Children are shaped in profound ways by the neighborhoods where they grow up. Perhaps this sounds like common sense (why else do we fret over where to raise them?). But it's borne out by research, too. High-poverty neighborhoods can be bad for children's health, school performance and even cognitive development. Low-poverty ones, meanwhile, often mean they have access to better schools and do better academically as a result.

It makes sense, then, that when we subsidize housing for poor families, we should try to help them into homes in the kind of neighborhoods that have lower poverty, less crime and higher-quality schools. Most government rental assistance, however, barely does this at all.

Federal housing programs help the families of about 4 million low-income children. But, according to a new analysis from the Center on Budget and Policy Priorities, a small sliver of those children -- about 15 percent -- actually live in stable, low-poverty neighborhoods, even with that help from the government.

While we're focused on helping these families into housing, in short, we're not paying much attention to where that housing is located. As a result, many poor families use federal subsidies to continue living in deeply poor neighborhoods — the very places that we know can hold back their kids.

The Department of Housing and Urban Development has three main programs (often run locally) that offer housing assistance to the poor. About a million households live in traditional public housing. Section 8 rental assistance subsidizes housing for another million families in designated but privately owned buildings. Then about 2 million households use Housing Choice Vouchers that they can spend on the private market.

The above chart, from the CBPP report, suggests that Housing Choice Vouchers are much more effective than the other two programs at helping families live in census tracts where many of their neighbors aren't poor, too. This isn't surprising, since vouchers are the most flexible form of assistance. The government can't very well pick up public housing projects and move them to better neighborhoods.

Housing vouchers, by CBPP's count, also make a significant difference in the ability of poor black and Hispanic families to raise their kids in low-poverty neighborhoods (this is less true for whites). Only about 7 percent of poor black children nationwide live in low-poverty neighborhoods. But the same is true of nearly 17 percent of poor black children in families using vouchers:

This data suggests we should be doing a lot more to leverage the power of the one housing program — also the largest housing program — that seems to help families get into better neighborhoods.

CBPP has several suggestions on this front, none of which require vast new resources or congressional action. HUD could better incentivize local housing agencies to help families deploy vouchers to move into low-poverty neighborhoods (and we could start measuring the success of such agencies by how well they do this). Similarly, local agencies could discourage families from using their vouchers in high-poverty places, while offering more assistance to help them figure out how to do that.

It would also help if these vouchers were even more portable — if families could more easily take them from the jurisdiction of one housing agency to another community nearby.