Schoolhouse Crock.

Why George W. Bush's education reforms won't change anything

My son Jacob will be 5 years old this summer, and I have had to face every parent's nightmare of discovering where and how to register him for kindergarten. As the director of an education policy program at a national think tank, I imagined that I had an advantage over the average parent. After all, my job is to evaluate charter schools and tax credits, public school choice and private school voucher programs, homeschooling efforts and privatized school management. Additionally, I am a member of a Los Angeles urban school improvement committee. Finding an acceptable school for my child, I assumed, shouldn't be difficult.

But my confidence began to wane when I started to explore the Web site for our neighborhood school, El Cerrito Elementary, in Corona, California. My heart sank as I reviewed test results and realized that the school to which my son is assigned does not have a Stanford 9 score above the 50th percentile. The Stanford 9 is a standardized test widely used by schools across the country to track educational achievement. El Cerrito Elementary's score is not that low by California's standards, but who wants to send her child to a below-average school?

I still held out the hope that I would be able to exercise some limited degree of school choice by enrolling my son in a quality public school. The school district had allowed the developer of a middle-class housing tract down the hill and across the freeway from my home to erect a brand new school. By doing so, the developer bypassed the normal per-dwelling school tax and finished building this state-of-the-art facility in less than one year. The new school, Woodrow Wilson Elementary, is even a little closer to my house (I live in a semi-rural area of Riverside County) than the "neighborhood" school that requires my son to ride the bus for 45 minutes each way.

Armed with this knowledge, I naively contacted the Corona-Norco Unified School District's "central registration" office to request an intra-district school transfer. A procedure I expected to be simple and direct turned out to be almost impossible. Presumably to discourage requests like mine, the district only accepts transfer applications one week per year, from the 1st to the 7th of December. The polite voice at the other end of the phone informed me that I would be welcome to apply in December for the following year.

Like millions of other parents stuck with low-performing public schools, my only options are relocation, home-schooling, or, given the decisive failure of the California school choice initiative last fall, investment of a small fortune in private school tuition.

What's more depressing is that there is no relief in sight. President Bush's widely touted education reform plan-- which has earned him kudos on the right and attacks on the left--will have minimal impact on American schools and will do little or nothing to improve education for the average kid. That's because the Bush plan deals almost exclusively with Title I, the major federal education program, which is designed to improve schooling for at-risk students. On paper, zeroing in on Title I, widely considered a colossal failure, makes sense. Ninety percent of America's school districts--some 23,000 schools nation-wide--receive grants under the program. But the Bush plan's focus is misguided and diverts attention from more sweeping education reform that would help more parents provide their children with a quality education.

What Is Title I?

In 1965, President Lyndon Johnson established Title I of the Elementary and Secondary Education Act as part of his Great Society program. The lofty goal of Title I has been to improve the basic and advanced skills of students who are at risk of failing in school. In particular, the program is designed to assist low-achieving children living in low-income areas where school funding is deemed to be inadequate. At $9 billion a year, Title I is the largest program of federal aid for elementary and secondary education. The money is used mostly to provide intensive math and reading instruction.

In Title I's 36-year history, the U.S. Department of Education has released two major longitudinal studies on the program's effectiveness: Sustaining Effects in 1984 and Prospects in 1997. The Sustaining Effects study demonstrated that the $40 billion spent on the program to that point had done little to improve the achievement of the children it was designed to help. Although the elementary school students showed slight gains over their peers, "By the time students reached junior high school, there was no evidence of sustained or delayed effects of Title I," wrote Launer R. Carter, director of the study, in Educational Researcher.

Thirteen years later, the most recent longitudinal study of the program found that even after the federal government spent another $78 billion (from 1984 to 1997), bringing the total spent on Title I to $118 billion, little had changed. "After controlling for student, family, and school differences between participants and non-participants, we still find that participants score lower than non-participants and that this gap in achievement is not closed over time," the authors of the Prospects study wrote.

Researchers could not discern any long-term achievement gains directly linked to the Title I program. The program tries to identify and serve the children who need the most help, but according to the study, "The services appear to be insufficient to allow them to overcome the relatively large differences between them and their more-advantaged classmates." Similarly, Wayne Riddle, an education analyst at the federal government's Congressional Research Service, analyzed the two federal longitudinal studies and five other Title I studies. His conclusion: "Title I participants tend to increase their achievement levels at the same rate as non-disadvantaged pupils, so gaps in achievement do not significantly change."

