Harvard Study Shows That Voucher Threat Helps Raise Test Scores In Florida

Issue Brief

Because of its extensive use of various forms of parental choice to reform public schools, Florida has become something of a laboratory for education researchers. Now two Harvard University scholars have found that the vouchers offered under Governor Jeb Bush’s A+ Accountability plan are helping to promote gains in student achievement.

Indeed, Martin West and Paul Peterson of the Program on Education Policy and Governance at Harvard’s John F. Kennedy School of Government found that Florida’s use of school choice has been more effective than the choice provisions of the federal No Child Left Behind Act (NCLB) in bringing about test score improvements. However, they acknowledged NCLB as a force for results-oriented school reform and a factor in states’ revisions of their school accountability plans, including Florida’s of the A+ plan in 2002. Since introduction of the revised A+ plan, overall test score performance in Florida has risen on both the FCAT (the state’s assessment) and the norm-referenced SAT-9, they noted.

Under the A+ plan, Florida students become eligible for vouchers to transfer to a private school if their public schools receive an “F” on accountability measures twice in a four-year period. West and Peterson found that Florida’s fourth and fifth graders made modest but significant gains in math and reading when their schools were in imminent peril of losing students to vouchers. Students in schools that received their initial “F” in 2002 scored from 4 to 5 percent of a standard deviation higher the following year than did students in “D” schools that did not face an imminent voucher threat.

Meanwhile, the stigma of publicly receiving a low grade seemed to provide some reform impetus to “D” schools as well. Their students improved by 5 percent of a standard deviation relative to students in “C” schools.

Under NCLB, schools that fail to make adequate yearly progress toward a state-set level of academic proficiency for two consecutive years are found to be “in need of improvement” and their students are supposed to be given a choice of a better-performing public school located in the same school district. But the Harvard researchers found that Florida schools subjected to this public-school-choice threat under NCLB showed no improvements in student achievement.

Why the difference between the A+ plan and NCLB in the impact of choice? Peterson commented on two factors:

(1) Almost 75 percent of Florida’s schools failed to make adequate progress in 2003. By contrast, only 2 percent of Florida schools received an “F” under the state’s accountability system in 2002, while 8 percent received a “D.” With regard to NCLB, “sky-high failure rates can undermine the accountability threat. When everyone is criticized, no one is going to take the criticism seriously.”

(2) NCLB’s school-choice provisions are too anemic to provoke much of a response from public schools in need of improvement. Fewer than 1 percent of students eligible to transfer from one public school to another under NCLB accountability are doing so, Peterson noted. (The relative roles of apathy and foot-dragging of school bureaucracies in informing parents of their rights may be debatable.)

In their paper, the researchers also pointed out that certain features of the A+ plan are “considerably more rigorous” than NCLB. For instance, students at the schools that fail twice under the state’s standards gain the opportunity “to receive a voucher to attend any school — public or private — within the school district or elsewhere.” NCLB choice extends only to public schools within a student’s school district, and some districts may not have enough high-performing public schools from which to choose.

West and Peterson stress that their study looks only at how NCLB choice operates in Florida, and does not gauge how that provision is working in other states or the impact that NCLB is having overall. But if robust choice in the form of state vouchers is having a positive effect on student achievement in Florida, it is reasonable to conclude that full-fledged federal vouchers could help make the promise of No Child Left Behind a more universal reality.

President Bush’s original NCLB proposal in 2001 included vouchers much like the reform brother Jeb championed in Florida. However, Congressional leaders balked at giving parents in substandard public schools that full measure of choice.

The next time NCLB is up for amendment — in the 2007 reauthorization of the law, if not sooner – policymakers would be well advised to seriously consider the beneficial impact vouchers could have on equal opportunity and student achievement.