Monday, October 18, 2010

UPDATED: Foster care in Los Angeles: LA Times gives Sam Zell just what he wants


At the end of this post, readers will find links to the superb database maintained by the Center for Social Services Research at the University of California at Berkeley.  Using that database, readers can check for themselves the long-term data cited in this post.

Earlier this month, there was a lot of buzz in media circles about a 4000-word piece by New York Times media critic David Carr about the culture of sleaze brought to the Tribune Company, owner of the Los Angeles Times, by Sam Zell, the man who bought the company – and drove it into bankruptcy. 

There are several ways a reporter can respond to a new boss like this.  One is to quit – but that’s extremely difficult in the current job market for journalists.  Another is to raise hell until you’re fired – but that leads to the same problem as option #1.  A third way is to just keep your head down and do the best work you can in spite of the climate.

Or you can just give the new guy what he wants.

Anyone who read the New York Times story knows what kind of journalism Sam Zell wants.  And apparently, Los Angeles Times reporter Garrett Therolf is determined to give it to him - no matter how many children may be endangered in the process.  Or, at least, Therolf’s latest story about child welfare “raises questions” about whether that is his goal.

To sum up the story: Therolf reports that, since 2008, deaths of children “known to the system” in Los Angeles County have increased.  Then County Supervisor Zev Yaroslavsky, who has been getting child welfare wrong for decades, obligingly says exactly what Therolf wants: He scapegoats the county’s efforts to keep families together.

But to the extent that the data in the story are evidence of anything, it is that Therolf's "reporting," and Yaroslavsky's demagoguery, may be endangering children. 

Ever since Therolf began his crusade to tear apart more families, we've warned that foster-care panics, sudden surges in removals of children by workers terrified of being on the front page, are followed by increases in child abuse deaths.  That's exactly what has happened in Los Angeles County. 


And that's almost certainly why Therolf's story mentions only the number of children in foster care on any given day (the "snapshot number")  - neglecting to mention that the number of children taken away from their parents over the course of a year increased in 2009 - thanks to the foster-care panic.    The number almost certainly is increasing again in 2010.   In other words, the increase in removals was accompanied by an increase in deaths of children known-to-the-system.

After I posted an early, abbreviated version of this Blog in the comments section below an earlier version of his story, Therolf e-mailed to say he’d posted a response in which he says:

Mr. Wexler is wrong to suggest that The Times has suppressed any news of a "foster care panic." The reality is that the detention numbers have seen no sustained rise since the newspaper began its series of reporting on child fatalities in the system. Readers can access the detention figures and judge for themselves here:  

I certainly hope Times readers call Therolf’s bluff.  Because what they’ll find are numbers almost identical to those in NCCPR’s own report that first revealed the foster-care panic.

As I pointed out in an e-mail to Therolf:

There always are fluctuations from month to month.  That's why you determine a panic by comparing a month to the same month the previous year.  We predicted the panic would begin in September 2009, and when you compare September 2009 to September 2008 there is indeed a spike.  Same with October, November, and December, comparing each year to the year before.  It stopped at the start of 2010 and started up again in April and May. [The most recent months for which data are available.]

            I explained this to you last year.  You chose to ignore it, and now, because it doesn't fit your master narrative, you decline to report it. 

            In fact the panic actually began even sooner, in August 2009, a response to stories that ran in the Times in July. In addition, the data NCCPR posted, which go back a year farther than Therolf’s link, show that, with one exception, in each of those months in 2009, the number of entries was the highest since 2006 – and the five-month total not only was a large increase in itself, it also reversed two years of small declines over those same months.

Indeed, if Garrett Therolf really wanted Times readers to “judge for themselves” he would have put these figures into the newspaper, with perspective from all sides, rather than only mentioning them now in the comments section of the Times website.

           It is true that the panic probably would be even worse if not for a waiver from federal funding restrictions – a waiver Therolf repeatedly has smeared, most recently in this latest story, with no meaningful response allowed from anyone who disagrees with Therolf (More on this below.)

            And if Therolf really wanted readers to “judge for themselves,” he’d report on how the rate of removal in Los Angeles County compares with other large metropolitan areas – L.A.’s is among the highest, more than double the rate of Miami-Dade County Florida and more than four times the rate of metropolitan Chicago – both places where independent evaluations have found that successful efforts to curb needless foster care improved child safety.

            But that’s far from the only failing in this story:

Even before the panic, the number of children torn from their families in L.A. County has increased every year but one since 2004.  (That one year was 2008, which, according to Therolf's own story, was the low point in deaths of children known-to-the-system).  And, as the story itself makes clear, the time from 2008 through 2010 is the only time period in which it's possible to do an accurate comparison, because the same reviewers appear to be applying the same standards to all cases.  Data for previous years probably can't be compared reliably to the more recent data - or to anything else.

