Manatee County is holding a town hall on 8/30 to discuss the drastic expansion of its out-of-home care population. According to news reports, the expansion has been driven largely by an ongoing heroin epidemic. I thought it would be useful to share what the Dashboard says about the situation. The spike may have more to do with changes in policy than with the rise in drug usage among residents.
First, the big picture: Manatee County has definitely hit historically high out-of-home-care numbers (in blue above). Its in-home care numbers (in green above) are high but not extreme. Manatee County may already be past its peak, as there have been multiple months of steady or decreasing numbers. It could rebound again, but a closer look at the fundamentals below suggest that this is not likely.
Investigations are a little up, verifications are fairly steady, and removals are way up.
Looking at the green line in this chart, it appears that the upswing in removals started sometime around September 2013, which has the lowest removal count on record (only 4) in this area. The sharp upturn the next month suggests a clear policy change occured. The peak appears to have occurred in October 2015, when 74 children were removed.
Notably, the number of screened-in investigations are up by about 33% from 301 in September 2013 to 366 in September 2015. Verifications, however, are trending very slightly downward from 19.93% in September 2013 to 17.76% in September 2015. This rise in screened-in investigations is seen across the state and could be a mix of policy decisions by the Statewide Abuse Hotline or a reflection of population growth. What is clear from these numbers of that the spike in removals is not a result of a drastic spike in the number of investigations. It’s instead likely a change in either the types of cases that are called in or a change in the State’s response to how risk is assessed. The sharp increase suggests a policy change.
Changes in OOHC is the simple difference of removals and discharges each month. Discharges and removals are both highly seasonal, with removals spiking at the ends of the school vacation months and discharges spiking in Novembers (adoption days) and June (end of a fiscal year). In the chart to the right, you can see Manatee County’s removals versus discharges, and the recent expansion is clearly a result in heightened removals, not a reduction in discharges.
Substance abuse allegations are down. Failure to protect allegations are up.
The explanation offered in the media by child welfare professionals is that the number of removals is a reflection of an on-going heroin and opiate problem in Manatee County. This would suggest an increase in substance abuse cases. DCF trend reports show that “Substance Misuse” as an allegation is actually on the decline through this period. The 3-month average number of allegations of substance misuse have gone up from 170.5 in September 2013 to 200.8 in July 2014, and are now at 143.5. This is not the surge I was expecting and is consistent with a statewide decrease in the number of cases marked as “substance misuse.” The verification rate for these cases, however, has gone up — from 10.91% in December 2013 to 23.33% in December 2015. Either cases are becoming more severe, or DCF is categorizing and assessing risk differently in these cases. The abrupt increase, again, suggests a policy change.
In comparison, “Failure to Protect” allegations appear to better match the removal curve, and are closed out as “verified” at much higher percentages. This could be a reflection of DCF’s policies on categorizing alleged maltreatment. “Failure to protect” is not a useful category if it does not include failure to protect from what.
Relatives to the rescue.
Since September 2006, relatives have been the predominate placement resource for kids in Manatee County. This is consistent with patterns across the state in areas facing similar expansions. Smaller expansions in Manatee County’s OOHC population from 2009 to 2013 were largely absorbed by relatives as well. Beginning in September 2013, however, all placement types began to expand, with relative placements expanding the most, and non-relatives and facility placements growing at slower rates. Family foster homes expanded until around August 2015, when they appear to have largely maxed out available placements at approximately 170 children; while all other placement types have continued to grow and show no current signs of reduction.
Relative placements may have peaked this summer and may be on the decline as cases close out and the removal rates come down. Family foster placements also have been on the decline as the numbers begin to drop. Notably, there is no similar drop in non-relative or facility foster numbers. This suggests that children in those placements experience different permanency paths from children in the homes of relatives or foster families.
The expansion clearly required a lot of reliance on relatives, but Manatee County and its CBC are not the state’s most heavily dependent on relative placements. In July 2016, Sarasota YMCA’s OOHC population was 42% relative caregivers. The areas marked red in the map to the right were all well over 50% relative placements as of July 2016. The CBCs with the lowest reliance on relatives have numbers below 40%, shown in deep green.
Discharges are lagging, but should be picking up soon.
As would be expected, the number of discharges has been rising with the increase in OOHC, but discharges as a percentage of OOHC has been on the decline. This indicates that the increase in OOHC has resulted in a slow-down of the normal permanency path of cases, perhaps due to lack of resources, lack of available pre-adoptive and guardianship placements, or just lack of room on the court docket to handle cases effectively. Expansions cause slowdowns across the board.