In 1999, the U.S. Department of Education released a congressionally mandated evaluation of Title I that seemed to show, based on results from the 1998 National Assessment of Educational Progress (NAEP), that the 1994 reauthorization of Title I had led to some increases in student achievement due to program reforms. The NAEP tests a sample of fourth-graders, eighth-graders, and 12th-graders from 40 states in writing, science, math, and reading. The test is considered the "nation's school report card" and is widely viewed as an independent measurement of public school achievement. The 1998 NAEP results initially appeared to show significant improvements in fourth-grade reading scores in nine states since 1994.

The progress reported in this study was largely fictitious, however. A skeptical parent in Kentucky, Richard Innes, discovered a problem with the 1998 NAEP reading scores. According to the official results, Kentucky was one of the most improved states in fourth-grade reading. But using data gleaned from the Internet, Innes discovered that the gains in some states, including Kentucky, resulted from the exclusion of students considered to be slow learners and those with learning disabilities. Innes asked this critical question: Can a state's scores be accurate when they don't include large numbers of low-scoring students? An analysis by the U.S. Department of Education confirmed that several states had inflated average reading scores by excluding greater numbers of special-education students from testing in 1998 than in 1994. The federal analysis established that more than half of the 36 states where the NAEP is administered had excluded significantly larger numbers of special-education students in 1998. Five states excluded substantially more non-English-speaking students than they had in 1994.

For example, Kentucky dumped test results for 10 percent of the students who were selected for its 1998 sample, compared with 4 percent in 1994. Louisiana ignored 13 percent in 1998, up from 6 percent in 1994. And Connecticut, the nation's highest-scoring state, removed 10 percent of the students selected to participate, compared with 6 percent in 1994. Not surprisingly, states with larger increases in total exclusions also tended to have larger score increases. When the test scores were compared on a realistic basis, Kentucky gained nothing.

Despite these flaws, the Department of Education's 1999 report showing student improvement from 1994 was widely cited in the education press and the general media. Incredibly, the Department of Education's own investigation into special education exclusions was never made public. I only discovered it accidentally when researching congressional testimony regarding the effectiveness of Title I.

The most recent results of the 2000 NAEP tests for fourth-grade reading paint a bleak picture for Title I: 63 percent of black fourth-graders, 58 percent of Hispanics, 47 percent of urban students, and 60 percent of poor children scored below "basic" in reading--which for all practical purposes means they cannot read.

Why Bush's Plan Won't Work

President Bush has proposed to increase the Department of Education's budget by 11 percent, to $44.5 billion. Assuming his budget is passed as is, Title I, which continues to be the largest single item in the federal education budget, would spend approximately $10 billion for a program that has consistently failed to produce any measurable results for close to four decades. (Democrats are pushing to boost Title I spending to $18 billion.) By comparison, charter schools in Bush's budget would receive around $175 million for start-up and facility costs.

The central component of the Bush plan would overhaul the Title I program for disadvantaged students by requiring states to develop systems of rewards and penalties to hold districts and students accountable for academic progress. Specifically, states would be required to test all students in grades three through eight in reading and mathematics every year, as a condition of receiving federal Title I aid. In a nod to federalism, Bush would not institute a national curriculum and would allow individual states to design their own diagnostic tests.

A voucher sanction in Bush's original plan has evolved, through the legislative process, into a $600 tutoring credit. It would allow parents at failing schools that do not make "adequate yearly progress" for three consecutive years to pay for extra reading and math instruction for their child.

That's not much help. But there's a bigger problem with the Bush plan: The current focus on failing Title I schools ignores the grim reality that many schools that have not been designated as failures are still not the kinds of places that most parents want to send their children. To fully appreciate this phenomenon, it is important to consider exactly how dysfunctional a school must become before it is designated a "failure."

Individual states vary widely in how they define what constitutes a "failing" school. Many states have set standards that deem a school's performance adequate even if less than half of its students meet state standards for proficiency. At least eight states have set their standard at or around the 40th percentile, and a few have set the standard even lower. In Alabama, for example, more than half of a school's students must score below the 38th percentile for the school to be put on an intervention track, and more than half must score below the 23rd percentile to immediately target a school for improvement efforts.

To appreciate how most schools escape the "failure" label, one must understand the notion of "adequate yearly progress" (AYP) and how it is measured. AYP is determined at the state level and is tied to meeting performance goals and state standards.