None of this is hindsight.  Based on what we've seen when similar lousy journalism set off similar panics all across the country, we warned more than a year ago, in this Times op ed column that

It's no wonder some lawyers say they're already seeing a foster care panic -- a sudden surge in removals of children from their homes. That only further overloads the system, making it even less likely that the next child in real danger will be found. That's why, across the country, such panics have been followed by increases in deaths of children "known to the system."

So if the harm to all the children needlessly taken isn't enough, the fact that foster-care panics make all children less safe ought to be reason enough for Therolf to report the full story, instead of just the parts that make the kind of headlines someone like Sam Zell will love.  And the fact that foster-care panics make all children less safe ought to be reason enough for Yaroslavsky to stop pouring gasoline on the fire, and start looking for ways to do more, not less, to safely keep families together.


 Yes, the number of children in foster care on any given day still is declining, but the decline is very slow, and has been quite slow for some time.  The really big plunge in foster care numbers took place all the way back in the time from 1999 through 2003, and Therolf's story says nothing about child abuse fatalities then.  Nor does it tell us anything about the number of deaths of children "known to the system" back in 1997 and 1998, when the number of children in foster care was at its height.


What all of this really shows is that fatalities simply can't be used to judge the overall performance of a child welfare system in assessing child safety.  The data are too unreliable, and too subjective.  (For example: If a toddler previously known to DCFS wakes up early one morning, unlatches the back door, wanders outside, falls into a body of water and drowns, is it neglect or an accident?) [UPDATE, OCT. 19: Indeed, as the competing Los Angeles Daily News reports, one reason the new figures are higher than the old figures is that the old figures didn't include drownings and the new ones do.]

All of this is explained a report cited often on this Blog, from a Texas group, the Center for Public Policy Priorities - a group that formerly was fanatical about child removal, but now seems to be changing its mind - as a result of the evidence they developed.  The report also concludes that none of the traditional investigative functions of agencies like DCFS has an impact on child abuse fatalities - but you can reduce them with more efforts at prevention, and more efforts to curb poverty and teen pregnancy.


There is a far better, though still imperfect, safety measure - the one the federal government uses: overall reabuse of children left in their own homes.  That has been steadily improving in Los Angeles County for a decade, all the way through September 2009, the most recent data available, a period that ends just as the panic was beginning.   This measure improved when entries into care went way down, in the early part of the decade, and it's continued to improve when entries went up, then down, then up again.

Since those data don't prove Garrett Therolf's point, Garrett Therolf declines to share them with readers.

Instead, he includes one brief quote from a professor evaluating the waiver who said reabuse increased since the waiver began.  But he used a different time frame from the standard used by the federal government.  In addition, the Berkeley database shows a much smaller increase, even when using this time frame.  And using either time frame, the highest rate of reabuse, by far, was in 1998 – the same year both the snapshot number and entries into foster care were at record highs.  So if the waiver is supposedly causing an increase in reabuse now, what caused an even higher rate of reabuse when there was no waiver, and vastly more children in foster care? Again, since this context spoils Therolf’s master narrative, he doesn’t share it with readers.

But suppose, hypothetically, real measures of safety had worsened in the last two years.  Instead of asking why and what can be done about it, Therolf, speaking through Yaroslavsky, simply offers his own "solution" - take away more children - instead of asking why other places with the same waiver, like Florida, have improved safety, and asking how Los Angeles County can do the same.

You can see all the long-term data for yourself here:

All of these links go to the Berkeley database mentioned at the top of this post.  In each case, after clicking on the link, follow the instructions to create a table involving multiple time periods and choose Los Angeles County.  (The pre-selected box under “interval” will give you data that end with the most recent full year available.)  In the case of “entries into care” we checked the box for entries of any duration, rather than the default, which counts only entries longer than a week.  For reabuse, we chose to the measure within six months because that is the standard time frame used by the federal government:

UPDATE, OCT. 19: Be sure to check out the story in the competing Los Angeles Daily News for some of the crucial context the Times left out.  For example, The Daily News reports that part of the reason for the higher numbers is a broader definition of child maltreatment deaths that includes drowning.  That isn’t wrong in itself, but drowning is among the hardest causes of death to categorize objectively, as indicated in the “accident or neglect?” example above.

Also, once again, Deanne Tilton Durfee is refusing to jump on the Times’ bandwagon. Durfee heads L.A.'s Interagency Council on Abuse and Neglect, serves on the county's child abuse death review team, heads the National Center for Child Fatality Review – and is no friend of family preservation. But she’s still not buying the snake oil the Times and Yaroslavsky are trying to sell.