Nonpartisan Education Review / Articles
The Costs of Federal Intervention in Local Education:
The Effectiveness of America’s Choice in Arkansas
Professor Emerita, University of Arkansas
Oklahoma Department of Education
ABSTRACT and EXECUTIVE SUMMARY
Since passage of the No Child Left Behind act in 2001, all states have been under federal pressure to identify “failing schools.” The US Department of Education has strongly encouraged “turnaround” as an option for schools to use in order to improve student achievement. The study reported here reviewed the research literature on school “turnaround,” with a particular focus on the results in Arkansas of the use of the same “turnaround” partner (America’s Choice) for over five years for many of the state’s “failing” elementary, middle, and high schools.
There was little evidence to support either this partner or this policy option. The lack of evidence raises a basic question about the usefulness and cost of the federal government’s intervention in local education policy. Nor is it clear how a new program titled Excellence for All now offered by the developers of America’s Choice will differ. It is described as follows: “Excellence for All provides a clear, practical strategy for every student to graduate from high school prepared to succeed in college or in a program of career and technical education that will lead to a successful career.” We are also told: “Excellence for All is aligned with the Common Core State Standards, enabling participating high schools to not just lay the foundation for the Common Core but to get a head start on implementation.”
According to a draft of a yet-to-be-voted-on re-authorization of the Elementary and Secondary Education Act (ESEA/NCLB), state departments of education are to submit “college-ready” standards for approval by the USDE if they want Title I funds. However, state legislators, local school board members, parents, and a state’s own higher education academic experts are excluded from approving the standards their state department of education submits, perhaps to ensure that only the “right” standards are submitted. If the language in the current re-authorization draft is approved by Congress, Arkansas may find Excellence for All one of the few intervention programs recommended for its low-income, low-achieving schools despite its costs and no evidence that this new program can do better than America’s Choice did.
At this point, it seems reasonable to suggest that there should be no federal or state requirement for "turnaround" partners or the "turnaround" model, whether or not the programs they promote address Common Core’s standards. We clearly do not need the federal government pushing states and local districts to pay for consultants and services to solve the problems in low-achieving schools for which they have had no solutions.
Since passage of the No Child Left Behind (NCLB) act in 2001, states across the country have been under intense federal pressure to identify “failing schools”—schools receiving Title I funds that do not make Adequate Yearly Progress (AYP) as a whole and in which all identified demographic subgroups do not make AYP--and to make changes. That was the impetus for states that requested a waiver from the U.S. Department of Education (USDE) in 2013. 2014 was the magic year by which all subgroups were to become proficient, and no state saw this miracle happen.
According to NCLB (a re-authorization of the Elementary and Secondary Education Act first authorized in 1965), if a school failed to meet AYP for two consecutive years, the school was deemed in need of improvement (Year 1) and needed to offer public school choice to its students. If a school failed to meet AYP for three consecutive years, the school was labeled in need of improvement (Year 2) and needed to offer public school choice and supplemental services, including tutoring. If a school failed to meet AYP for four consecutive years, the school was labeled in need of improvement (Year 3), and needed to take corrective action. If a school failed to meet AYP for five consecutive years, the school was labeled in need of improvement (Year 4), and needed to plan its restructuring. If a school failed to meet AYP for six consecutive years, the school was labeled in need of improvement (Year 5), and needed to implement a restructuring plan. A school could exit Program Improvement when it met AYP for two out of three years. However, the G.W. Bush administration did not spell out all the kinds of restructuring plans from which failing schools could choose for implementation.
The Obama administration, in an August 2009 Press Release (USDE 2009), announced it would provide $3.5 billion in Title I School Improvement Grants to turn around the nation's lowest performing schools and identified four basic types of interventions school districts must choose from. Indeed, the criteria in its Race to the Top (RttT) competitive grants in 2010 awarded 50 points to identifying and doing something with failing schools, and the 12 states that won the RttT competition needed to overhaul the bottom five percent of their schools to secure their full share of the $4.3 billion in prize money.
The four basic types of interventions from which states are required to choose for implementation are as follows:
Turnaround Model: This would include among other actions, replacing the principal and at least 50 percent of the school's staff, adopting a new governance structure, and implementing a new or revised instructional program.
Restart Model: School districts would close failing schools and reopen them under the management of a charter school operator, a charter management organization, or an educational management organization selected through a rigorous review process. A restart school would be required to admit, within the grades it serves, any former student who wishes to attend.
School Closure: The district would close a failing school and enroll the students who attended that school in other, high-achieving schools in the district.
