Are Student Absences Worth the Worry in U.S. Primary Schools?

ECLS-K nationally representative kindergarten/grade-1 cohort plus North Carolina statewide administrative records for grades 4-5

US primary school students, kindergarten through grade 5

Observational, with rich covariates and prior achievement, in two independent datasets. This is the canonical "how much does absence matter" descriptive paper, and it is explicitly framed as a relationship rather than a causal estimate. No instrument, no lottery, no within-family comparison. Included here as the observational benchmark against which the quasi-experimental estimates are compared, not as evidence for a causal claim.

Absences are associated with modest but statistically significant decreases in achievement, and the association is approximately linear in the number of absences. The gradient is two to three times larger for North Carolina 4th-5th graders than for ECLS-K kindergarten/1st graders, i.e. the association grows with grade level. Association is similar in urban, rural and suburban schools; larger for low-income students and English language learners, especially in reading. Unexcused absences are associated with twice the harm of excused absences.

High, and the excused/unexcused split is the tell. An unexcused absence differs from an excused one mainly in what it says about the family and the child - parental follow-through, child conscientiousness, externalising behaviour - not in the quantity of instruction missed. A design that finds unexcused absences "twice as harmful" as an identical number of excused absences is measuring child and family type, not lost instructional time. The larger association for low-income students runs the same way. Treat every number here as an upper bound.