On the Determinants of Young Adult Outcomes: Impacts of Randomly Assigned Neighborhoods For Children in Military Families

Using the quasi-random assignment of 760,000 children in U.S. military families, we show that neighborhood attributes experienced during childhood have powerful impacts on SAT scores, college-going and earnings. For earnings and college going outcomes, location during high school is twice as important as location during elementary school, and for SAT scores, location during middle school has the strongest impact. There is little evidence of positive interactions in neighborhood quality across ages groups. Importantly, the same locations benefit children with equal potency across race or sex. Twenty years of exposure in a 1 standard deviation "better" county raises SAT composite scores by 10 points (1.8 percentiles), raises college attendance by 1.7 percentage points, earnings by 2.2 percentile points, and lowers EITC receipt by 10%. Impacts are three times more potent when we measure neighborhood quality at the zip code level: twenty years of exposure to a one (county level) standard deviation better zip code raises college going by 6.7 percentage points, SAT composite by 38 points and income percentile at age 25 by 6.1 points. By equalizing average neighborhood quality for Black and White families, we estimate that the Army's quasi-random assignment reduces Black-white earnings gaps among the children of Army personnel by 23%.