Transformational Model: Districts would address four specific areas: 1) developing teacher and school leader effectiveness, which includes replacing the principal who led the school prior to commencement of the transformational model, 2) implementing comprehensive instructional reform strategies, 3) extending learning and teacher planning time and creating community-oriented schools, and 4) providing operating flexibility and sustained support.
The Obama administration appeared to believe that high-poverty low-performing schools can reach high academic standards if they have strong expectations, strong leadership and staff, improved instructional programs, extended class time, engaged families, and changed governance (e.g., as charter schools—separately funded private or public schools whose students were typically selected by law by lottery). However, it presented only selected "case studies" to support these features, as was pointed out in a review of the research base for its proposed policies for a re-authorization of NCLB (Ravitch and Mathis 2010). It did not offer evidence to support the effectiveness of any specific "transformational" or turnaround "model." Nor did it show a body of research evidence on the effectiveness of the turnaround "partners" that schools had been using since 2006 to help them address chronically low-performing schools. In fact, questions had been raised about companies certified by states as school turnaround partners because some of them had no experience improving the fortunes of low-performing schools—or any school, for that matter, as reported in Education Week (Aarons 2010). As Education Week also reported, in August 2010, Congressman George Miller of California expressed plans to investigate them. Since the November 2010 election, there has been no further movement on this topic.
Purpose of Study and Sources of Information
The claim has been made by organizations selling turnaround services to low-performing schools that time is needed to show solid improvement. This is a reasonable claim. It is built into the useful definition that one such organization, Mass Insight, offered in a 2010 report (Mass Insight 2010). It defined "school turnaround" as a "dramatic and comprehensive intervention in low-performing schools that produces significant gains in achievement within two years and readies the school for the longer process of transformation into a high-performance organization." Organizations selling turnaround services and programs also claim that the extent of implementation by a low-performing school is another factor to take into account. This is also a reasonable claim, although it requires an enormous amount of classroom observation to verify what administrators or teachers may report on a survey.
To find out if there was evidence to support these claims, the co-authors of this report reviewed the research and related commentary on a turnaround partner heavily used in Arkansas from 2006 to 2010. Limiting our focus to the same state and to same turnaround partner for over five years would be more informative, we decided, than combining results from different turnaround models and states. We focused on America's Choice (AC), a for-profit private organization begun in 1998 as a program of the non-profit National Center on Education and the Economy (NCEE) and then sold in 2010 to Pearson, a leading world-wide education publisher. AC sells consulting services, school designs, instructional systems, curricular programs, and professional development. In 2006 then Arkansas Commissioner of Education Kenneth James (in 2009 he became executive vice president of AC and in 2010 its president) signed a multi-million dollar contract with AC, paying it to serve as a turnaround partner for 30 to 40 low-performing elementary, middle, and high schools in the state for four years. We first reviewed several sources of information on turnaround models themselves, then the research on AC as a turnaround partner, and, finally, a report by the Arkansas Bureau of Legislative Research.
Systematic Review of Information on Turnaround Models
We began with what is known about turnaround models based on the publications of Mass Insight, an organization in Massachusetts providing consulting services nationally to the schools, including strategies for turnaround. Its first major publication on the topic, The Turnaround Challenge (Mass Insight 2007), revealed no successful examples of any turnaround model at that time. Its most recent report on the topic, School Turnaround Models: Emerging Turnaround Strategies and Results (Mass Insight 2010), "focuses on emerging examples of effective school turnaround from the field…" This June 2010 report identified six urban school districts across the country (Baltimore, Chicago, New York City, New Orleans, Los Angeles, and Charlotte-Mecklenburg) then undertaking long-term turnaround initiatives to improve outcomes in some or all of its low-performing schools, as well as four "lead partnership" models involving a school district and several local education agencies or charter management organizations. As the report indicated, they were all "emerging examples" and their turnaround initiatives were all "necessarily" long-term. None is as yet a successful example of either a district-based turnaround model or a partnership-led turnaround model.
A second source of information on turnaround addressed the lack of a research base to support turnaround as one of the four intervention approaches spelled out by the Obama administration for chronically low-performing schools. The October 2010 report by Diane Ravitch and William Mathis titled Review of College- and Career-ready Students, released by the National Center for Education Policy in Boulder, Colorado, reviewed a May 2010 "research summary" produced by the USDE (USDE 2010a; 2010b; 2010c) supporting its proposals for the re-authorization of NCLB. One of six summaries prepared by the USDE for this purpose, the summary reviewed by Ravitch and Mathis, addressed three key areas: Common Core standards, rewarding progress and success, and turning around low-performing schools. They described the research base the USDE provides as "superficial and inadequate" and noted that "for two major issues, national standards and school turnarounds, the research cited does not support the document’s conclusions." They went on to remark that "fewer than 15% of the report’s references rely on independent, peer-reviewed research" and that the document "advances rhetorical ends and political goals rather than providing a sound research base for the proposed policies." They concluded: "Overall, the document is of little or no value for those who seek evidence of the soundness of the Obama administration’s proposed legislation."
