Broken adoptions, buried records: How states are failing adoptees

Once a child is adopted from foster care, it’s as if they are reborn in the eyes of many child welfare agencies.

In required data reports to the federal government, these agencies remove evidence that would illuminate the child’s past journeys through the system. They take away the ability to trace details of a child’s round trip from foster care to adoption and back again when an adoption fails, or to examine what might have led to a better outcome.

The result, a USA TODAY investigation found: No one knows how well each state is fulfilling its mission of finding children their forever homes.

At stake are the fates of more than 50,000 children adopted out of foster care every year. The federal government funnels about $3 billion a year to families who adopt from foster care, now the leading type of adoption in most states, according to data compiled by USA TODAY.

Cortney Jones, a child welfare advocate who spent 10 years in foster care and lived through a broken adoption, said following the paths of foster children into adoptions could boost the odds that adoptions succeed.

“The kids end up back in the system because you didn’t track it,” Jones said. “You didn’t provide the support. You didn’t do your due diligence.”

There’s no reason that tracking can’t happen, experts say. A national adoption and foster care tracking system has been in place for nearly three decades. The federal government is in charge of making sure states supply accurate data.

Yet numerous states erase information essential to linking adoptions from foster care to adoption outcomes. They create a brand-new identification number for the child in the federal system, either at adoption or on reentry to foster care. Some, drawing on policies rooted in an era of secretive adoptions, break the data connection deliberately.

It’s just one example of how the federal Adoption and Foster Care Analysis and Reporting System falls far short of its aspirations, creating a jumble of disconnected facts instead of a reliable resource.

Federal authorities have allowed the inadequate reporting to continue, despite a law Congress passed in 2003 to seize foster care money from states with bad data. Audits since then by the U.S. Administration for Children and Families, which hands out adoption subsidies and manages the database, have flagged states with a wide array of data problems.

States have dragged their heels on making improvements, and under a new rule, the children and families administration says it won’t start imposing fines until 2023.

Confronted with the problem of spotty case tracking, the agency downplayed its legal authority over adoptions, which it said were “largely a state issue.”

“Data like this is not all subject to federal oversight and, therefore, is not necessarily in the purview of state or federal child welfare systems,” Aysha Schomburg, associate commissioner of the Children’s Bureau at the children and families administration, said in a written response.

How prescriptive federal authorities can be about consistent ID numbers is murky. But the agency said that if states don't keep consistent IDs, they still are required to report whether children entering foster care were previously adopted and whether they had other visits to foster care. That, the agency acknowledged, may not always happen.

Penelope Maza, a statistician who helped launch the system, called the federal government’s failure to crack down on states “a slap in the face.”

“Our whole strategy for data quality was through penalty,” said Maza, who retired from the children and families administration in 2008. “If they thought there was a penalty, they’d improve their data.”

‘You want us to make it up?’

Maza earned a Ph.D. in sociology from Ohio State University, studying intergenerational social mobility. At the federal children and families administration, a project she pitched became the one that spanned her career: creation of a national database to track every U.S. foster child and every adoption from foster care.

The work wasn’t quite what she had imagined.

“It was a nightmare,” she said.

The plainspoken researcher said she had to compromise constantly in the face of competing demands from social workers, politicians and fellow academics. She learned to prod and push to keep the ball moving among bureaucrats who she thought didn’t seem to care.

Up until that point in the 1980s, state and federal child welfare officials kept spotty state data with wildly varying definitions. The national statistics were so poor that if someone called her agency from Capitol Hill seeking a figure, she used to say, “You want us to make it up?”

Maza was frustrated. The country desperately needed a way to see how each state was serving children from abusive homes.

“No one knew how fast kids went through their system, if they were slower than other states and why were they?” Maza said.

She decided to keep the database simple enough for everyone to participate. Federal officials wouldn’t demand every clinical detail about every child. They couldn’t house notes from every case file. They had to accommodate varying levels of data knowhow among the states.

The Adoption and Foster Care Analysis and Reporting System launched in 1993, creating a standardized list of data each state had to submit twice yearly.

The government invested heavily in the technology that helps states manage adoptions and foster care stays internally.

The U.S. Department of Health and Human Services reported that from 1994 to 2001, it paid states $821 million to create the in-house databases that now feed AFCARS. States told the Government Accountability Office they spent another $1.1 billion in state and local funds.

Once the reporting system was in place, Maza took to the road. She crisscrossed the country to convince child welfare workers of its value. She felt good about what she considered her conservative approach. States were getting to work.

Then, she said, things got ugly.

No penalties for states’ messy records

Over the years, auditors with Maza’s agency performed routine checks on how each state was doing. Problems surfaced frequently.

In Georgia, 70% of foster care data required technical corrections, according to a 2011 report. Ohio underreported children’s mental health diagnoses, a 2014 audit found. Missouri “needs to ensure that all parties involved in a case take ownership for the accuracy of the data,” federal reviewers wrote in 2009. New York failed to report runaways from foster care, the federal government found in 2013.

Incomplete and inaccurate record-keeping persists.

The federal data system is supposed to flag, yes-no-can’t tell, whether a child entering foster care had been adopted before. USA TODAY found the adoption flag consistently missing or marked as “unable to determine” in the records of more than 400,000 kids served by the child welfare system from 2008 to 2020.

Some Washington foster children were listed in the federal database year after year as having had no adoption in their past. Yet they were marked as previously adopted the moment a new family adopted them. Alerted to the problem by USA TODAY, the state’s child welfare agency found a “system error” and plans to address it, said spokeswoman Nancy Gutierrez.

Early on, Maza’s agency tried getting states to shape up by handing down fines. That, she said, is when states pushed back – hard.

After 12 states joined forces to appeal their penalties – and won – the children and families administration announced in 2002 that it would abandon fines. The agency instead would “encourage states to make the improvements in as timely a manner as possible.”

But the idea of penalizing states hung on. Noting a GAO report pinpointing widespread problems, House Majority Leader Tom Delay, R-Texas, and Sen. Chuck Grassley, R-Iowa, in 2003 publicly called on human services officials to consider reviving fines for bad foster data. Months later, Delay added mandatory fines to an adoption incentives bill signed into law in December 2003.

Making it happen was far harder than it sounded, though. In 2008 and again in 2015, the federal children and families agency made a dead-end attempt to enact rules that would expand the data standards Maza created and enforce the penalties Congress had authorized. States responding to the 2015 proposal called it “onerous,” “of little value” and an “added burden.”

Without new rules for penalties, states took their time.

A decade after receiving a negative audit of their foster data, North Carolina authorities still hadn't made fixes, noting in 2017 that changes were "imminent.” Work then “paused in 2018 due to issues with the system’s usability for county child welfare workers,” Bailey Pennington, a spokesperson with North Carolina’s Department of Health and Human Services, wrote in an email to USA TODAY.

Federal officials told Pennsylvania’s child welfare agency in 2015 that its decision to leave data collection to counties was creating inconsistent and incomplete data. Seven years later, Pennsylvania still has not built the statewide system recommended by the children and families administration.

“We anticipate the system development of this project to begin in 2023,” Brandon Cwalina, an agency spokesperson, said by email.

New IDs erase children’s pasts

One of the biggest shortcomings in the national foster care database is that in most cases, it can’t be used to track children after an adoption fails. One reason may be an early 20th-century ethos that walled off any detail of a child’s biological family after an adoption.

States remove names before sending records to the federal foster care and adoption database, which charts the histories of individual children and is open to anyone. Federal guidance tells states instead to assign a unique but encrypted ID number that travels with the child “regardless of where the family/child moves” within the state.

But most states send the feds a fresh ID, either when an adoption from foster care is finalized or when a former adoptee reenters the foster system. A 2020 report submitted to the children and families administration found that only 16 states had AFCARS IDs that allowed kids from failed adoptions to be linked to prior foster care records.

Some states sever any link between pre-adoption and post-adoption records, even in their own databases. Others preserve IDs for internal use. Nancy Rolock, a professor at Case Western Reserve University in Cleveland, said she was able to trace adoption failures in Illinois because a state employee happened to be keeping children’s new and old IDs in one Excel spreadsheet.

“I spent two years trying to do the same thing in Wisconsin,” Rolock said. “It was impossible."

Information about the child’s birth family, early cases of abuse and mental health diagnoses are washed away because the old ID is unknowable by practically anyone but a caseworker around long enough to remember the child.

