By Tiffany Danitz Pache, VTDigger.org
Is Vermont spending too little or too much on schools? The answer, according to experts hired by the state, is that Vermont may have spent about 10 percent more than necessary in the 2014-2015 school year. However, they say the findings don’t mean the state should start sharpening the budgetary scythe.
Last year, lawmakers included $300,000 in Act 46 to study better ways to finance the state’s schools. A nearly 300-page report titled “An Evaluation of Vermont’s Education Finance System,” authored by Allan Odden, Lawrence Picus and Michael Griffith, says that by applying an “evidence-based” school funding model, Vermont could have saved $164 million, or 10 percent, on pre-kindergarten through 12th grade education during the school year they examined.
Special education is the reason for the lion’s share of the costs. The next big-ticket item is the high number of staff in relation to students in Vermont’s schools. Costs associated with numerous small systems that lack efficiencies also drive up the price of education, the researchers said.
“I think that you have a well-funded system that is highly equitable, which is what your court requires it to be,” Picus told lawmakers on the House Education Committee last week. “Proceed cautiously in attempting to achieve savings, because the complexities of school finance may lead to unintended consequences. This will not happen tomorrow. It will take time.”
Picus encouraged lawmakers to continue considering organizational changes such as unifying school districts into a more centralized system in an effort to provide additional resources and services more effectively.
“Focus on management and then on how to ensure that all teachers are really highly qualified,” he said. “I would start by having coaches in schools to help teachers be better teachers and adding resources in schools for struggling students.”
The “evidence-based model” developed by Picus, Odden & Associates does not directly link funding to student performance, but identifies a base per-pupil spending level as well as any needed extra resources that will provide all students with the opportunities they need to become ready for college and a career.
William Mathis is a member of the State Board of Education and teaches on the subject of education finance. He said the study’s basic conclusions are right but no surprise: staffing ratios are too low, and schools have too many special education aides. Speaking from his own expertise, not for the board, he suggested legislators should now do some studying.
“They should look at what is going on with special education and why those costs are there,” he said.
Vermont is spending $294 million for special education and identifies 16 percent of the state’s students as needing special education services. The Picus-Odden model spends about $140 million less and reflects a belief that the state is identifying too many students as needing these services. For their model, they use an estimate of 12 percent for special education students. “That takes $20 million off the table right there,” according to Picus.
In the study, they rely on new research showing that struggling students often can be helped with extra tutoring and other resources and can then return to the general education population—never being labeled as needing special education, which ushers in prescribed resources and services at additional costs. For this reason, Picus said, their model includes $95 million for extra help to bring struggling students up to speed. The investment is offset by savings on special education, he said.
Their model also reduces paraprofessionals and eliminates aides. They said the latest research shows that special education students with the greatest needs learn more and do better in a classroom led by a skilled classroom teacher as opposed to employing paraprofessionals and aides.
Mathis agreed that Vermont can reduce the number of paraprofessionals working in special education, but he said he doesn’t believe it’s a good idea to “categorically” cut out aides. “You still need someone to push a wheelchair or monitor an emotionally disturbed student,” he said. “We can and should reduce the number of aides over time, but to say they should go out completely is probably unrealistic.”
The evidence-based model also assumes that children with severe disabilities—estimated at about 2 percent of the student population—are fully funded through the state at a cost of $100 million. “That way if they end up in a district it doesn’t wipe out the school’s budget,” Picus said.
But Mathis doesn’t think basing state funding for severely disabled students on a percentage of the general student population would reflect school reality.
“Special education kids are not evenly distributed throughout the state,” he said. “There are pockets, and so for people with low [special education] child counts they would have extra money, and those with a high child count [in special education] would not have enough, so that would be dis-equalizing.”
Regardless, the House Education Committee chair, Rep. David Sharpe, D-Bristol, said during last week’s committee meeting that he wants to move in a direction where “we spend less on special education.”
The other big contributor to Vermont’s education costs is student-to-teacher ratios, which are higher in the evidence-based model even after modifying it to better adapt to Vermont’s recommended class sizes. For pre-kindergarten through sixth grade, the Picus study used a student-to-teacher ratio of 17-1. In the upper grades it used a 20-1 ratio.
While Mathis agrees the student-to-teacher ratio of 10.03-1 in Vermont is one of its biggest cost drivers, he also said Picus’ choice of 17-1 could account for the 10 percent savings he found in his report. “If you change that number even by one pupil it is going to have a huge impact,” Mathis said. “If Picus had used, say, 16-to-1 then his cost estimate of Vermont’s spending being 10 percent over could disappear or be substantially reduced.”
The size of individual schools is another area where the researchers adjusted national norms to suit Vermont.
National research indicates that elementary schools with between 400 and 600 students and secondary schools that range from 500 to 1,000 students are the most effective and cost-efficient. Accordingly, the evidence-based approach for most states has been to build model elementary schools with 450 students and secondary schools serving up to 600 students. To scale down for Vermont’s sake, the researchers created a “prototypical” school serving 357 students in pre-kindergarten through fifth grade, as well as one with 238 students and one with 119.
At the middle school level, the model has three prototypes with 450, 300 and 150 students. The high school models serve 600, 450, 300 and 150 students.
This estimation is the biggest fault Mathis finds with the report. “Basing it on a model school of 357 students—which would be by Vermont standards huge—the question becomes: can you extrapolate from a model school of 357 down for Vermont,” he said.
The bottom line is that small schools cost more to run. “When you get down to a school with 50 or less children, it is a whole different problem,” Picus said. “It adds up to a fairly high amount of per-pupil spending.”
While playing with the data to see what would happen, Picus created a supervisory union with 4,000 students and found $32 million in savings.
Sharpe found this interesting since, he said, during Act 46 deliberations Legislators were told that if it were fully implemented, Vermont would see $30 million in savings. The report states that an “important part” of the study included support for the anticipated cost savings of reorganization.
The researchers offered a final warning about comparing Vermont’s school funding system with the evidence-based model because certain Vermont features such as tuitioning, Small Schools grants, “phantom students” and local control “make a perfect comparison impossible.”
In other areas as well the model differed from Vermont’s current situation. It assumes that supervisory unions operate as one school district, establishing economies of scale in central operations. It offers full-day pre-kindergarten that operates within the public school system, instead of Vermont’s blended model where parents have vouchers for their children to attend private or public school pre-kindergarten. It does not include instructional aides at any education level, whereas most Vermont school districts employ “substantial numbers of aides in their instructional programs,” the report states. Finally, the evidence-based model appears to have fewer administrators than are generally found in Vermont school districts.
Mathis said he hopes lawmakers don’t react to the report by mandating staff or teacher ratios. “Be careful. You cannot micromanage effectively from Montpelier,” he warned. “There are too many variations from communities to towns, from one place to another. Treating it as a uniform system won’t fit particularly when you have such small schools and small populations like we do.”