Reducing student absences at scale by targeting parents' misbeliefs

parents of 28,080 high-risk students, grades K-12

US public school students identified as at risk of chronic absenteeism, one large district

Individually randomised mailing experiment. Parents received one of three personalised information treatments repeatedly across the school year, or no additional mailing (control). The active ingredient is correcting parents' systematic underestimate of how many days their child has already missed. Single-district trial; the reported outcomes are attendance outcomes only - total absences and chronic absenteeism. No academic achievement outcome is reported.

The most effective treatment arms reduced chronic absenteeism by 10% or more. Effects were comparable across grade levels. There was a household spillover - absences also fell among untreated cohabiting siblings. Cost-effectiveness on the attendance outcome is more than an order of magnitude better than conventional absence-reduction practice. The paper does not test, and does not claim, that achievement improved.

Low for the attendance effect itself - randomisation makes genotype orthogonal to treatment assignment. But this trial cannot speak to whether raising attendance raises achievement, because achievement was not measured. The household spillover is worth noting for the opposite reason it is usually cited - it demonstrates that absence is a family-level behaviour, which is exactly why cross-sectional absence-achievement gradients are confounded by shared family characteristics.