Review finds states failing new test of child welfare system

WASHINGTON -- Not a single state has passed a rigorous test of its ability to protect children from child abuse and to find permanent homes for kids who often languish in foster care.

The 32 states evaluated so far could lose millions of dollars from the federal government if they fail to fix problems within a few years. Neither Missouri nor Illinois has been evaluated yet.

The problems of child welfare get periodic attention, usually following the tragic death of a child. The Child and Family Service Reviews are the first time federal officials have tried to measure how well children are faring across state systems created to protect them -- but that often fall short.

The reviews ask whether children are bouncing from one foster home to the next, never able to put down roots; whether siblings taken from their parents are kept together or pulled apart; whether it takes a state too long to finalize adoptions or to send children back to their biological parents.

Affected are nearly 550,000 children in foster care and an estimated half million others living at home but under state supervision.

"There is a lot of work to be done," said Joan Ohl, commissioner of the Administration for Children, Youth and Families. "It's a daunting task."

In the past, states were evaluated on bureaucratic benchmarks. Now, the questions are how many children are abused again after entering the system and whether parents are getting promised help.

The reviews merge dozens of questions into seven "outcomes" measurements.

Fourteen states have failed all seven. An additional 14, plus the District of Columbia, have failed six of the seven, and four states failed five. No state has passed more than two.

"We set a very high bar and we don't apologize for that bar," Ohl said in an interview.

Problems were found in every state:

In Tennessee, the agency did not respond to abuse reports in a timely manner nearly 30 percent of the time.

In Michigan, more than one in four parents with children in foster care said they had not received needed services such as parenting classes or drug treatment.

In Ohio, 27 percent of the time the agency did not make a diligent effort to help children in foster care maintain connections to family and community.

States acknowledge the problems and welcome a clear set of benchmarks for improvement, said Robert Lindecamp, director of the National Association of Public Child Welfare Administrators. "States don't have a problem with having a high standard," he said.

One problem common to all states is the huge load handled by child welfare caseworkers. The reviews found that families do better when caseworkers make more visits, but that requires additional money that budget-strapped states are not inclined to spend.

After the first round of reviews, scheduled for completion next year, states must write improvement plans. A second round of tests will determine if states made promised changes. If not, they could lose some of their federal child welfare money.

While the seven outcome measurements are the heart of the reviews, states are evaluated on their overall systems -- for instance, do computer systems work and is training done properly. That brings the number of benchmarks to 14.

Maximum penalties proposed range from $130,000 in Delaware, which failed six of seven measures, to more than $18 million for California, which failed all seven.

Whether states will make significant changes is an open question. Ohl says the examples of innovation by the states "are still more of the exception than the rule."

"We are still receiving program improvement plans that merely scratch the surface in terms of the real improvements that must be made," she said.

Critics, including state officials and outside advocates and experts, say the reviews themselves are flawed.

The grades are based on statewide data submitted regularly to the federal government plus in-depth reviews of 50 cases selected randomly from each state.

Much of the state data is widely considered unreliable. The critics also say 50 cases, a fraction of any state's caseload, do not accurately represent the state.

The measurements are essentially snapshots of a moment in time, which can be misleading, rather than a look at what happens to a child over years.

For example, the reviews count how many of the children were reunited with their parents within a year and how many adoptions were finalized within two years. But neither measure looks at the entire caseload to calculate the likelihood of reunification or adoption.

Federal officials say the review paints an accurate picture and that the process marks a turning point in child welfare.

But it will take even well-meaning states a long time to fix the problems uncovered, said Richard Gelles, dean of the School of Social Work at the University of Pennsylvania.

"Some state systems are truly horrible," he said, "and no amount of accountability is going to make them jump from horrible to good in one leap."

On the Net

Child and Family Service Reviews: www.acf.hhs.gov/programs/cb/cwrp/