A new NBER working paper by Brian Jacob, Jens Ludwig and Douglas Miller examines how improved housing conditions impact child mortality rates in Chicago. The improvement in child mortality seems to apply only to girls, and not boys. The data come from Chicago’s resuscitated housing voucher system, from 1997 through 2005. Here’s the abstract:
In this paper we estimate the causal effects on child mortality from moving into less distressed neighborhood environments. We match mortality data to information on every child in public housing that applied for a housing voucher in Chicago in 1997 (N=11,848). Families were randomly assigned to the voucher wait list, and only some families were offered vouchers. The odds ratio for the effects of being offered a housing voucher on overall mortality rates is equal to 1.11 for all children (95% CI 0.54 to 2.10), 1.50 for boys (95% CI 0.72 to 2.89) and 0.00 for girls – that is, the voucher offer is perfectly protective for mortality for girls (95% CI 0 to 0.79). Our paper also addresses a methodological issue that may arise in studies of low-probability outcomes – perfect prediction by key explanatory variables.
The study builds on the findings of the federal government’s Moving to Opportunity experiment, which started in the mid-1990s and offered randomly chosen residents of public housing the chance to move to a wealthier neighborhood (poverty below 10%). Among adults, rates of obesity and mental health problems declined, but the effects were mixed on the risky behaviors of kids. Girls did better, while boys did worse.
Jacob, Ludwig and Miller looked deeper into the mortality statistics, and found the same thing. While moving from a disadvantaged neighborhood proved perfectly protective for girls, (none of the female youths in the study’s sample died after their families were offered housing vouchers) it actually increased the mortality risk for boys.
We also find that moving out of disadvantaged public housing does not have the same protective effects on mortality outcomes for male youth.
…It is interesting that in our data the suggestive (but not statistically significant) indications of increased mortality to male youth from residential mobility are concentrated among homicides, while declines in mortality to female youth are concentrated among deaths due to disease and accidents.
…Although our findings for homicides and accidents for boys are limited in their statistical power, the point estimates are consistent with the MTO studies, finding that boys were more likely to be injured or engage in other problem behaviors.
Further evidence of the “accident hump” at work?
The authors conclude that if the mortality rate in public housing overall was similar to their study’s control mean, then if every public housing family that applied for a voucher in 1997 had been offered one, the mortality rate for all youth black females in public housing (9,269 black females 18 and under living in Chicago public housing in 2000) would have declined by 5 per 100,000 (17 percent).