In order to understand patterns of reunification from foster care, it is useful to consider reunification in the context of all exits from care, and in the context of the individual characteristics and the case histories of the children involved.2 Table IIA.1 presents an overview of exits from foster care. The left side of the table contains counts of spells by type of exit, classified first by state, then by four child characteristics, and finally by each of four case characteristics. This series of two-way tables is based on a universe of 404,416 distinct spells in foster care. These are all spells that began with a foster placement between 1990-94 in any of the nine states, and are enumerated in the first column, labeled "All Entries. These spells are each tracked until the child is discharged from foster care or through the end of 1997. So the potential period of observation is at least three years (for children who entered care at the end of 1994), and at most eight years (for children who entered care at the start of 1990.
The second column, labeled "No Exit, Still in Care," indicates how many of these spells were still open at the end of 1997. Just under one-fifth (75,199 or 18.6%) of the children who entered care between 1990 and 1994 remained in care through the end of observation. It is important, to note that exits from some long-term stays in care are not observed and are not reported in this table, so that the completed duration and the eventual discharge destination at exit for these spells are censored from this analysis. However, we do know that all of these censored spells, by definition, have a duration of three years or more, and as will be seen, reunification is a fairly rare outcome after three years. Therefore, we assume that these censored spells would probably add little new information to most of this reunification analysis. However, it should be noted that exits from stays of over three years are incompletely described in these tables. The proportion still in care is computed against a base of all entries.
The exit distribution is presented with six categories for destination at exit:
- Reunification is defined as a return to the parent or family of origin.
- Relative care is full custodial placement with a relative outside of the foster care system. It is not formal kinship foster care, which involves state custody and control.
- Adoption is finalized legal adoption of the child, either by relatives or by unrelated persons. Pre-adoptive placements are considered as foster care.
- Reach majority occurs when a child ages out of foster care.
- Runaway occurs when the child leaves placement for a significant period of time without agency authorization.
- "Other exit" is a catch all for a variety of categories, including "unknown," "service completed," death, and transfer to other agency.
Looking at the first row of Table IIA. I , we see that of the 404,416 spells observed, 187,406 ended with a family reunification. There were also 23,398 exits to other relatives, 38,291 completed adoptions, 9,583 children "aging out" of care, 23,635 runaways, and 46,904 spells that exited to "other" destinations. On the right side of the table, each exit type is presented as a proportion of all observed exits. Therefore, the statistic reporting that the proportion of reunification spells for all nine states combined is .569 should be interpreted as meaning that 56.9 percent of all spells where an exit was observed were exits by reunification.
During the period of observation, the majority of exits (57%) were achieved through family reunification. Of the remainder, 7 percent were to relatives, 12 percent via adoption, 3 percent reached majority, 7 percent involved runaways from care, and 14 percent were in the "other" category. This exit distribution for the spells from the pooled nine-state data provides a yardstick against which the distributions for each state or subgroup will be compared.
It is important to note that the distributions presented in Table IIA. I describe each type of exit as a proportion of all observed exits, not as a proportion of all entries. If the 9-state reunification statistic (.569) were recomputed as a proportion of all spells, we would see that 46.3 percent of all entry spells were observed to end in reunification by the end of 1997. This latter statistic is usually what would be considered as a reunification rate, in that it describes the likelihood of reunification for an entire population at risk. The difference between the two measures is caused by spells where an exit is not observed. The relationship of these statistics is presented in Figure A. Here, we can see that the reunification rate (reunifications/entries) can be described as the product of a) the likelihood of exit (exits/entries), and b) the likelihood that an exit is by reunification (reunifications/exits). Alabama has one of the highest exit rates, so the low reunification rate must be interpreted in the context of knowing that many children exit foster care in other ways. In contrast, the low reunification rate in Illinois must be understood in the context of low overall exit rates -- as relatively few children leave foster care to other destinations.3
