Why do poor parents have children who also grow up to be poor? One possible reason is that poor families do not have access to credit that would allow parents to invest more in the improvement of the human capital of their children. The conventional policy recommendation for this diagnosis is to increase transfers to poor families in order to remove their credit constraints.

The expansion of the Earned Income Tax Credit (EITC)—which uses the tax system to transfer money to low-income households—has been shown to increase standardized test scores. But critics of this research argue that factors unobservable to researchers but correlated with EITC receipt are responsible for children’s success, not the EITC transfers themselves.

Increasingly, economists use clever research designs that involve an element of random assignment, much like clinical trials of new pharmaceuticals, to provide more conclusive evidence of a program’s effectiveness or ineffectiveness. Recently, three researchers used a policy change in Chicago to test the effects of a change in housing subsidies.

Unlike many other welfare programs, housing subsidies are not given to everyone who qualifies for them, but are handed out on the basis of availability. In 1997, for the first time in 12 years, Chicago accepted applications for housing vouchers. About 82,000 people applied out of 300,000 poverty households in Chicago at the time. The applicants were randomly assigned a position in the waiting list. The first 35,000 on the list were told their number and that they would be offered a voucher within three years. The rest were told that they would not receive vouchers.

By 2003, 18,000 of the first 35,000 applicants had received vouchers. The Chicago Housing Authority had issued as many vouchers as it could fund, and stopped offering any new vouchers.

In a study I review in my “Working Papers” column in the current Regulation, Brian Jacob, Max Kapustin, and Jens Ludwig examine the outcomes 14 years later for children whose families “won” the Chicago housing vouchers versus the children of families that were told they would not receive a voucher. Families that won the lottery received a very large positive income shock—the equivalent of $12,000 a year—relative to the average income in the sample ($19,000 a year). If income alone allows families to improve the human capital in their children, we should see results from this experiment.

The authors find very few effects on schooling, crime, or health outcomes—and none were significant. “Our estimates imply that extra cash transfers beyond the current level provided in the United States are likely to have a smaller impact per dollar than the best-practice educational interventions explicitly designed to improve children’s human capital,” they write. Their results are consistent with the findings of sociologist Susan Mayer, who concluded in her 1997 book What Money Can’t Buy (Harvard University Press) that there is “little reason to expect that policies to increase the income of poor families alone will substantially improve their children’s life chances.”