A third source of information on the effectiveness of either a turnaround model or school closure as an intervention type came from a December 2010 report released by the Thomas B. Fordham Institute (Stuit 2010). Its author tracked 2025 low-performing charter and district schools in ten states from 2003/2004 through 2008/2009 to determine how many were turned around, shut down, or remained low-performing. (It should be noted that he looked only at elementary or middle schools and used a formula to identify them that differed from the one used by the USDE to identify low-performing schools for school improvement grants.) He found 72 percent of the original low-performing charters remaining in operation and still low-performing five years later. So were 80 percent of the district schools. Stuit concluded that most low-performing public charter and district schools are resistant to significant change.
Stuit further concluded that: "Rather than pushing dubious turnaround efforts, charter authorizers and education policy makers alike should ramp up their efforts to close bad schools, particularly in cases where higher-performing schools are nearby." This advice was based on his finding that "the charter sector has done a slightly better job of eliminating low-performing schools…In all ten states, low-performing charter schools were likelier to close than were low-performing district schools…And in both sectors, the majority of schools that closed were lower-performing than their neighboring schools; thus, students leaving closed schools had better academic options nearby."
A fourth source of information should not be ignored—the conclusions of a September 2008 report by the Center on Education Policy titled A Call to Restructure Restructuring: Lessons from the No Child Left Behind Act in Five States (Scott 2008). This study examined 42 schools in 19 districts judged by state departments of education for taking restructuring seriously. However, the researchers found no one federal restructuring option more effective than another in helping the 42 schools make AYP. In addition, none of the staff interviewed in schools that had exited restructuring could point to a single strategy that they believed was the key to improving student achievement. The researchers concluded that any re-authorization of NCLB should refrain from codifying in law these four types of intervention as requirements for low-performing schools (p.27).
Nevertheless, the authors of a March 2011 Issue Brief from the National Governors Association's Center for Best Practices, titled "State Strategies for Fixing Failing Schools and Districts," clearly suggested that we know the causes of what it calls 'school failure" and, therefore, what more federal funds should aim for (NGA 2011). They claimed, without any body of credible evidence to support its claims, that: "The underlying causes of school failure are similar, regardless of whether the schools are located in urban, rural, or suburban communities. First, the schools are characterized by weak leadership, starting with a principal who fails to keep the school focused on teaching and learning. Second, teachers’ skill levels tend to be inadequate, and teachers are not afforded the coaching necessary to help them improve as professionals. Third, failing schools do not pay enough attention to using high-quality curricula and instructional materials and often lack the capacity to analyze student test results to guide improvement. Finally, compared with a typical school, a failing school often has twice the number of high-poverty students and many more students who enter the school below grade level."
Despite the certitude of the Center for Best Practices, most of these "causes" are not established by any systematic research as causes. For example, how do we know that the teachers' skills are inadequate? Simply because the students are below grade level? How do we know that these teachers (as opposed to teachers in non-failing schools) "do not pay enough attention to using high-quality curricula and instructional materials…"? And what is "enough attention?" Without any evidence, the authors of this Brief nevertheless recommended that we invest enormous funds to try to "Build state capacity to support the turnaround of failing schools and districts" by (1) Engaging external partners to manage school and district turnarounds; (2) Setting ambitious but realistic goals for school improvement that incorporate multiple measures; (3) Developing a human capital strategy to improve the quality of leadership and teaching; and (4) Increasing state authority to intervene in failing schools and districts, if other approaches prove insufficient. The authors of this Issue Brief honestly admitted that whatever strategies have been used work no more than "10% to 20% of the time," that "no strategy currently works all the time or even on average."
Yet another study released in 2011, referring to some of the studies described above, recommended "five steps districts can take to improve their chances of success” (Baroody 2011). But it modestly described them as steps that will "increase the probability" that efforts to transform low-achieving schools will achieve lasting improvement. Titled Turning Around the Nation's Lowest-Performing Schools, and released by the Center for American Progress, its five steps are a bit more specific than those recommended by the Center for Best Practices (e.g., "quantify what each school gets and how it is used"). But its authors, too, suggested that it is not clear to anyone how to turn around a "failing" school.
We do not know from a review of the research and related literature whether an external partner can accomplish anything with a low-performing school. But, we were able to find out what one long-term external turnaround partner accomplished in one state—Arkansas—with a large number of schools.
