes, G.M., Overman, L.T., & Brown, S. (2003) Comprehensive school reform and achievement: A meta-analysis. Review of Educational Research, 73 (2), 125-230.

Borman, G.D., Slavin, R.E., Cheung, A., Chamberlain, A., Madden, N., & Chambers, B. (2005a). Success for All: First year results from the National Randomized Field Trial. Educational Evaluation and Policy Analysis, 27 (1), 1-22.

Borman, G., Slavin, R. E., Cheung, A., Chamberlain, A., Madden, N. A., & Chambers, B. (2005b). The national randomized field trial of Success for All: Second-year outcomes. American Educational Research Journal 42 (4), 673-696.

Borman, G.D., Slavin, R.E., Cheung, A., Chamberlain, A., Madden, N.A. & Chambers, B. (2006). Final reading outcomes of the national randomized field trial of Success for All. Manuscript submitted for publication.

Brownstein, A., & Hicks, T. (2005 a, August 29). IG targets conflicts of interest: Limits on local control. Title I Monitor.

Brownstein, A., & Hicks, T. (2005b, August 29). When research goes to market, is it a good thing for education? Title I Monitor.

Comprehensive School Reform Quality Center (2005). CSRQ Center report on elementary school comprehensive school reform models. Washington, DC: American Institutes for Research.

Foorman, B.R., Francis, D.J., Fletcher, J.M., Schatschneider, C., & Mehta, P. (1998). The role of instruction in learning to read: Preventing reading failure in at-risk children. Journal of Educational Psychology, 90 (1), 37-55.

Herman, R. (1999). An educator's guide to schoolwide reform. Arlington, VA: Educational Research Service.

Manzo, K. (2005 a, June 22). Complaint filed against reading initiative. Education Week.

Manzo, K. (2005 b, September 8). States pressed to refashion Reading First grant designs. Education Week.

Manzo, K. (2005 c, October 12). GAO to probe federal plan for reading. Education Week.

National Reading Panel (2000). Teaching children to read: An evidence-based assessment of the scientific research literature on reading and its implications for reading instruction. Rockville, MD: National Institute of Child Health and Human Development.

Salvato, N. (January 18, 2006). Effective reading programs share common characteristics: An interview with Reid Lyon. Education News.

Slavin, R. E., Madden, N. A., Cheung, A., Borman, G.D., Chamberlain, A. & Chambers, B. (2006). A three-year randomized evaluation of Success for All: Final reading outcomes. Baltimore, MD: Center for Data-Driven Reform in Education, Johns Hopkins University.

Toppo, G. (2005 a, August 8). Federally funded Reading First called into question. USA Today.

Toppo, G. (2005 b, October 10). Reading program raises questions for lawmakers. USA Today.

Traub, J. (1999). Better by design? A consumer's guide to schoolwide reform. Washington, DC: Thomas Fordham Foundation.

U.S. Department of Education (2002). Guidance on the Comprehensive School Reform Program. Washington, DC: Author.

— Success for All
Revision, June 2006


This site contains copyrighted material the use of which has not always been specifically authorized by the copyright owner. We are making such material available in our efforts to advance understanding of education issues vital to a democracy. We believe this constitutes a 'fair use' of any such copyrighted material as provided for in section 107 of the US Copyright Law. In accordance with Title 17 U.S.C. Section 107, the material on this site is distributed without profit to those who have expressed a prior interest in receiving the included information for research and educational purposes. For more information click here. If you wish to use copyrighted material from this site for purposes of your own that go beyond 'fair use', you must obtain permission from the copyright owner.