An Analytical Approach to Setting School Boundaries


When the nation’s eighth-largest school district, Hillsborough County Public Schools in Tampa, Fla., set out to establish new high school boundaries in 2008-09, administrators expected to receive numerous complaints from concerned parents and community members.

Revising boundaries, closing schools and choosing sites for new ones to increase diversity and improve efficiency are complex and sensitive procedures. Hillsborough County administrators knew it would be a challenge to evaluate and convey complicated policy trade-offs and, at the same time, gain community support for the boundary decisions for nine of the district’s high schools.

LazarusWilliam Lazarus

To reduce the emotion from the process and optimize efficiency, the school district partnered with a research firm, Seer Analytics, to conduct rigorous, evidence-based analytics. This allowed the district to take an impartial, analytical approach to the boundary-changing process and put off introducing maps until optimal boundary solutions were found using values-based decision criteria. Without maps, no one would know which households and neighborhoods would be affected by the boundary modifications until after the best solutions had been chosen — based on the decision-making rules agreed upon by district staff, board members and the community.

Mapping Alternative
In most boundary-change procedures brought about by the opening of a new school or the need to close one, staff members begin with maps. After reviewing current attendance lines, they sketch and modify new boundaries, run geographic information system queries, produce spreadsheets, and decide on a set of proposed attendance lines.

SeerAnalytics approached the process in a different way by starting with this question: Based on which criteria and decision-making rules do we want to develop the boundaries?

The project team engaged the community early on, conducting a series of public meetings to discuss priorities and determine the criteria that analytical software would use to generate possible boundary scenarios. With input from board of education members, staff and parents, the team concluded that transportation costs, utilization balance, (i.e., avoiding the situation where some schools were overcrowded when others were under-enrolled) and diversity criteria would form the basis for the evaluation of boundary options.

By removing maps from the equation and setting decision rules based on community values, the project team communicated the message that boundary solutions would be generated without considering specific communities and households. Everyone would be treated impartially and fairly. As one team member said, the team “couldn’t guarantee equity of outcome but could ensure the basic fairness of the process.”

After the criteria were accepted, the project team began a high-tech analysis. First, it designed an index to measure relative diversity without using racial or ethnic data of individual students. Using the decision rules, the team then built a multi-objective geo-spatial optimization model. After thousands of model runs, multiple boundary options were produced, which gave different weights to the criteria.

Seventy-nine of the best boundary solutions, which were fully documented in terms of the school-specific utilization and transportation trade-offs, were presented to district staff and board members for review. Diversity index calculations also were provided, which allowed decision makers to assess the diversity results of each possible boundary solution. Staff chose four scenarios they considered most effective in balancing building use while reducing transportation costs and increasing the diversity mix.

After these four scenarios were chosen, maps were introduced for the first time. Staff determined which maps to produce only through use of the decision rules and the weightings agreed on by the school board and community members at the beginning of the process.

No Complaints
After staff created maps of the potential boundary solutions, the maps were presented and discussed at community meetings. The project team listened to parental concerns about boundaries splitting subdivisions in two, transportation challenges, and the impact the new boundaries would have on the availability of some academic programs and extracurricular activities. Staff addressed concerns on a case-by-case basis and made some minor boundary modifications as a result.

Involving the community through the entire process helped the district gain broad support for the new school attendance lines. When the boundaries were presented to the school board for final adoption, the result was unprecedented in Hillsborough County history — not a single parent or community member spoke out against them.

Through the new boundaries, building utilization was balanced with the maximum difference between “full” and “empty” schools at only six percentage points. In addition, the boundary changes enabled the district to trim annual transportation costs by hundreds of thousands of dollars. Diversity was increased by as much as 11 percent at some schools, with only a 2 percent increase in the cost of transportation at those sites.

Because of the project’s success, the Hillsborough County Public Schools was subsequently awarded one of 10 federally funded demonstration project grants to develop new boundaries for the county’s 46 middle schools.

As they made decisions through unbiased analytics centered on community values, Hillsborough County Public Schools proved that school districts can overcome emotionally charged school boundary battles and achieve positive results.

William Lazarus is CEO of SeerAnalytics in Tampa, Fla. E-mail: