Monday, November 15, 2010

Foster care in Los Angeles: What Garrett Therolf won’t tell you about the LA child welfare waiver


Read the full evaluation here.

            For much of the time since Los Angeles Times reporter Garrett Therolf launched his crusade to get more children taken from their families he’s been targeting a waiver from the usual restrictions on the largest category of federal child welfare funds, Title IV-E.

            In most of the country, such funds are an open-ended entitlement, covering a large share of the cost of foster care for every eligible child.  No matter how many eligible children a state or county tears from her or his parents, the federal government will keep right on picking up a big part of the tab. 

            This does not mean, as some have charged, that state or local governments “make money on foster care” – since those governments still have to pay some of the costs.  (Private agencies, paid for every day they hold a child in their group homes or in foster homes they oversee do make money on foster care, but that’s a separate set of incentives. Details on all of this are in our publication on child welfare finance.)

           The open-ended entitlement does mean, however, that while safe, proven alternatives to foster care cost less in total dollars, foster care sometimes can be a money-saver for states and counties.  And all that state and local money wasted paying the portion of foster care costs not paid by the feds can’t be used for better options.

          Here’s how an independent evaluation of the Los Angeles waiver puts it:

         Absent a source of funding for community-based services that support families, child welfare departments are left with little choice but to remove children from questionable environments and place them in foster care. When restrictive funding drives what programs and services are offered by the State and counties, it also defines the philosophy and mission of child welfare in California.

         But Garrett Therolf has never told his Times readers about the profound incentive for needless foster care that existed before the waiver – since Garrett Therolf has no problem with that.  Instead, Therolf smears the waiver every chance he gets.

        Under the waiver, Los Angeles County agreed to take its Title IV-E money as a flat grant, based on the amount the county had received in previous years.  The county was free to use the money on better alternatives to foster care.  If it reduced foster care numbers, the county could keep the savings, as long as that money was plowed right back into child welfare services.  But if foster care numbers go up, the county has to pay the full cost of the increase.

        Early on, Therolf’s stories repeatedly described reunifying families as “risky” and the waiver as a “wager” on reunification.  There is, of course, no mention of the profound risks caused by needless foster care before the waiver.  But he sank to a new low in October when he took one tiny piece of the waiver evaluation out of context.  Therolf wrote:

Others who have studied the waiver program say that earlier indicators were already suggesting problems.

         As is discussed in detail below they actually said no such thing.   But for now, back to Therolf:

Child safety indicators under the waiver have become "an area of concern," said Charlie Ferguson, a San Jose State University professor and the state's independent evaluator of the waiver program.
Ferguson said officials had studied data involving children who had not been removed from their homes after an allegation of child abuse had been substantiated. They found that those children experienced an increased rate of substantiated abuse within a year. The rate has increased by 19% since the Title IV-E waiver began in 2007, he said.

          Therolf repeated the claim in a story on November 12:

Since 2007, children found by the department to be victims of abuse who were left in their homes have increasingly experienced abuse again within a year, according to a researcher hired by the state. The increase was 19%.

         He repeated it again, word-for-word, the following day in a story about the likely imminent replacement of DCFS Director Trish Ploehn.


        What Therolf neglects to mention is that the evaluation also states, on page 2, that there is no evidence that the waiver, which the evaluation refers to as the Capped Allocation Plan (CAP), is responsible for this change.  According to the evaluation:

[C]hanges observed in the participating counties and in outcomes for children and families cannot be directly attributed to the CAP.

        Of course, it’s a pretty long document – maybe Garrett just overlooked it.  Except that the evaluation also made the same point on Page 137:

The design of the evaluation makes it impossible to attribute changes in outcome trend lines to a department’s participation in the CAP.

And, in more detail on page 11:

[T]he overall design [of the evaluation] does not allow the evaluation to determine causality. In other words, it cannot be said that changes observed in the patterns of outcomes of the participating counties post-CAP as compared to their pre-CAP pattern of outcomes were caused by the CAP.

