Limitations of the available data sources have severely hindered our capacity to describe and analyze the characteristics of children living in kinship care settings. As seen in Section I, large national samples such as that drawn for the Current Population Survey do provide detailed characteristics for individual children living in kinship arrangements. However, because kinship living rarely occurs in the general population, the resulting sample sizes for children in this category do not encourage detailed comparison much beyond univariate description at the national level. Although data from the 1990 decennial census are built on a much broader sampling base, the public-access tabulations produced from the census provide minimal detail. They classify the full range of child living arrangements only by two wide age categories, children 0-5 and 6-17 years of age.(16) Census data can be extracted for many different geographic places, and analyzed in the context of other characteristics associated with those places, but such ecological inference poses its own limitations and potential pitfalls.
As we have seen in the previous section, the Multistate Foster Care Data Archive provides some individual-level information on the children living in formal kinship and non-kinship foster care settings in four states. This information enabled us to separate the formal kinship foster care component from the larger "relative care" category of the census in four states, and to separate all non-relative foster care from the broader "unrelated child" category of the census. But, in order to consider the differences between children living in formal and informal kinship arrangements, it is also necessary to obtain information of a similar quality describing the population of children living in informal kinship care. Such data has not been located from existing sources.
In lieu of a comprehensive data source, we take an initial step towards the formal-informal comparison here by introducing information drawn from one special sub-population of children in Illinois, namely those who recently received public services in the form of AFDC grants or foster care. Over one-quarter of all children living in the care of relatives in Illinois in 1990 received AFDC grants paid to the kinship care household, and over one-eighth of the relative care children in the state were in formal kinship foster care. Child recipients of AFDC and foster children are clearly not representative of the kinship care population as a whole. But this group contains a substantial proportion of Illinois's children in kinship care settings and it can easily be construed as a most important segment of the kinship care population from the vantage point of informing public policy.
The primary reason for choosing this population for study is simply that data describing it is available. Individual case records for AFDC children living with kinship caregivers (parent absent) is available from the Child Multiservice Database at Chapin Hall. Facing an extreme paucity of available information in this area, the capacity to draw comparisons at all is a significant improvement.
Illinois AFDC Data
The information presented here is drawn from an archive containing full histories of all AFDC grant cases from 1990 through 1995. The data points analyzed here are based upon an extraction of records for all children who were active recipients in an AFDC grant case in any of six successive annual June cross-sections. The history of the grant unit itself can be traced fully between these June time-points, but the actual attachment of the individual to the grant during the unobserved time must be inferred. The use of annual cross-sectional pulls inevitably implies that some child participation in the AFDC program is missed entirely, such as when a grant is both opened and closed during the 10-month time period between any July and the following May. Given certain constraints, we believe that the group of cases developed for this analysis presents a relatively accurate picture of the Illinois child AFDC population. As will be seen in the following analysis, even though we are missing some of the rapid "on-and-off" movement that occurs with some welfare participants, there is a marked overall continuity and stability in the welfare histories of the children we observe.
For each AFDC-recipient child, we can identify their age, ethnicity, geographic area, gender, program participation (Medicaid, Food Stamps), and relationship to the official grantee. We can also identify the same information for the grantee, all recipients in the grant unit, and some other persons associated with the grant unit. Because individuals often move between grant units, and grant units often split apart or reform into new grant units, a careful unduplication of child records was pursued to insure that each child history is complete (to the extent possible), and that the same child does not appear multiple times in the data just because he/she received benefits under different grant units. For analysis, children were restricted to persons between the ages of 0 and 17. Even if grant support continued beyond the eighteenth birthday, 18-year-olds were excluded from this analysis, because the custodial nature of care is redefined when a child reaches the age of majority.
A "relation to the grantee" field was used to obtain a preliminary classification of children into "parent" cases, "relative" cases, and "other" cases. All relationships coded as parent, mother, father, step-parent, etc. were classed as parent cases, and those classified as grandparent, aunt/uncle, sibling, cousin, etc. as relative-only cases. After this initial classification, each relative case was screened specifically for the presence of a "credible mother" in the grant unit.(17) For the purposes of analyzing kinship care, we assumed that classification errors in the direction forcing a few "relative-only" cases into the "parent" population should have much smaller potential corrupting effect on our conclusions than would result from classifying a number of parent-present cases into the "relative-only" population. Not only are "relatives" are the focus of study, the pool of relative cases is also much smaller than the pool of parent cases.(18)
Analysis of Illinois Living Arrangement Groups
The child cases examined here appear in either the AFDC or Foster Care tracking data between 1990 and 1995. At any specific point in time, a child can be classified uniquely as being enrolled in one of four program statuses -- AFDC Parent Grant, AFDC Relative Grant, Kinship Foster Care, or Other Foster Care.(19) The populations of children tracked in these statuses in each successive June is presented in Table 4.1.
