« Back to Results

Race and Family Background as Determinants of Educational Outcomes

Paper Session

Saturday, Jan. 7, 2023 10:15 AM - 12:15 PM (CST)

Hilton Riverside, Grand Salon D Sec 19 & 22
Hosted By: American Economic Association & Committee on the Status of Women in the Economics Profession
  • Chair: Francisca Antman, University of Colorado Boulder

Ending Exclusionary Discipline in the Early Grades: Effects and Implications

Sarah Komisarow
Duke University
Ezra Karger
Federal Reserve Bank of Chicago


School suspensions – and racial disparities in their use -- have received increasing public attention and prompted states and school districts across the country to reform school discipline policies. Despite state- and district-level initiatives aimed at reducing the use of suspensions, little is known about the causal effects of suspensions on student outcomes, particularly for students in early grades.

In this paper we take advantage of a policy change in Charlotte-Mecklenburg Schools (CMS), a large district serving nearly 150,000 students in North Carolina, in order to measure the short- and medium-run effects of early grade suspensions on student outcomes. In 2017, CMS passed a new policy that removed principal discretion and centralized out-of-school suspension decisions – instead, the school district superintendent was charged with deciding on out-of-school suspensions for all kindergarten through second grade students. We investigate the effects of this policy on the use of exclusionary discipline (e.g., suspensions and expulsions), student behavior, and test scores.

Using student-level data from North Carolina, we investigate the effect of this policy change on the use of exclusionary discipline, student behavior (e.g., reported disciplinary infractions), and test scores. We find that the policy change dramatically reduced out-of-school suspensions in grades K-2 and – importantly – did not produce substitution to other forms of classroom exclusion or discipline (e.g., in-school suspension). We do not find evidence of any statistically significant effects on test scores.

Mitigating Racial Bias in Teachers' Assessments of Students

Maria Zhu
Syracuse University


Substantial research has found that teachers display racial biases in their assessments of student achievement. This study provides novel evidence that these biases are consistent with the predictions of a statistical discrimination framework, in which teachers rely on group averages to inform their evaluation of individual students in the presence of imperfect information. A key finding in this analysis is that increasing the amount of interaction teachers have with students reduces biases. Specifically, reducing class sizes decreases the racial gaps in a teacher's student assessments between White and minority students.

Effects of Parental Disability on Children’s Schooling: The Surprising Role of Parental Education

Leah Lakdawala
Wake Forest University
Prashant Bharadwaj
University of California-San Diego


We show that the effects of parental disability on children’s schooling investments are much more negative when parents are highly educated using repeated cross-sectional data from the American Community Survey (ACS) from 2008 to 2019. To isolate the causal effects of parental disability, we focus on children of veterans who become disabled during military service. Within this group, we find that the gradient in children’s private school attendance with respect to the severity of parental disability is much steeper for children with college-educated parents than for children with non-college-educated parents. We provide evidence that these heterogeneous effects are driven by differences in foregone earnings across more and less educated parents. Though parental disability generates larger reductions in parental labor supply for less educated parents, the reduction in earnings is greater for more educated parents. These findings illustrate that parental disability may have larger negative impacts on children that are not traditionally considered to be part of a vulnerable group.

Family Background, Educational Attainment and Earnings: The Limited Value of “Test-Score Transmission”

Naomi Friedman-Sokuler
Bar Ilan University
Moshe Justman
Ben Gurion University


Universal public education is the great equalizer, leveling opportunities among children from different family backgrounds. In practice, the ability of affluent parents to privately invest resources in their children’s education persistently generates differences in children’s educational outcomes. The effect of these educational gains on intergenerational mobility remains elusive because the distinction between human capital enhancing investments and those that can loosely be described as investment in test-taking skills is difficult to pin down. This paper analyzes “test-score transmission” in secondary education and the extent to which this translates into higher labor market earnings, separating the role of achievement, attainment and study field choices. We develop an analytical model of family background, achievement and earnings and utilize administrative longitudinal data from Israel’s centrally administered and funded public education system to estimate the relative importance of these different types of investment. Using a differences-in-difference style setup, we define two types of family background: “first-generation” (FG) students, low income parents with no education beyond high school; and “second-generation” (SG) students, high-income parents with at least some post-secondary education, focusing on students in the top two eighth grade achievement quartiles. We find evidence of test-score transmission, SG students score higher than FG students on matriculation and university entrance exams, conditional on eighth grade achievement. As a result, SG are 14.7 percentage points more likely than FG to be accepted to a university, as opposed to college, a gap largely accounted for by the differences in the entrance exam. However, these advantages have a much smaller impact on students’ study field choice and labor market earnings, observed as they approached thirty. We find that in 2018, FG students in the top quartile of eight-grade test scores earned 6% less, on average, than SG students in the top quartile, and 20% more than second-quartile SG students.

Briana Ballis
University of California-Merced
Carycruz Bueno
Wesleyan University
Anita Mukherjee
University of Wisconsin-Madison
Rebecca Dizon-Ross
University of Chicago
JEL Classifications
  • I2 - Education and Research Institutions