Research on America's Choice as a Turnaround Partner
We begin with a brief description of the goal and key feature of AC's comprehensive school improvement model, the "literacy workshops" it mandates as part of this model, and the deliverables in the Arkansas Department of Education's four-year contract with America's Choice (AC), from 2006-2010 (a contract later amended to extend to 2011). We then review the evidence for AC's effectiveness as a turnaround partner in research studies that precede the present study.
AC’s School Design is a K-12 comprehensive school model designed by the NCEE. Its stated goal is to “make sure that all but the most severely handicapped students reach an internationally benchmarked standard of achievement in English language arts and mathematics by the time that they graduate.” To implement the program over a three-year period, each school “must assign personnel as coaches.”
The following description of AC’s Readers Workshop appears in a NCEE-sponsored evaluation in 2002 of the way AC “literacy workshops” are to be implemented.
Readers Workshop is structured to begin with a whole-class meeting in which the class might do a shared reading and have a mini-lesson in a 15-20 minute time period. The mini-lesson can cover phonics-based skills, decoding word analysis, comprehension skills, or procedures. This mini-lesson is usually followed by a period of independent/guided reading and/or reading conference period in which a number of activities like partner reading or book talks occur for about 45 minutes. In independent reading, students focus on reading appropriately leveled texts for enjoyment and understanding. Partner reading allows students to work with slightly more difficult texts, practice reading aloud, and model “accountable talk” and “think-aloud” strategies. Reading aloud provides an opportunity for the teacher or other proficient reader to introduce authors or topics and model reading for the whole class. Shared reading allows the teacher to work with smaller groups of readers on reading strategies. Readers workshop may end with a book talk in which students share reactions to books read independently or to a book read aloud to the group.
AC’s original four-year contract with the Arkansas Department of Education beginning in June 2006 indicated that AC would:
(1) deliver the AC comprehensive school improvement model to 46 low-performing public schools in Arkansas: 36 are in restructuring and will receive the intensive version of the model; 10 other schools will receive the basic model.
(2) provide professional development and
training to the target schools in the critical areas of
(3) provide on-site technical assistance and coaching from AC Cluster Leaders to the targeted schools. Cluster leaders will facilitate school Leadership Team meetings, hold Teacher Meetings, model and demonstrate strategies in classrooms, coach the school Coaches and support full implementation of the model.
In March 2009, a review team came from the USDE Office of Student Achievement and School Accountability Programs to monitor the state's compliance with federal regulations on the use of school improvement funds. In the cover letter for its report dated May 2009, the Director of Student Achievement and School Accountability Programs pointed out that eligible schools that did not want the services of AC were given no school improvement funds in the previous two years because all the funds set aside for this purpose had been awarded to AC (Stevenson 2009). The relevant comments appear on pp. 11-12 of Stevenson’s letter:
“The ADE did not ensure that it properly allocated its school improvement funds to its LEAs. Following SASA’s March 2006 onsite monitoring review, the ADE chose a vendor, America’s Choice, to enhance its statewide system of support. America’s Choice provides services and support to the Title I schools that meet the ADE’s criteria of having been identified for improvement for three or more years. America’s Choice is funded through a contract using 95 percent of the four percent of school improvement funds set aside under section 1003(a) of the ESEA. In 2007-2008, several LEAs with schools that met the ADE’s criteria chose not to receive services and support from the America’s Choice vendor. The ADE, however, awarded the entire reservation for school improvement to America’s Choice, leaving no funds for assistance to the schools choosing not to participate in America’s Choice even though they met the eligibility criteria.
“Additionally, for the school year 2008-2009, the ADE did not provide ED staff with documentation that
1) the ADE had requested approval from some or all of its LEAs to use their share of the 1003(a) school improvement funds for America’s Choice services and support; and/or
2) the ADE had allocated any of the school improvement funds to LEAs meeting the selection criteria to participate in America’s Choice. Again, America’s Choice was awarded the full amount of the school improvement funds.”
The relevance of this report for the present study lay in the fact that by 2009 there was negative evidence on AC's effectiveness as a turnaround partner for Arkansas's low-performing high schools, according to a WestEd report dated 2008 titled America’s Choice in Arkansas: Implementation and Achievement after One Year (McCrary, Ziobrowski, and Bojorquez 2008). More to the point, the funds available by law for schools not wanting AC's services might have been used by these schools in productive ways.
