Voucher Effects on Educational Attainment (Graduation, College Enrollment)

What the evidence says

The available evidence on "voucher effects on educational attainment" in this corpus comes entirely from housing-mobility voucher research (the Moving to Opportunity experiment and its literature), not education/school vouchers. Within that body of evidence, the clearest and most quantified result is an increase in college attendance tied to a specific condition: children whose families moved with an experimental housing voucher before the child turned 13 show a 2.5 percentage point increase in college attendance measured between ages 18-20, a 16% increase relative to the control group's attendance rate of 16.5% (Chetty 2016). The same age-restricted group is also reported to attend better-quality colleges, though the source does not quantify that effect. A second source, drawing on the same research base, corroborates the direction: children who moved before age 13 were more likely to attend college and had higher earnings in their mid-twenties than control-group children, though it likewise does not give a magnitude for the college-attendance increase (Baldassarri 2017).

Set against this age-restricted finding, two sources report no measured effect when the outcome or population is not restricted to early movers. Housing vouchers show no measured effect on high school graduation, with the underlying estimate not statistically different from zero and carrying larger confidence intervals than other outcomes examined in the same analysis (Page ME 2024). Separately, a review of the housing-voucher literature reports no measured effect on children's educational attainment in medium- and long-term follow-up analyses, distinguishing this null result from measured beneficial effects on mental health and risky behaviors among young women in the same studies (Baldassarri 2017).

Taken together, the evidence in this corpus does not support a general claim that housing vouchers increase educational attainment. The positive, quantified result is confined to college attendance (and reported college quality and later earnings) among children who moved before age 13. On high school graduation specifically, and on educational attainment measured across the full study population, the evidence shows no measured effect.

Where the evidence splits

No contradictions in the current sources. The apparent tension between the general null effect on educational attainment (Baldassarri 2017; Page ME 2024) and the positive effect on college attendance (Chetty 2016; Baldassarri 2017) is accounted for by the sources themselves: the positive effect is reported specifically for the subgroup of children who moved before age 13, not for the study population as a whole.

Thresholds and dose-response

Age at time of move sets the condition under which the college-attendance effect appears: children assigned to the experimental voucher group before turning 13 show a 2.5 percentage point (16%) increase in college attendance and are reported to attend better colleges, while the same body of research finds no measured effect on educational attainment when analyzed across the medium- and long-term follow-up population without this age restriction (Chetty 2016; Baldassarri 2017).

Open empirical questions

• Whether the 2.5 percentage point college-attendance increase persists beyond ages 18-20.
• Whether the reported college-attendance effect for pre-13 movers reflects college enrollment, college completion, or both.
• Whether the null effect on high school graduation reflects a true absence of effect or insufficient statistical power, given the larger confidence intervals noted around that estimate.
• Whether the null effect on general educational attainment holds across different levels of voucher compliance and take-up (one source notes 47% voucher take-up and segregated neighborhood placement in the studies it reviews) or across subgroups beyond age at move.