Seen below, reunification has overwhelmingly become the predominate discharge type, recently hitting 77% of all discharges. Statewide, reunifications tend to be approximately 50% of all discharges. Guardianships and aging out have remained constant, while adoptions have risen very slightly. These numbers go against statewide trends, where reunifications have been steady and guardianships significantly down in recent years. The large number of reunifications raises the question of whether these children could have been served in their homes as opposed to removed. For the children who are found to have been appropriately removed, credit is due for high reunification rates.
Disproportionality issues galore.
Drastic expansions are interesting opportunities to study disproportionality effects. As the graph to the left shows, the expansion saw a significant increase in the number of white children (orange line) in OOHC. Initially there was a slight increase in black children (blue line), but that appears to have reduced and then stabilized. The numbers of “other race”children (green line) appears to have remained steady to slightly rising.
Viewed as a percentage of OOHC, the initial expansion raised the relative population of black children, but that number quickly returned back to lower levels as the number of white children continued to increase through 2015. The current expected OOHC composition in Manatee County is consistent with its historical values: 72% white, 20% black, and 7% other race. Statewide, the composition is approximately 60% white, 30% black, and 8% other, and growing whiter.
Interestingly the expansion seems to have affected children ages 10-14 (yellow line) the most. Before the expansion, this population regularly comprised about 25% of the OOHC population, but is now projected to be approximately 28%. After an initial spike in children ages 0-4 (blue line) beginning in November 2013, the next spike is found in ages 10-14 (yellow line) beginning in March 2014. Interestingly, ages 5-9 (light blue line) have seen steady decrease over the years and now make up approximately only 8% of the OOHC population.
Finally, the number of children in OOHC for 12 or more months shows significant age effects. The 10-14 and 15-17 populations who have been in care longer than 12 months are expected to increase drastically over the next 16 months, while the numbers of children ages 0-4 and 5-9 are expected to remain fairly constant. These trends are reflective of the fact that older children face significant barriers to permanency once removed from their homes. The effects of this expansion will be felt for years to come as children in these cohorts eventually age out.
Is this just a Manatee County thing?
Curiously, this same expansion pattern is found in other counties around the state, but by no means in all of them. The following circuits and counties saw spikes; the dates in parentheses are eyeballed estimates of when the spikes began:
- Twentieth Circuit – Charlotte (August 2014)
- Seventh Circuit – Flagler (September 2014), Putnam (January 2014), St. John’s (July 2014), Volusia (July 2014)
- Third Circuit – Columbia (September 2014), Dixie (September 2014)
- Fifth Circuit – Hernando (February 2015), Marion (February 2015), Sumter (March 2015)
- Eighteenth Circuit (September 2014) – Brevard (September 2014)
As an example, here is the Putnam County OOHC/IHC chart:
The Putnam County spike is by far the most pronounced. Putnam County’s OOHC population went from 62 to 248 children in just two years. Its IHC population, however, did not rise by equal amounts. Putnam went from 27 children placed with relatives to 161 during this time. Putnam also experienced similar demographic waves: first a surge of black children, then an uptick of white children — first a surge of infants, then 10-14 year olds. Putnam County has dealt with discharges differently from Manatee, relying more on adoptions and permanent guardianships, but in many months the county has few if any discharges at all. Based on my google news searches, though, Putnam did not experience any similar drug problems. This spike is something else.
Is this a drug epidemic or a policy change? (Yes)
What happened in Florida Child Welfare policy in September 2014 that so many counties experienced extreme spikes in their OOHC populations all at once? SB 1666 happened, a new Secretary had been recently appointed, and the debate about how to handle abuse investigations continued. SB 1666 formalized and, in some areas, greatly reduced the Department’s ability to rely on safety plans in lieu of formal court intervention. The expansion in removals into relative placements in lockstep with the change in law suggests that these are cases that would have been resolved with a safety plan — “grandmother agrees to supervise contact between mother and child” instead of court involvement: “grandmother is ordered to supervise contact…”. Manatee County may have had the unfortunate misfortune to experience a heroin epidemic at the same time as their discretion to work with families informally was reduced. This is what the pendulum of child welfare looks like.
This look at the numbers does not solve these counties’ placement problem, but it does suggest one important question: do all of those children really need to be removed and court involved?