The report found evidence of achievement gains in the elementary and middle schools AC worked with but unsatisfactory outcomes in the high schools it worked with. It implied that implementation was the problem. Analyzing data from the 34 schools participating in the first year of the project, the authors found, among other things, that:
“Students in elementary and middle schools that adopted the America’s Choice school design (ASCD) had one-year achievement gains more than twice as high as their counterparts in comparable schools. School-level effect sizes were 0.04 in literacy and 0.09 in mathematics. …While these effect sizes are very small according to commonly accepted practices…, it must be emphasized that the ACSD takes three to five years to fully implement and these are achievement gains after one year.”
Review of the Diagnostic Assessment Tool data supplied by America’s Choice reveals several interesting patterns in implementation of the ACSD. Implementation of the ACSD is highest at the elementary school level and lowest in high schools. Elementary schools seem to be embracing the ACSD in their schools, providing the necessary training and resources to teachers to successfully implement the design, and involving parents and the community in the implementation to ensure success.
“…after one year America’s Choice schools at the elementary level were well on their way to implementing the ACSD with fidelity and were seeing achievement gains in literacy and mathematics. Middle schools were also far along in implementing some components of the ACSD, and also realized achievement gains. High schools had not made as much progress in implementation, and student scores in the America’s Choice schools were lower than those in comparable schools.”
Interestingly, a comprehensive examination of the research on secondary reading programs by Robert Slavin and his associates published in 2008 found “no qualifying studies” on AC's Ramp-Up Literacy program (Slavin, et al 2008). The criteria in this study that eliminated existing studies on Ramp-Up Literacy for further examination were (1) use of randomized or matched control groups, (2) a study lasting at least 12 weeks, and/or (3) valid achievement measures that were independent of experimental treatments (in other words, the studies on Ramp-Up Literacy did not meet one or more of these criteria). We do not know whether Slavin and his associates had examined some or all of the studies CPRE had included in its evaluative studies before determining that there were no "qualifying studies" on AC's secondary reading program.
The most positive evidence for the effectiveness of AC's programs but only at the elementary level and again only at the level of statistical (not practical) significance came from a study released by the Consortium for Policy Research in Education (CPRE) in 2009 comparing three models of comprehensive school reform programs (Accelerated Schools Project, America's Choice, and Success for All) that had been implemented from 1999-2004 in 115 elementary schools across the country (CPRE 2009). The researchers found "statistically significant differences in patterns of achievement growth for students in AC schools in the upper [elementary] grades. … From the beginning of third grade to the end of fifth grade …students in AC schools, on average, scored an additional nine to 12 points on the reading comprehension outcome, depending on the model adjustments."
The researchers accounted for these changes by noting AC's emphasis on "a significant amount of guidance and press for instructional standardization as part of its instructional improvement strategy… not by emphasizing scripted instructional routines, but rather by encouraging development of strong instructional leadership in schools…" However, they also noted that this emphasis on standardization and leadership "worked against the formation of strong professional communities and also decreased the press for innovation and autonomy in AC schools."
Some information on AC at the high school level came directly from teachers in high schools under contract with AC in Arkansas. A 2009 survey of over 400 high school English teachers in grades 9, 10, and 11 in Arkansas by researchers at the University of Arkansas was followed up by eight focus group meetings in all four of the state's Congressional districts (Stotsky, Goering, and Jolliffe 2009). A number of teachers at these meetings called attention to the influence of AC on both the content of the high school English curriculum and teachers’ instructional strategies. Many of their comments were highly negative (see pp. 39-41 in the report). What may account for the positive effects researchers have found for AC at the elementary school level—instructional standardization—may be exactly what aroused a great deal of hostility to it at the high school level in Arkansas, as well as a lack of results.
The positive evidence from the CPRE studies was balanced by the results of a study on seven Arkansas middle schools under contract with America's Choice. Undertaken for a dissertation by a staff member of the Arkansas Department of Education, the study examined the impact of AC's School Design on middle schools under contract for three complete years, from 2006 to 2008 (Cox 2010). Devonda Cox first compared student scores in grades 6, 7, and 8 in seven low-performing middle schools with the state average on state assessments. She then compared their scores with those of students in comparable grades in schools that had been officially "waivered" from implementing America's Choice School Design in 2006. Finally, she compared their scores with those of students in schools with similar demographics, using the Diagnostic Assessment Tool (DAT) to determine the extent to which the schools under contract had implemented the AC design.
She found, first, that schools in the AC School Design had a lower mean score in 2008 than in 2006, while students across Arkansas had a higher mean score in 2008 than in 2006. Moreover, the gap between the America's Choice schools and those across the state of Arkansas appeared to increase for literacy at all three grade levels (p. 101). Second, she found that schools receiving a waiver in 2006 had a higher mean score at each higher grade level each year, while the schools under contract with AC showed a decrease in the mean score at each higher grade level each year. Third, she found that as implementation increased in schools under contract with AC according to DAT indicators, scores decreased in grades 6 and 8. She concluded that "change in average score for these schools mostly depends on factors other than implementation of the specified America's Choice indicators found on the DAT" (p. 94).
