Our full response to Innocents Lost


A Response to the Miami Herald from
the National Coalition for Child Protection Reform
June, 2014


In March, 2014, the Miami Herald published a series of articles called “Innocents Lost.”  The central claim of the stories is this:

The children were not just casualties of bad parenting, but of a deliberate shift in Florida child welfare policy. DCF leaders made a decision, nearly 10 years ago, to reduce by as much as half the number of children taken into state care, adopting a philosophy known as family preservation. They also, simultaneously, slashed services, monitoring and protections for the increased number of children left with their violent, neglectful, mentally ill or drug-addicted parents.

The result: Many more children died.

This claim is almost entirely false.  Indeed, it is a repeat of The Big Lie of child welfare -- that family preservation and child safety are at odds and doing more to keep families together makes children less safe.

To bolster this error, the lead reporter for the series, Carol Marbin Miller, and her colleagues have  distorted data, taken information out of context, gotten time frames wrong, and systematically left out facts that contradict their point of view.  We document all of this, in detail, in the critique that follows.

Miller didn’t do this out of a desire to sensationalize. She didn’t do it to “sell newspapers” (or generate pageviews).  We know Miller well and we know she sincerely wants to make children safer.  But that does not make her failings any less serious.

If Miller’s claims were true the series would have been a huge success.  In the first full month since the series appeared the number of children torn from their families in Florida skyrocketed by 40 percent.[1]  No doubt the Herald will claim this makes children safer.

But it does nothing of the kind.


The problem with a rush to needless foster care is not that it hurts parents, though of course it does.  The problem is that it hurts children.

It hurts children because removal from their homes is so inherently traumatic that two major studies of 15,000 typical cases found that even maltreated children left in their own homes with little or no help fared better, on average, than comparably-maltreated children placed in foster care. 

The Herald, and some readers of this rebuttal, might respond: That’s too bad, but what else can we do to save children from the kinds of horrors described in the Herald stories?  But the horrors described in the Herald stories involve only a tiny fraction of the cases seen by workers for agencies like the Florida Department of Children and Families (DCF) and its counterparts across the country.  No one disputes that children in these situations should be removed from their homes. 

Where the Herald errs grievously is in claiming the reason these children were not removed was because of some overarching philosophy of family preservation.  The real reason, discussed below, is very different.

Typical cases are nothing like the horror stories described in the Herald series.  Far more often, they involve workers confusing poverty with neglect.  (See these NCCPR Issue Papers for details: http://bit.ly/1nwOFn8http://bit.ly/1tjMDqg.)  That very fact helps explain those study findings – the ones that conclude that, in typical cases, family preservation is a far better option for children.

All that harm of foster care can occur even when the foster home is a good one, as a majority are.  But study after study has found abuse in one-quarter to one-third of foster homes, and the record of group homes and institutions is worse.

Most of all, the massive needless removal of children to foster care hurts children by so overloading caseworkers that they have less time to find children in real danger.  That is the actualreason for the failures documented in the Herald stories.

All over the country, these kinds of foster-care panics – huge surges in child removals in the wake of reporting like that of the Herald – have been followed by increases in child abuse deaths.  And contrary to the Herald’s claim, that also has been the case in Florida

Indeed, what is happening now is tragedy repeating itself; a replay of events that took place in 1999 – and an undermining of what had been one of the nation’s most successful efforts to reform child welfare and make children safer.


A decade ago, the state of Florida was the prime national example of failure in child welfare.  News organizations across the state flocked to Florida after it was revealed that a foster child had been missing for 14 months – and was presumed dead – before DCF even noticed.

In fact, NCCPR had predicted the collapse of the Florida system in 1999, shortly after Kathleen Kearney, a former Broward County judge, was named to run DCF.

Kearney’s approach to child welfare could be boiled down to a single sentence: Take the child and run.  During her first year as DCF Secretary, the number of children torn from their homes soared by 50 percent, the worst statewide foster-care panic we’ve ever seen – though the one underway now comes close.  And entries into foster care stayed at this obscenely high level for seven years.

