Whether the size of a school district has any effect on academic performance or cost efficiency is unclear, because the research isn’t conclusive, UNC-Chapel Hill representatives told a legislative committee.

Too little information exists for anyone to claim victory, said Rep. Bill Brawley, R-Mecklenburg, co-chairman of the Joint Legislative Study Committee on the Division of Local School Administrative Units.

“We know enough to know that we don’t really know,” Brawley said in a meeting Wednesday, March 28.

The committee is tasked with studying the pros and cons of breaking up large school districts and identifying challenges resulting from a deconsolidation.

The Charlotte-Mecklenburg and Wake County school districts are the two largest districts in the state, each serving more than 100,000 students. The average number of students in districts statewide is about 12,000.

Kevin Bastian and Eric Houck shared their findings with the committee.

Bastian is senior research associate and associate director of Education Policy Initiative at Carolina, UNC-Chapel Hill. Houck is associate professor of Education Leadership and Policy at UNC-Chapel Hill.

Bastian and Houck told the committee more research is needed to determine what effects, if any, school district size has on performance. This is in part because deconsolidation is a fairly new and uncommon occurrence.

Available research is mixed.

“There is no real consensus about the relationship between district size and cost,” Houck said. “The overall finding, although this is not universal, [is] the consensus seems to be that district size operates on efficiency on a U-shaped curve.”

Houck said some research shows districts are inefficient when they are very small or very large. A 2002 study published in the Economics of Education Review suggests the U curve starts at 4,000 students and ends about 15,000.

Houck said robust literature on the size of schools indicates smaller schools are sometimes better for student learning, but little attention has been paid to the size of schools within smaller and larger districts.

“It may be the case that smaller districts seem to be more efficient in terms of producing student test scores because they contain smaller schools, and it is the school’s size driving the findings not the district size,” Houck said.

If larger districts are deconsolidated but school size remains unchanged, breaking up those districts may have little to no impact.

In terms of student achievement, Bastian said there’s no real consensus on the impact of school district size, but several studies how that smaller districts are associated with more desired outcomes. A few other studies, including one from Denmark, suggest larger school districts have a positive impact on student outcomes.

Bastian said some studies found the socioeconomic status of a district determined whether the district’s size would have a negative or positive effect. As the socioeconomic status increased, the effect of a school district’s size goes from negative to positive.

Bastian and Houck cautioned lawmakers about some limitations of the studies they cited. A lot of the data comes from the 1980s and 1990s and doesn’t control for other variables.

“There are real questions about the extent to which they are isolating size as impacting the outcome they are looking at versus other things they are not controlling for,” Bastian said.

Furthermore, most of the studies don’t look at student-level data or performance growth. Few districts have deconsolidated, making it hard to find case studies.

“What is really key here is regardless of researchers of different perspectives —  researchers who have come to different conclusions about size — they will still say there is no optimal size,” Bastian said. “There is no one size fits all.”

At the next meeting, the committee will hear about innovative programs in various schools, such as Sugar Creek Charter School and Project Lift of Charlotte-Mecklenburg schools.