Lack of significant results above the elementary school level is not just an Arkansas phenomenon. AC was a turnaround partner with the public schools in Holyoke, Massachusetts from 2006-2008 under an initial $2 million contract with the Massachusetts Department of Elementary and Secondary Education (DESE). AC implemented its Ramp-Up Math and Ramp-Up Literacy curricula in six schools, with two of these under-performing schools receiving an additional 30 days of support from AC coaches. As the DESE put it in its Race to the Top application in January 2010, “results were mixed.” As noted by a local reporter: "Few improvements were seen when the schools tried the America's Choice math program so that was ended” (DeForge 2010). After further decline during the years Holyoke was implementing Common Core’s standards following state adoption of them in 2010, the Holyoke, Massachusetts, Public Schools were put under receivership by the state’s board of education in April 2015.
Schools using AC in Arkansas in 2006-2011 and Results
The results of our review of the research literature and related commentary on AC were serendipitously confirmed by a research report titled Review of School Improvement Consulting Expenditures and Results, issued on February 7, 2012, by Arkansas's Bureau of Legislative Research. At the request of a state legislator, the Bureau had examined the results of the work of school improvement providers in the 300 Arkansas schools receiving School Improvement funds from 2006-2011. The review included only the organizations that provided a "comprehensive set of turnaround services, including some combination of professional development, coaching, diagnostic analysis of schools' weaknesses, special curriculum, and regular visits by consultants." The Bureau compared the 192 schools receiving at least $50,000 worth of services in a year with 190 randomly selected schools receiving no services (from a total pool of 925 schools). This left five major vendors, including America's Choice, plus a sixth category called "multiple providers" for schools receiving at least $50,000 worth of services in a year from more than one vendor.
The Bureau examined school improvement providers' results by looking at changes in their schools' overall percentage of students scoring proficient or advanced on state tests in literacy and mathematics.
The Bureau found that vendor schools had significantly higher gains in both math and literacy in the last five years than schools with no services. But some vendors were more effective than others. When the results of the five major providers plus "multiple providers" were individually compared with schools receiving no services for average annual literacy gains, only three of the providers were superior to schools that received no services and the average gains in America's Choice's schools were the lowest in the six comparisons. The gains of only one of the other vendors were statistically superior in mathematics to the comparison group, although none of the differences between providers in math was statistically significant.
The costs of these services in Arkansas were not commensurate with the results. About half of the state and district funds used for School Improvement in 2006-2010 (about $25,000,000) went to America's Choice—over $6,000,000 a year for work with 30-37 schools. Calculating per/school costs, the Bureau found that the state paid AC:
$190,357 per school (31 schools) in 2006-07
$208,084 per school (30 schools) in 2007-08
$167,655 per school (37 schools) in 2008-09
$178,358 per school (35 schools) in 2009-10
The Bureau estimated a total expenditure of about $70,000,000 for the five-year period, with $20,000,000 alone for 2010-2011. In 2010-2011, after the contract ended, school districts were allowed to choose their turnaround partner. AC was paid in new federal funds for its services to four schools in the Little Rock area: $310,000 for Cloverdale Middle School, which saw a decline in both literacy and mathematics; $394,000 for Hall High School, which saw a slight decline in mathematics and a slight increase in literacy; $310,000 for Rose City Middle School, which saw a decline in both literacy and mathematics; and $409,000 for J.A. Fair High School, which saw an increase in both subjects, math especially. In other words, only one of the four schools AC served in 2010-2011 clearly appeared to benefit from its services.
The Bureau found that only 39 schools in school improvement status for more than three years were removed from that status since 2006. Over 50 percent (21) of these schools had no school improvement provider. Six of the other 18 (or 17 percent of the 39) had used only the services of AC. AC's costly services, together with the uneven or inadequate achievement made by the schools it served in Arkansas, probably helped to reduce the number of voluntary contracts with it in 2010-2011, which amounted to about $1,400,000, all in the Little Rock area.