Some states require schools to meet an absolute target or performance threshold. In the president's home state of Texas, for example, a rating of "acceptable" means that at least 50 percent of students at a given school must pass the state assessment in reading, writing, and math. The 50 percent standard doesn't address the obvious: Do the parents of the 50 percent of students who did not pass the Texas Assessment of Academic Skills find this level of achievement to be "acceptable"? Such a low standard means that less than 1 percent of Title I schools in Texas are labeled as failures.

California schools must improve every year by 5 percent of the difference between their academic performance index and the state's performance target of 800. The formula is complex and is based on student scores on the Stanford 9 test. The bottom line is that in order to show AYP, my local school must gain a few percentage points each year on a standardized test.

Assuming that the school can actually achieve the mandated minimums each year, it could be two decades before my local school meets the state target score of 800. In effect, the state and the school district are willing to sacrifice the minds of the children who will be forced to attend this school over the course of the next 20 years while the school struggles to meet minimum AYP standards.

The real failures of El Cerrito Elementary, which can stand in for many schools across the country, become increasingly apparent when actual improvements in reading scores on the Stanford 9 test are considered. In 1998, the average second-grade test scores at El Cerrito Elementary were at the 36th percentile. By 2000, average second grade reading scores jumped to the 38th percentile. A similar small improvement in achievement has occurred for the other grades at El Cerrito Elementary.

As a parent, perhaps I should be proud of the progress my local school is making. Instead, I am shocked that the school's accepted rate of yearly academic progress could mean that my son will spend his six years of elementary school at an institution that will not come close to meeting California's statewide target goal of 800 by the end of his tenure there.

Why Vouchers Hardly Mattered

Even if the voucher component of Bush's plan had not died in Congress, it would have had little impact on parents who are stuck with low-performing schools. President Bush's plan mirrored Florida Gov. Jeb Bush's A-Plus education program, so the Florida plan is instructive as a model of how the president's plan would likely work over time. By offering vouchers to students at failing schools, the Florida plan is intended to motivate those schools to improve their academic performance. Each public school in Florida is assigned a grade, from A through F, based on the proportion of its students earning a passing grade on the Florida Comprehensive Assessment Test (FCAT). Students attending schools that receive two F grades in four years are eligible to attend a private school or to transfer to another public school.

In practice, few Florida students have actually received the promised vouchers. In the first year that students were eligible to receive vouchers under the program, a total of 53 students from two schools statewide got tickets to go elsewhere. In 1999, there were 78 public schools that received a failing grade based on their FCAT scores. If those schools got the same grades in 2000, they would have been sanctioned with vouchers. Miraculously, by year two of the A-Plus program, every school in Florida (including the 78 schools that had a failing grade the year before) managed to pull test scores up enough to avoid the voucher sanction.

Apparently, the public school establishment in Florida sensed an end to their monopoly and reacted accordingly. Dr. Jay P. Greene, an education researcher at the Manhattan Institute, recently analyzed FCAT test scores covering the initial two years of the A-Plus program. He found that "schools that received F grades in 1998-1999 experienced increases in test scores that were more than twice as large as those experienced by schools with higher state grades."

All of this suggests that the public school establishment will react to the threat of a club, particularly the voucher sanction. But, beyond the minimal threshold of sanction, inertia seems to set in. The never-mentioned tragedy is the fact that, despite improvement, hundreds of thousands of students are still attending inferior schools.

Imagine the frustration of Florida parents with children in those mediocre schools who are now denied the opportunity of a real education for their children simply because an 18-point improvement earned a formerly failing school a D grade? The average failing school did improve but still showed an average score of only 272 out of a possible 500 points. That is the effective equivalent of a doctor being wrong in her diagnosis almost as often as she is correct. Will it soothe those parents to know that President Bush plans to continue with Title I spending of $10 billion a year to ensure that their failing local school, and others like it around the nation, will improve from 50 percent of students failing a standardized test to 47 percent failing the same standardized test? Will they rest easy knowing that average reading test scores will possibly improve from the 39th percentile to the 42nd? Will such results persuade inner-city parents, and all parents for that matter, that truly robust school choice is not a necessary option in the United States?

Failing History

President Bush's proposed program is just the latest attempt to fix Title I. The program has been reformed several times over the last 30 years. Completely absent from the reform debate are the Department of Education's own Title I program evaluations, which demonstrate that after spending more than $150 billion, the program has not improved achievement for disadvantaged students.

Aside from that, education reform is so focused on poorly performing schools and students that education for average and above-average students doesn't even make it to the radar screen. Even if the purest form of Bush's education plan was implemented and the sanctions were enforced, most schools will still be mediocre and most parents will still have few options between their state-mandated public school or paying private school tuition on top of public school taxes.