When the data are examined in full, not selectively as the Herald has done, it is clear that as entries into foster care soared, child abuse deaths increased – as we noted above, this is a pattern seen across the country.  This should come as no surprise.  The real reason workers sometimes leave children in dangerous homes is that those workers are too overloaded to investigate any case properly.  A foster-care panic only overloads them some more.

NCCPR issued report after report on Florida’s failure.  And ultimately those reports, and other factors had an impact.

A new governor, Charlie Crist, brought in new leadership.  First Bob Butterworth and then George Sheldon reversed course and embraced safe, proven approaches to keeping families together.  Independent outside evaluations  found that the reforms improved child safety.  The New York Times did an in-depth story on the state’s remarkable turnaround.

Child safety improved and, again, contrary to recent claims by the Herald, child abuse deaths decreased.

If anything, there is a need for more such reform.  Even in 2009 and 2010, when entries into care were at their lowest, Florida’s statewide rate of removal still was 10 percent above the national average, and significantly above the rate in states that take, proportionately, far fewer children – including the very state that the Herald now singles out as a model.  So the notion that some kind of pendulum swung too far toward preserving families is a myth.

But ever since Butterworth and Sheldon changed course, opponents wedded to Kearney’s discredited approach have tried to undermine the reforms.  They’ve found an eager ally in Miller, the longtime reporter on the child welfare beat for The Miami Herald.  Miller is a skilled and tenacious journalist.  But somewhere along the line Miller went from reporter to advocate.  She decided that she knows so much about Florida child welfare that she has the right to draw the conclusions herself, instead of giving readers all the information they need to draw their own conclusions.

That’s why in 2011 we created a website, www.heraldvsfacts.blogspot.com in an effort to set the record straight.

“Innocents Lost” repeats the same failings of Miller’s earlier reporting:

● Data are taken out of context.  Data that would contradict Miller’s thesis are omitted by choosing only start and end points that support Miller’s claims.

● Similarly, Miller misstates the time frames concerning when Florida DCF embraced a take-the-child-and-run approach to child welfare, and when it was doing more to keep families together.  She draws conclusions based on what DCF officials said, rather than the data that reveal what they actually did.

● Miller and her colleagues did their own examination of deaths during the years that, she claims, DCF was emphasizing family preservation.  They conclude there are more deaths than shown in official statistics.  But they do no such analysis for the years in which the take-the-child-and-run approach dominated the system, making an apples-to-apples comparison impossible.

● Miller attributes a drop in deaths of children “known to the system” (an inconvenient fact that contradicts her entire thesis) to a narrowing of the definition of “neglect” – but never mentions that, earlier, the definition had been vastly broadened.  The change Miller cites simply returned the definition to a middle ground.

● The stories rightly point to Alabama as a model of doing child welfare right – but chose not to mention that Alabama’s success is a result of doing far more, not less, to keep families together.  Indeed, Alabama takes away children at one of the lowest rates in the nation, a rate 20 to 40 percent lower than Florida, depending on how the data are calculated.  (NCCPR can provide full details on rates of removal in different states to anyone interested.)

● Florida has a commendably broad open records law.  But instead of using this law to examine a statistically-valid random sample of cases, in order to see how and how often the system errs – in all directions – the Herald looked only at deaths of children previously known to DCF.  In story after story, the Herald isolates a few such cases and generalizes, with no evidence that the problem cited permeates the system.  No one would laud an article on air travel safety that looked only at crashes, and failed to gauge the overall record of safe flights.  Just as air travel is the safest form of transportation, family preservation is the safest intervention for the overwhelming majority of children, the overwhelming majority of the time.

[1] To view the most recent entry into care data, click on this link on the Website of Florida’s Center for Child Welfare: http://centerforchildwelfare.fmhi.usf.edu/Datareports/TrendReports.shtml Then click on Child Welfare Services Trend Report.