As the Bureau noted, with access to new and greatly enhanced funding, and with intense pressure to pull up test scores, districts have found themselves facing "an aggressive school improvement consulting industry vying for their business." The Bureau further noted that many school improvement companies started out providing professional development and gradually began to offer more comprehensive services as more federal school improvement funding has become available. It would seem that, at least in the case of the companies selling their services in Arkansas, their eagerness to use the money stream made available chiefly but not only by the federal government apparently outstripped the effectiveness of the services they sold the state, if indeed there was an early record of effectiveness for any of them. Caveat emptor would clearly appear to be the case with America's Choice, which alone received about half of the $50,000,000 the state and school districts spent on turnaround services from 2006-2010 because of a contract signed with it in 2006 by the then commissioner of education in Arkansas.
What do these results suggest might be done in order to help students in low-performing schools? There are no clear clues, for Arkansas or for the rest of the country. Perhaps academically adequate students need to be separated from those with serious behavioral and academic problems in the elementary grades so that the low-performing students can be dealt with intensively, whether or not they are in special classes. Surely, some students in these low-performing schools do not warrant the intense efforts at remediation that they may receive if the whole school is declared “failing.” Perhaps young adolescents need a range of program choices for their high school curriculum (as offered in modern vocational/technical high schools for grades 9-12) to motivate stronger study habits.
We badly need some research on what was done in the schools served by America’s Choice in Arkansas. We don't know what the interventions were. Were teachers dismissed? New teachers hired? Principals changed? New materials bought? What kinds of consultants brought in? How evaluated? We have no independent reports on what teachers did in their classrooms as a result of the consulting, coaching, curricular, and professional development services they received from America's Choice. Nor do we have accounts from their administrators. At least we should know what was or was not done in the schools partnered with America's Choice, and what will be done differently by a newly developed successor program titled “Competency-Based Education Framework and Readiness Assessment for Post-Secondary Education.” This new program is described as addressing the post-secondary problems faced by low-achieving students enrolled in post-secondary institutions.
Arkansas also needs to know what will be done differently in another new program now offered by NCEE titled Excellence for All. It is described as follows: “Excellence for All provides a clear, practical strategy for every student to graduate from high school prepared to succeed in college or in a program of career and technical education that will lead to a successful career.” Moreover, we are told: “Excellence for All is aligned with the Common Core State Standards, enabling participating high schools to not just lay the foundation for the Common Core but to get a head start on implementation.”
A revision of ESEA/NCLB is now underway. According to a recent draft of a yet-to-be-voted-on re-authorization, state departments of education are to submit “college-ready” standards for approval by the USDE if they want Title I funds. However, state legislators, local school board members, parents, and a state’s own higher education academic experts are excluded from approving the standards their state department of education submits. (See comments on the problems in ESEA’s re-authorization drafts in Stotsky 2015.) If this draft is approved by Congress, Arkansas may well find Excellence for All one of the few “intervention” programs recommended for its low-income, low-achieving schools despite no evidence that this new program can do better than America’s Choice did.
What do the results of our review of the research and related literature on the school turnaround model in the context of the report by the Arkansas Bureau of Legislative Research suggest with respect to federal policy? At this point, it seems reasonable to suggest that there should be no requirement for "turnaround" partners or for the "turnaround" model. We clearly do not need the federal government pushing states and local districts to pay for consultants and services to solve the problems in low-achieving schools for which they have had no solutions.
To make matters worse, readers can infer from a newly released study by the Institute for Education Sciences (IES), reported on by the Washington Post on May 5, 2015, that the federal government is trying to blame the states for the failure of "turnaround" schools to improve student achievement, not its own policy or the inadequacies of the turnaround partners that the states and schools were often coerced into using. This implies that the federal government is incapable of learning from its own mistakes in education policy making from its own research facilities (IES).
We need new ideas and more active roles for parents, their children, and their teachers in policies addressing the chronic problems of low achievement. Maybe our public schools need a respite from bureaucratic diktats in Washington, DC and their conduits in state departments of education. So-called reformers have run out of gas in their efforts to address low achievement. Maybe our schools need to be more responsive to what local administrators, teachers, parents, and students themselves see as the causes of low achievement. And maybe they should all be anonymously surveyed for their opinions on whether the curriculum and tests based on Common Core’s standards can address these causes.
Citation: Stotsky, S., & Holzman, T. (2015). The Costs of Federal Intervention in Local Education: The Effectiveness of America’s Choice in Arkansas. Nonpartisan Education Review/Articles, 11(2).
“Congress to Investigate School Turnaround Companies.” Education Week. http://blogs.edweek.org/edweek/campaign-k-12/2010/08/congress_to_investigate_school.html
Baroody, K. January 2011. “Turning Around the Nation’s Lowest-Performing Schools: Five Steps Districts Can Take to Improve Their Chances of Success. Center for American Progress. http://files.eric.ed.gov/fulltext/ED535862.pdf
Consortium for Policy Research in Education (CPRE). 2009. A study of instructional improvement. Philadelphia: University of Pennsylvania.