Real education reform would give parents a way to find a better quality education now, instead of waiting years for their failing or simply mediocre public school to improve. Until the federal government allows real education reforms--such as universal tax credits or actual vouchers that are at least equal to the federal portion of per-pupil spending--it will have little impact on the educational experience of students who need better schools while they're still in school.

Meanwhile, the rest of us are supposed to settle for the status quo because El Cerrito Elementary School and others like it around the nation can claim they have made adequate yearly progress (second-grade reading scores average in the 42nd percentile instead of the 38th!). My son Jacob would still be forced to attend a school that is not considered failing even though reading scores average well below the 50th percentile. For most parents and students, whether eligible for Title I or not, the Bush plan is, at best, too little, too late.

Lisa Snell is the director of the Education and Child Welfare Program at the Reason Public Policy Institute.

Free Lunch

Title I's formula for determining aid--and its recipe for fraud.

Individual schools receive Title I funding based on the percentage of students that are eligible for the federally subsidized free-lunch program. Though the lunch program is designed to provide food to low-income students who might otherwise go hungry, its guidelines do not require schools to verify the parental income of students who enroll. The process to qualify for a free lunch comes down to parents self-reporting their income on a form that is turned in to their local school. Federal free-lunch program administrators argue that the program has Little potential for abuse because "the worst that happens is a kid gets a free lunch."

Federal free-lunch data, however, are used as one of the main poverty indicators for school districts and are linked to many other local, state, and federal funding streams. So any fraud in the free-lunch program is quickly multiplied. And rest assured that school districts recognize the program's multiplier effect and work hard to sign up students. Consider this typical account from the St. Petersburg (Florida) Times last summer: "When Gulf High School assistant principal Pat Haynes sees needy children lunching on cookies or a bag of chips, she knows the kids are jeopardizing more than just good nutrition. Those kids are also cutting into their school's ability to cash in on its share of millions of dollars in grants and government rebates designed to benefit low-income schools."

The result of such incentives is both soft and hard abuse of the free-lunch program and, by extension, many other programs designed for students from low-income households. Many school districts offer free ice cream and other tokens to kids who return their forms, even if the students aren't eligible for the program. Schools have mailed multiple enrollment forms to parents and some have even taken to calling families at home to ask them to enroll.

For example, in October 1999 The Baltimore Sun reported that the principal of Patterson High School took to the intercom and announced: "Guess what is coming? Pizza Party! Everyone connected with the school from parents to staff members will eat pizza, get a free T-shirt, and listen to a disc jockey if poor students can get their parents to fill out an application for free and reduced-price school lunches." In 1999 about 50 percent of Patterson High School's 2,200 students qualified for free meals. By using the pizza party strategy, school principal Laura D'Anna increased that to more than 70 percent of students in 2000.

White fraud from the federal free-Lunch program may have small consequences for the program itself, the cost of fraud to other education programs such as Title I may be much greater. Districts that have tried more strictly to verify parental income have met with resistance.

When the Bergenfield school district in New Jersey required parents to submit more extensive income documentation after the number of students in the free-Lunch program doubted in one year, the New Jersey state nutrition program forced the district to reinstate all students who were disqualified from the program. Bergenfield district business administrator Tom Egan argued that there were inconsistencies in some of the applications, including applications from students whose parents had homes valued at $350,000.

In some cases, however, school boards openly overreport their school lunch data. In 1999, the Clifton school board in Bergen, New Jersey, voted 5-4 to report that exactly 20.16 percent of public school students were poor enough to qualify for free lunches, instead of the actual number, which was 19.19 percent. The difference was significant: If the number dropped below 20 percent the district would lose $4 million in aid. As the Bergen County, New Jersey, Record reported in November 1999, board president Wayne Demikoff said while casting his vote for the higher number, "I cannot, in good conscience, vote for something that is going to devastate the budget."

Strangely, the board did not necessarily break the law by reporting the higher number. The lower figure was arrived at by verifying the income of 80 percent of the families that applied for a free lunch. The federal government requires districts to verify only 3 percent.

The free lunches bring the district both short-term fiscal benefits and long-term financial consequences. Each year, Clifton receives millions of dollars in aid for needy students. At the same time, New Jersey districts that have more than 20 percent of their students on free lunches are supposed to begin providing full-day kindergarten and half-day preschool.

Lisa Snell