Traditional school districts have increasingly embraced giving students more options within their systems in recent decades, in response to growing charter and private school competition and the desire of district parents for more specialized programs and services for their children. How does district-centric school choice impact the pupil demographics of individual buildings, long governed by race- and income-based residential proscriptions and district attendance lines? A trio of researchers has released a new working paper examining that question.
The research team, led by Francisco Lagos of the Inter-American Development Bank, looks at the 300 largest school districts in the United States to identify which of them implemented a choice-driven admission system any time between 2000 and 2022. They sought districts where at least 50 percent of schools participate in a centralized application process; where applications are submitted through a single, districtwide platform (rather than through individual schools); where parents are allowed to rank multiple choices for their children; and where student assignments occur via random allocation—including lottery for oversubscribed seats—without regard to academic, behavioral, or attendance records.
Seventy-one districts met all these criteria. The remainder comprised a comparison group of so-called “non-choice” districts, which may have incorporated some parental options but did not meet all the criteria. Choice districts were larger on average than non-choice districts (88,942 students compared to 42,844), number of buildings (128.7 versus 57.2), and total area population (611,000 versus 264,000). Choice districts had a higher percentage of charter schools within their borders and exhibited higher levels of school-aged poverty. The researchers use a difference-in-differences design, exploiting the staggered adoption of school choice programs over time, to determine the impact of these programs on school composition, racial exposure[1], and racial/ethnic distribution.
Overall, the introduction of a centralized school-choice admissions system was found to “modestly redistribute students” across district buildings. The exposure of non-White students to White students increased by 1.8 points in choice vs. non-choice districts, and Black students’ exposure to White students specifically rose by 1.6 percentage points. However, White students experienced reduced exposure to all other racial and ethnic groups, likely led by the somewhat surprising finding that the introduction of school choice boosted the number of White students enrolled in those districts. In total, although centralized school choice did impact individual buildings, it did little to alter district-wide patterns of racial and ethnic segregation.
These findings lead Lagos and his team to conclude that school choice should not be introduced into traditional districts as a means to reduce pre-existing segregation patterns, which—it should be stated—is their area of concern and motivation for undertaking the study. But it should not be construed as the primary concern of any of the districts involved. As pervasive as residential- and income-influenced segregation is within traditional school district boundaries, other research has shown that school choice for parents is more about academics, extracurriculars, and school quality than it is about racial or ethnic concerns. It seems more likely that districts are concerned primarily with these important points—specifically to keep their families from moving to charter, private, and micro-schools, or to homeschooling—than they are with unwinding decades of history.
SOURCE: Francisco Lagos, Jason Saltmarsh, and Jing Liu, “The Effect of Centralized-Admission School Lotteries on Between-School Segregation: Evidence from 300 Largest School Districts in the United States,” Annenberg Institute at Brown University (December 2025).
[1] “Exposure” is defined as “the average racial and ethnic composition of the schools attended by students of a given group. This index reflects the probability that a randomly selected student from group A attends a school with students from group B, weighted by the group A population.”