The practice of issuing new IDs to children after they’ve been adopted from foster care is well known to federal authorities. Schomburg, at the children and families administration, acknowledged by email that the fact a child came from a broken adoption “is not always accurately obtained by the states" because of the ID gap.

The agency said further that it does require states to accurately report whether an adoptee has previously spent time in foster care but that it cannot make states keep a consistent ID that would unlock detailed histories.

Some states, when asked why they issued new IDs after adoption, cited technical limitations they’re working to overcome. Others said issuing new IDs was a deliberate choice.

In South Dakota, “the state treats the entry as a new case because the original record is sealed,” wrote Laurie Gill, cabinet secretary for the state’s Department of Social Services, in an email.

Help USA TODAY investigate adoption

Are you an adoptee, parent, community member or public and private employee who can help us learn more about adoption issues? We want to hear from you about disrupted and dissolved adoptions.

Fill out this form

Iowa “has a practice of delinking the identity of the adopted child by issuing a new unique ID in the system at the time of adoption,” said Alex Carfrae. As in South Dakota, Carfrae cited a state law sealing records of an adoption and the termination of biological parents’ rights.

Arkansas Department of Human Services spokesman Gavin Lesnick cited “Arkansas business practice.” Oklahoma issues a new ID number because it is “a closed adoption state,” spokesperson Casey White said.

It’s unclear how submitting an encrypted ID number to the AFCARS system would reveal an adopted child’s identity.

According to Mark Testa, a professor emeritus of the School of Social Work at the University of North Carolina at Chapel Hill, these policies are a throwback to when society shamed young women who became pregnant “out of wedlock.”

“It had to do with attitudes of secrecy and confidentiality over adoption,” Testa said.

Today, however, attitudes have changed. Many adoptions are “open” from the start – meaning birth parents have a relationship with the adoptive parents – and some states now grant adoptive children the right to open their birth records.

Clinging to outdated state policies on privacy for reasons few can articulate has led to overprotection of children’s identities, Testa and other child welfare researchers said. It comes at the expense of accurately documenting each child’s path through the state system.

Federal authorities cannot scrutinize what happened to people like Alyssa Davis, whose adoption from foster care in New Mexico failed when she was 14. Researchers can’t see what led her to be removed from her birth family, they can’t evaluate risk factors in Davis’ demographic profile, and they can’t study whether policies such as pre-adoption supports and adoptive parent screening are helpful in cases like hers.

Even Davis can’t find out all that happened. She remembers snippets, but when she finally received her foster care records at 18, only her most recent placements came back. Her foster history before the adoption was invisible.

Richard Barth, a professor at the University of Maryland’s School of Social Work, said there’s no excuse for state child welfare officials to hold back data revealing what happens post-adoption.

“Since most, if not all, of children adopted out of foster care are receiving subsidies, the states should have the responsibility to track the success of those adoptions,” said Barth, who has written dozens of studies on child welfare and adoption.

Changes coming but debate goes on

People who oversee the nation’s foster care and adoption records succeeded in enacting rules for an upgrade, including penalties, in 2020. States can start being fined for chronically bad data next year.

The plan was supposed to be wrapped up in 2016. But when President Donald Trump took office, his executive order limiting federal regulation put the rules on ice. In the meantime, agency officials scaled back the new data they would collect.

“It was bonkers,” said Kathryn Fort, who represents the plaintiffs, including the Cherokee Nation, in a lawsuit challenging the government’s decision not to require the reporting of data on foster and adopted children’s gender identities and tribal citizenship. Without that information, the government doesn’t have an accurate picture of these marginalized populations in foster care, Fort says, which means children are losing out on accessing services.

The 2020 rule is final, but the debate continues over how much data will be required and how to make sure states report it.

Some say the child welfare system is already overloaded.

“We demand a lot from social service providers and caseworkers,” said Linda-Jeanne Mack, who spent six years as the director of a Massachusetts program intended to ensure foster children find permanent homes.

“You have a bunch of people who know absolutely nothing about the data compiling the data,” Mack said. “None of us were data scientists.”

Mark Testa, the former UNC professor, said he’s not sure what will work going forward.

“How do we get states to improve their data collection?” Testa said. “I think we have to continue to say why it’s good for kids that we do this, why it’s good for families, why it's good to correct the inequities in our system.”

Contributing: Mark Nichols and Matt Wynn.

.