Sibling Spillovers of Integrated Student Support

• In Event: Identifying Peer Effects: Evidence from Classrooms, Friendships, and Families

Abstract

Evaluations of school-based interventions typically focus on direct participants, often overlooking within-family externalities (Figlio, Karbownik, & Özek, 2023). While research indicates that school policies can produce significant spillover effects on untreated siblings (Nicoletti & Rabe, 2019), such dynamics remain understudied in the context of Integrated Student Support (ISS) models.

ISS models address out-of-school barriers to learning through comprehensive needs assessments, individualized support plans, and tiered service delivery (Walsh et al., 2014). Although ISS is known to improve academic and behavioral outcomes for treated students, it remains unclear whether these benefits extend to untreated siblings in the same household.Because ISS delivery frequently requires family engagement, benefits may diffuse to untreated children through shared resources, reduced parental stress, or improved household dynamics. Consequently, traditional evaluations restricted to direct participants may significantly underestimate total program impact (Angelucci & De Maro, 2015).

This study examines whether City Connects, a prominent ISS intervention, produces spillover effects for untreated children living with a sibling who directly participated in the ISS intervention. We utilize school attendance as our primary outcome, as it is an earlier predictor of academic success (Balfanz & Byrnes, 2012) and highly responsive to shifts in family support systems (Kearney et al., 2023).

Using six years of administrative data from a large urban district in the U.S., we identified sibling pairs via address and surname matching. Our analytical sample includes 1,836 untreated children whose siblings received the intervention within the preceding two years, compared against a control group of 23,429 children from households with no ISS intervention exposure. Employing a cohort difference-in-differences design on pooled cross-sectional data (grades 1–12), we estimate the impact of having an ISS-treated sibling on attendance while controlling for individual characteristics—such as race/ethnicity, gender, English learner status, and special education status—alongside year, grade, and school fixed effects.

Findings reveal that untreated children with an ISS-treated sibling attended school significantly more often, averaging 4.25 additional days per year (SE = 0.94, p < 0.001) compared to the control group. These results remain robust across multiple specifications. Our findings suggest that ISS interventions generate meaningful within-household spillovers, implying that the social return on investment for integrated student support is substantially higher than previously estimated.

Author

• Haibin Jiang, Boston College
  Presenting Author