|Counts||Proportion of Entries||Proportion of Exits|
|All Entries||No Exit||Exit Type||No Exit||Exit Type|
|Still in Care||Reunify||Relative Care||Adopt||Reach Majority||Run Away||Other Exit||Still in Care||Reunify||Relative Care||Adopt||Reach Majority||Run Away||Other Exit|
|Nine State Total||404,416||75,199||187,406||23,398||38,291||9,583||23,635||46,904||0.186||0.569||0.071||0.116||0.029||0.072||0.142|
Age at Entry
|< 3 mos.||51,521||13,215||15,095||2,772||15,492||0||0||4,947||0.256||0.394||0.072||0.404||0.000||0.000||0.129|
|1 to 2 yrs.||57,188||13,032||28,155||3,773||6,835||0||0||5,393||0.228||0.638||0.085||0.155||0.000||0.000||0.122|
|3 to 5||61,649||15,148||31,406||3,897||5,525||0||9||5,664||0.246||0.675||0.084||0.119||0.000||0.000||0.122|
|6 to 8||46,372||11,556||24,199||2,924||3,036||0||209||4,448||0.249||0.695||0.084||0.087||0.000||0.006||0.128|
|9 to 11||40,943||8,822||21,637||2,545||1,514||165||1,817||4,443||0.215||0.674||0.079||0.047||0.005||0.057||0.138|
|12 to 14||62,742||5,498||31,067||3,315||601||2,294||10,611||9,356||0.088||0.543||0.058||0.010||0.040||0.185||0.163|
|15 to 17||52,973||1,093||22,114||2,112||127||7,124||10,989||9,414||0.021||0.426||0.041||0.002||0.137||0.212||0.181|
Age at Exit
|1 to 2||47,325||0||29,317||4,090||8,375||0||0||5,543||0.000||0.619||0.086||0.177||0.000||0.000||0.117|
|3 to 5||80,501||17,848||35,440||4,578||15,529||0||0||7,106||0.222||0.566||0.073||0.248||0.000||0.000||0.113|
|6 to 8||62,543||20,868||26,189||3,160||7,041||0||0||5,285||0.334||0.628||0.076||0.169||0.000||0.000||0.127|
|9 to 11||46,650||13,449||22,288||2,592||3,484||0||359||4,478||0.288||0.671||0.078||0.105||0.000||0.011||0.135|
|12 to 14||58,176||10,620||27,935||3,101||1,883||59||7,220||7,358||0.183||0.587||0.065||0.040||0.001||0.152||0.155|
|15 to 17||69,444||8,747||30,330||3,016||682||1,960||14,573||10,136||0.126||0.500||0.050||0.011||0.032||0.240||0.167|
|Over 17 yrs||19,290||3,667||2,094||303||80||7,564||1,483||4,099||0.190||0.134||0.019||0.005||0.484||0.095||0.262|
|< 1 mo||49,767||0||34,820||5,027||271||90||4,048||5,511||0.000||0.700||0.101||0.005||0.002||0.081||0.111|
|1 to 3 months||46,432||0||30,905||4,336||510||216||4,297||6,168||0.000||0.666||0.093||0.011||0.005||0.093||0.133|
|4 to 6 months||35,492||0||23,517||2,862||459||334||3,309||5,011||0.000||0.663||0.081||0.013||0.009||0.093||0.141|
|6 mos to 1 year||48,657||0||33,358||3,197||1,706||872||3,661||5,863||0.000||0.686||0.066||0.035||0.018||0.075||0.120|
|1 year to 18 mos||36,557||0||24,329||2,210||2,170||1,041||2,165||4,642||0.000||0.666||0.060||0.059||0.028||0.059||0.127|
|18 mos to 2 year||26,752||0||15,077||1,655||3,141||1,097||1,529||4,253||0.000||0.564||0.062||0.117||0.041||0.057||0.159|
|2 to 3 year||36,103||0||14,327||2,040||9,777||1,948||2,098||5,913||0.000||0.397||0.057||0.271||0.054||0.058||0.164|
|3 years +||124,656||75,199||11,073||2,071||20,257||3,985||2,528||9,543||0.603||0.224||0.042||0.410||0.081||0.051||0.193|
Likelihood of Reunification from Foster Care Components of Reunification Rates
The proportions in Figure A.1 report the likelihood that a child or youth entering foster care between 1990 and 1994 exited before 1998 via reunification with their family.
One decomposition of the proportions in Figure A.1 is represented in Figure A.2, which presents the likelihood of any exit from care and the proportion of exits that are achieved via reunification. Mathematically, the likelihood of reunification is the product of these two separate proportions.
2. Ideally, we would also consider attributes of the family of origin, the parents (or other home caregivers), and the home environment; but this type of information is not generally available through child welfare tracking systems.
3. The analysis in later sections will focus almost exclusively on reunification rates (the likelihood of reunification among entrants). The focus on exit distribution in this section is to highlight the differences between children who are reunified and children who exit in other ways.