The Risks & Benefits of School Integration for Participating Students: Evidence from a Randomized Desegregation Program
Researchers
Peter Bergman | UT-Austin
Key Finding
Across-district school integration programs can boost academic outcomes for male minority students but need to address other risks to student well-being. Male students who were offered a spot in a cross-district racial integration program in California had better state test scores than those who didn’t win a spot and were 8 percentage points more likely to enroll in college. However, they were also more likely to be classified as “special needs” and be arrested for non-violent offenses.
* While the study examined male and female students, the transfer program had no effect on female students. Effects described in this summary pertain to male students only.
. . .
Other Findings
The cross-district transfer program resulted in minority students attending schools that are 73 percentage points more white.
English, science, and history scores on state tests for students offered a spot in the program were at least 0.20 standard deviations higher. However, students offered a spot in the program were also 8 percentage points more likely to be classified as “special needs” at one point in their K-12 academic career and 4 percentage points more likely to be arrested after age 18, though overwhelmingly for driving offenses related to traveling further from home to attend college.
School resources, though important, were unlikely to be the sole driver of the program’s impact.
. . .
Methodology & Data Highlights
Field experiment examining a cross-district school transfer program for students in the Ravenswood City School District, Northern California.
Summary
Fifty years after the Supreme Court mandated racial integration of America’s public schools, many remain largely segregated. In 1968, 77% of Black students and 55% of Hispanic students attended majority-minority public schools. Four decades later, 74% of Black students and 80% of Hispanic students attended majority-minority schools. [1]
A broad swath of research has demonstrated how “white flight” and suburban migration have hampered efforts to integrate schools within majority-minority districts. [2] In the first paper to study short- and long-term effects of desegregation across districts, Peter Bergman finds that transferring students to low-minority share, higher-income school districts introduces both risks and benefits to participating male students.
Short and long-term impacts of the cross-district integration program
The study reveals that the cross-district transfer program resulted in minority students attending schools that are 73 percentage points more white while also boosting test scores and college enrollment rates for male students.
While no detectable transferring effect is shown on state math test scores, Bergman finds that English scores of transfer students were 0.20 standard deviations higher, science scores were 0.15 standard deviations higher, and history scores were 0.28 standard deviations higher.
However, male students offered a spot in the program were also 8 percentage points more likely to be classified as special needs at one point in their K-12 academic career–a rate that is 57% higher than the rate for students who were not given a spot. In contrast, there was no significant impact of a transfer offer on students ever receiving gifted and talented classification.
Overall, 14% of transfer students were classified as "special needs" at least once over the course of their high school career, while only 6% were classified as "gifted" at least once. While research on these classifications hasn’t produced definitive answers on how these they affect student outcomes, there is evidence that minority students are disproportionately classified as “special needs” and that the classification contributes to racial achievement gaps.
Over the long term, the study found that students who were offered a spot in the transfer program were 8 percentage points more likely to attend college. For students who accepted the spot and ultimately transferred, this effect was concentrated at two-year colleges. Students given a spot in the program were also significantly more likely to attend three or more semesters of college. Overall, 39% of transfer students enrolled in college at one point.
Previous research has shown that school-choice lotteries reduce the likelihood of arrest. [3] Bergman finds evidence to the contrary, showing that male students offered a spot in the program were 4 percentage points more likely to be arrested after age 18. Importantly, however, these arrests were overwhelmingly driving-related, for example driving on a suspended license. While an offer to transfer did increase the likelihood of drug-related arrests, there was no significant effect on violent or property-related arrests.
Bergman presents evidence that most of the increase in arrests is driven by students who attended school farther away from home (an average of 28 miles from the Ravenswood City school district). He argues that the increase in arrests is best explained by transfer students spending more time in areas with higher police presence where they are more likely to encounter racial profiling.
Finally, the study finds small–and not statistically significant–effects of an offer to transfer on the likelihood of registering to vote (1 percentage point) and voting (negative three percentage points).
On the whole, the benefits—and the risks—of the integration program accrued almost entirely to male students, a commonly observed phenomenon in educational studies.
Background on the program
The program studied in this paper grew from a class-action lawsuit against a group of school districts and two counties in Northern California. The lawsuit settlement required that each of the eight districts participate in a transfer program if a district’s student body is more than 40% white.
As a result, students in Ravenswood City School District were given the opportunity to apply to transfer, via a lottery, to one of seven other school districts, and each district was required to receive a fixed number of students according to their student enrollment at the time of the settlement. [4]
Applications were restricted to rising kindergarten, 1st, and 2nd-grade students. The research sample included 2,410 applications to the transfer program from 1998 until 2008. Most applicants to the program were Hispanic, Black, or Pacific Islander.
To study short-term program outcomes, Bergman analyzed state test scores in math, English, science, and history from 2nd through 8th grades. This data contained information about students’ special education status (if any).
To study long-term program outcomes, Bergman linked application records to National Student Clearinghouse data on college attendance. This data includes the length of enrollment, enrollment status, and degree obtained for more than 3,600 public and private institutions across the United States. He also matched application data to arrest records for students aged 18 or over at the time of the data merge, and to voter registration and voting history records from a state database.
School spending matters, but doesn’t tell the whole story
To understand why the transfers were so impactful, Bergman ran a test to understand the role of spending per pupil, exploiting the fact that spending per pupil varied widely across receiving school districts.
Figure 2 from the study shows that differences in school resources, though important, were not the sole driver of the program’s impact.
Ultimately, he finds that, for a given student, moving from Ravenswood’s spending per pupil to Palo Alto’s, which is 62% greater, would boost college enrollment for the student by three percentage points. That increase is less than half of the program’s real-life impact, suggesting that school resources, though likely important, were not the sole driver of the program’s impact.
Cross-district transfer programs must address risks to student well-being
Given the limited potential for integrating schools within school districts that are majority-minority, this paper provides important perspectives for policymakers on the potential to improve student outcomes by integrating schools across districts.
This paper finds, from the standpoint of student well-being, the overall impact of a cross-district transfer program in California was decidedly mixed and suggests that integration programs should consider how to mitigate the outside-school risks participants may encounter.
Footnotes:
[1] Segregation 2.0: The New Generation of School Segregation in the 21st Century, Dana N. Thompson Dorsey, 2013.
[2] The Supreme Court’s 1974 decision in Milliken v. Bradley also played a role by determining that public school systems were not responsible for desegregation across district lines except under very specific circumstances.
[3] See Cullen et al (2006), Deming (2011) and Dobbie and Fryer (2015)
[4] At the time of the study, Ravenswood, which had a student body that was predominantly Hispanic (64%) and Black (24%) had the second-highest student-teacher ratio of any school in the program, the lowest proportion of students classified as special education, the highest students classified as Limited-English Proficiency (LEP), the second-lowest per-pupil spending, and the lowest average proficiency level.