Cox, D. K. 2010. The Impact of America's Choice Design. Unpublished dissertation, University of Arkansas.
Mass Insight. June 2010. School Turnaround Models: Emerging Turnaround Strategies and Results. http://www.massinsight.org/publications/stg-resources/112/file/1/pubs/2010/07/20/Turnaround_Models_7_19_10.pdf
Mass Insight. November 2007. The Turnaround Challenge Supplement to the Main Report. http://www.massinsight.org/publications/turnaround/52/
McCrary, J., J. Ziobrowski,, and J.C. Bojorquez. May 2008. America's Choice in Arkansas Implementation and Achievement after One Year. WestEd. http://www.industrycortex.com/datasheets/profile/1254217250/third-party-wested-americas-choice-in-arkansas-implementation-and-achievement-after-one-year-joseph-mccrary-jaclyn-ziobrowski
NGA Center for Best Practices. March 2011. "State Strategies for Fixing Failing Schools and Districts.” Issue Brief. http://www.nga.org/cms/home/nga-center-for-best-practices/center-publications/page-edu-publications/col2-content/main-content-list/state-strategies-for-fixing-fail.html
Ravitch, D. and W. Mathis, 2010. Review of “College- and Career-ready Students.” Boulder, CO: National Education Policy Center.
Scott, C. 2008. A Call to Restructure Restructuring: Lessons from the No Child Left Behind Act in Five States. Center for Education Policy. http://www.eric.ed.gov/ERICWebPortal/recordDetail?accno=ED503798
Slavin, R., A. Cheung, C. Groff, and C. Lake. (2008). Effective reading programs for middle and high schools: A best-evidence synthesis. Reading Research Quarterly, July-September 43 (3), 290-322.
Stevenson, Z. May 18, 2009. Office of Elementary and Secondary Education, USDE, Cover letter to Commissioner Ken James.
Stotsky, S., C. Goering, and D. Jolliffe. (March 2010). Literary Study in Grades 9, 10, and 11 in Arkansas, Unpublished monograph, University of Arkansas. http://www.uark.edu/ua/der/People/Stotsky/ArkansasLiteracyStudy_9-11.pdf
Stotsky, S. April 8, 2015. Blog on Education News. Why Don’t Senators Alexander and Murray Want to Improve the Education of Low-Income Students? http://www.educationviews.org/dont-senators-alexander-murray-improve-education-low-income-students/
Stuit, D. 2010. Are Bad Schools Immortal? The Scarcity of Turnarounds and Shutdowns in Both Charter and District Sectors. http://edexcellence.net/publications/are-bad-schools-immortal.html
U.S. Department of Education. August 26, 2009. “Obama Administration Announces Historic Opportunity to Turn Around Nation's Lowest-Achieving Public Schools.” Press Release. http://www2.ed.gov/news/pressreleases/2009/08/08262009.html
U.S. Department of Education. 2010a. A Blueprint for Reform: The Reauthorization of the Elementary and Secondary Education Act. http://www2.ed.gov/policy/elsec/leg/blueprint/index.html
U.S. Department of Education. 2010b. Research Behind the Obama Administration’s Proposal for Reauthorizing the Elementary and Secondary Education Act. Washington, DC. http://www.ed.gov/blog/2010/05/research-behind-the-obama-administration
U.S. Department of Education (2010c). College-and Career-Ready Students. Washington, DC. http://www2.ed.gov/policy/elsec/leg/blueprint/college-career-ready.pdf
 Donald Trae Holzman died in 2013 before we could finish this paper. It is dedicated to his memory. We gave a preliminary version of it at the Arkansas Political Science Association Conference in Conway, Arkansas, February 25, 2012, titled “How Effective Are For-Profit Education Turnaround Partners? America's Choice in Arkansas, 2006-2010.” At the time, Holzman was a graduate student at the University of Arkansas.
 NCLB requires states to measure "adequate yearly progress" (AYP) for school districts and schools receiving Title I funds with the goal of all students reaching the proficient level on reading/language arts and mathematics tests by the 2013-14 school year. States must define minimum levels of improvement as measured by standardized tests chosen by the state. AYP targets must be set for overall achievement and for subgroups of students, including major ethnic/racial groups, economically disadvantaged students, limited English proficient (LEP) students and students with disabilities.
 As is well known, statistically significant effects, a common finding in research studies using large numbers of students, are not necessarily practically significant effects. Statistically significant effects for a new program in quasi-experimental studies using large numbers of students do not necessarily translate into meaningful increases in student achievement in reading or mathematics.