But just because scientifically “it cannot be said” doesn’t mean Garrett Therolf won’t say it anyway.

And Therolf’s claim that “others who have studied the waiver program say that earlier indicators were already suggesting problems” with the waiver simply is untrue.  Those studying the waiver explicitly state that their evaluation can’t determine the reasons for any problems that may exist.

Oh, and one more thing: More recent and more comprehensive data, from the Center for Social Services Research at the University of California at Berkeley, show that recurrence of maltreatment within 12 months actually is down very slightly since the waiver began.  (To see these data, click here and follow the instructions for creating a table.)  Furthermore, as can be seen in the table below, reabuse within 12 months has been cut by more than 20 percent when compared to 1998, the earliest year for which data are available and a time when Los Angeles County was taking away more children and holding vastly more in foster care.

1998 (Earliest data available) Year ending June 30, 2008 (first full year of waiver) Year ending March 31, 2009
Entries into care 11,977 10,650 10,427
Total # of children in foster care 50,125 (1/1/98) 24,654 (7/1/07) 20,344 (4/1/09)
Reabuse of children known to DCFS w/in 12 months. (%) 12.4 (Jan. to June, 1998) 9.8 (July to Dec. 2007) 9.7 (Six months ending March 31, 2009 – most recent available)
Source: University of California and Berkeley Center for Social Services Research,

The waiver evaluation uses a more limited time frame – their data go back only to July 2002 and end at the end of 2008, and they use a slightly different methodology.  But even taking their data at face value, the evaluation points out that reabuse within 12 months remains below the level they found in 2002, when DCFS had far more children in foster care.

So if the waiver supposedly caused reabuse to increase, why was the rate of reabuse higher before the waiver, when the county was doing exactly what Therolf wants?

The waiver evaluation does say that the safety data “present a concern” on which DCFS needs to focus – just as Prof. Ferguson told Therolf.  Obviously that’s true.  But the fact that DCFS needs to be concerned about these data and needs to do something about the rate of reabuse, whatever that rate may be, is vastly different from claiming that the waiver is responsible for what those data show.

But Therolf’s smear campaign doesn’t stop there.  In both his recent stories, the boilerplate about the waiver is followed immediately by the claim that child abuse deaths of children “known to the system” have increased since January 2008, implying that, somehow, this too is linked to the waiver.  In fact, the waiver evaluation does not use fatalities as a measure of overall system performance, indeed no scientific evaluation of any system does this, probably for the simple reason that labeling a cause of death as child maltreatment is highly subjective and, though each is the worst possible tragedy, in any one community, even one as huge as Los Angeles County, the numbers are, thank God, low enough to fluctuate due to random chance.

Indeed, the only pattern in deaths of children known to the system, one that emerges by looking at multiple localities over many years, is that they tend to increase in the wake of foster-care panics – like the one triggered in Los Angeles by Garrett Therolf’s shoddy reporting.


As for the waiver evaluation, here are a few other details Therolf leaves out.

●Another safety indicator used in the evaluation – and the one used by the federal government to evaluate all child welfare systems, remained constant through the end of 2008. 

­●A third indicator, foster-care recidivism; that is, the proportion of children sent home from foster care who enter the system again, also did not change during the first 18 months of the waiver.  If one can reduce foster care without a safety indicator worsening due to that reduction, that is an improvement, since you’re getting the same level of safety while subjecting fewer children to the inherent trauma – and high risk of abuse – in foster care itself.

This does not mean all is well with Los Angeles child welfare, or even with the waiver.  But the problems involve doing too little, not too much.

Los Angeles still takes away children at one of the highest rates among America’s largest metropolitan areas – a problem made worse by the foster-care panic.  That has overloaded workers, creating a crisis in the system.

The evaluation also points to other problems which I’ll discuss in a future post.  But what the evaluation makes clear is that when it comes to Los Angeles County child welfare, the solution to the problems of reform is more reform.

Tomorrow: Why Trish Ploehn must go (and who should replace her)