We Want Kids to Grow Up in Safe Families. So Let’s Measure That.

Our cultural norm of relatives and kin stepping in to help struggling families is ingrained in our history and tradition. The U.S. Supreme Court has conferred constitutional protections on the sanctity of the extended family [See Moore v. City of East Cleveland, 1977]. Federal law encourages states to “consider giving preference to an adult relative over a non-related caregiver when determining placement for a child.”

Placements with relatives and kin provide children with greater stability, which results in fewer behavioral problems and greater ties to communities. Yet our formal child protection systems continue to undervalue and underutilize relatives and kin.

The Family First Prevention Services Act, the new law of the land, is our golden opportunity to rethink our child protection options, move beyond our heavy reliance on foster care, correct our metrics, and make our reliance on relatives and kin great – and meaningful – again.

Georgia’s child welfare leadership set a target of 50 percent relative placement. In other words, half of foster kids placed with a relative on any given day, a point-in-time measure. Does that target metric reflect our cultural norms, legal and policy directives, or what we’d want for our own kids? Sort of … maybe … well, not really.

If Georgia doubles the number of kids removed from their homes (which it did) and then places a higher proportion of them in relative foster homes, some flaws in the strategy emerge. Georgia might meet its target, but consider the trauma of removal, the burden on casework staff, and the cost to taxpayers. Also, don’t assume relative foster caregivers are well compensated – one-third receive no foster care maintenance payments in Georgia.

What if we instead made our metric – our child protection bottom-line, if you will – maximizing the time that all kids spend safely in a family: birth, relative, kin or adoptive?

That’s known as a person-time metric, in contrast to point-in-time. Administrative data systems are actually well suited to measure this one. They’re built to track payments for each child’s foster care placement setting (even at $0) over time. They don’t track the complex relationships between the caregiver and the child well, but they do distinguish related from unrelated placements.

So minimizing the time kids spend with non-family is a better fit to state tracking systems. The picture this person-time metric paints is inconsistent with the picture painted by the point-in-time metric (percent of kids in relative foster care) that Georgia and others have chosen.

Kids in Maryland spend the least time with non-family (half the national rate), and a high proportion (37 percent) are placed with relatives while in foster care. So that’s somewhat consistent, but the two metrics barely correlate when states are compared. Consider the outliers at the extremes: Virginia and Arizona.

At 6 percent of kids in relative foster care, Virginia ties South Carolina for the second-lowest (thanks to Kentucky) point-in-time proportion. But Virginia removes so few kids to foster care that, overall, Virginia’s kids get to spend one-third less time with non-family than kids in other states. Virginia also has very low recurrence of maltreatment, so their low removal rate may be helping with safety as well as avoiding trauma.

Arizona leads the nation on the point-in-time measure (48 percent in relative foster care), but, paradoxically, kids in Arizona spend 49 percent more time with non-family than in other states. How could that be? It only starts to make sense in the context of Arizona’s high removal rate, high rate of adoption, and long stays in foster care. And let’s not forget that adoptions from foster care carry the cost of terminating rights, even in the third of those adoptions where youth end up with relatives.

Virginia and Arizona take different approaches to child protection, and their differing degrees of reliance on relatives become clear only with metrics that are unambiguous, supply rich context, and reflect our cultural norms and legal and policy directives.

Starting from a clear statement of our desired outcome, a bottom-line if you will, and some concrete examples, it’s clear that we can go terribly wrong in our interpretations of the data we are given. The most common stat we see in evaluating our child protection systems’ use of relative and kin resources is the percentage of kids in foster care placed with relatives on a day.

At best, this stat is incomplete; at worst, it’s misleading. A higher percentage of kids in foster care placed with relatives isn't necessarily good because each of those kids still endured the trauma and disruption of removal, and a lower percentage isn't necessarily bad because it allows for the possibility that children are living with family without the disruptive state intervention of foster care. If we ask instead for states and the feds to measure the outcome we want – less time with non-family, more time with family – it's clear that more of that is good and less is bad. Isn't that simpler?

Melissa Carter is executive director of the Barton Child Law and Policy Center at the Emory University School of Law. Andy Barclay is the founder and advisory board chair of the center.