The two most important categories for this study are the AFDC/Relative and the Kinship foster care groups. The Kinship foster care group (Kin/FC) is a complete enumeration of the formal kinship care population of Illinois during the period of study. This population has been expanding rapidly. The number of children in kinship foster care increased from 8,150 in June 1990 to 27,054 in June 1995, a cumulative growth of over 230 percent across the 5-year period. Although this explosion of kinship foster care in Illinois is far more dramatic than the experience of most other states where these trends have been monitored, a general pattern of growth in kinship care is one of the dominant national trends observed in child welfare in the early 1990s.(20)
The AFDC/Relative group is a non-random subset of the informal kinship care population of Illinois, which might best be described as "semi-formal" because of their reliance on some public supports. The U.S. Census in 1990 estimated that 56,793 children in Illinois were living in the care of relatives with no parent present in the household. Subtracting the children who were in formal kinship foster care from this total leaves 48,643 children in informal kinship care. The AFDC/Relative group numbered 16,058 in June 1990, almost exactly one-third the size of this estimate of the state's informal kinship population. Unlike the formal kinship population, the size of the AFDC/Relative group has remained virtually unchanged between 1990 and 1995.
Size Trends: The AFDC/Relative population has remained stable at about 16,000 children between 1990 and 1995, while during the same period the Kin/FC population has grown over 230 percent, from 8,150 to 27,054. In comparison, the AFDC/Parent population grew 12 percent and the FC/Other population grew by 71 percent. These relations suggest several preliminary findings. The increase in formal kinship care in Illinois during the early 1990s was not apparently part of a general shift to kinship care that extended outside of the child welfare system. The increase in formal kinship was not paralleled by a decrease in informal AFDC/Kinship, which some had predicted. Finally, we can see that formal kinship is associated with (and probably pacing) an overall growth in Illinois child welfare that is reflected in the remainder of the foster care population.
Characteristics of Children by Living Arrangement
Tables 4.2 through 4.4 present distributions of certain characteristics among these two populations and several important comparison groups for time points in 1990 and 1995. The comparison groups available include all children active in an AFDC/Parent grant and all children living in Other (unrelated) Foster Care. Also, for 1990 only, the census estimates for all children in kinship care with no parent present, and an enumeration of all children in Illinois are available. We use all four groups as benchmarks by which to understand different aspects of the characteristics of children in formal (Kin/FC) and informal (AFDC/Relative) kinship care in Illinois.
Table 4.2 has three sections. The upper panel is a tabulation of counts of children in the four programs, in June 1990 and June 1995,(21) subclassified by a number of characteristic traits (region, age, race, gender, and race/region). The middle panel presents percentage distributions within each group, across these characteristic traits (e.g. the percentage of AFDC/Relative children who were ages 6-11 in 1990). The lower panel presents trait-specific prevalence for each program compared to the larger reference child population. For example, .9 percent of Cook County children were in AFDC/Relative settings in 1990, and 14.0 percent of all 6-11 year-olds in Illinois were recipients in an AFDC/parent grant in 1995.
Size: In absolute size, the AFDC/Parent population with over 400,000 clients dwarfs all of the other groups being discussed here. About 15 percent of the children in Illinois received an AFDC grant through a parent at any one time between 1990 and 1995. At no time in the period of study did any of the other groups include as much as 1 percent of the children in the state, although the foster care programs are each approaching that number.
Age: Using three 6-year age groups for comparison, the AFDC/Relative group is clearly older than the Kin/FC group and each of the other comparison groups in Table 4.2 (including all Illinois children). One-fourth (26 percent) of the children in AFDC/Kinship care are under the age of 6, compared to around 40 percent of the Kin/FC population and almost half (48 percent) of the AFDC/Parent population. The evidence that informal kinship care is more likely to be utilized as a caretaking response for older related children is further substantiated in Table 4.3, which imputes age and regional characteristics for the NON-AFDC informal kin group for 1990. Within this "unobserved" remainder of the Illinois informal kinship group, only 20 percent of the children are under the age of 6. The age composition of the formal Kin/FC group is almost a mirror image of the AFDC/Relative group with just under one-quarter of these cases falling in the older age group. The "youngest" group among those observed here is the AFDC/Parent category.
Gender: The levels of either type of kinship care have not appeared to vary by gender of the child. Among the groups described, only FC/Other appears to have a small gender gap with males, at 53 percent, being slightly more prevalent than females.
Region and Race: Compared to the child population of Illinois, all service-receipt defined groups described here are disproportionally African American and disproportionally located in Cook County (including Chicago) rather than in the balance of the state. These two effects are difficult to disentangle because while Cook County contains 43 percent of the state's child population, it includes almost three-fourths of Illinois's African American children. By 1995, all four programs observed were composed of almost two-thirds (or more) African Americans and almost two-thirds (or more) Cook County residents.
Overall, the kinship-based programs showed a slightly stronger racial pattern than comparable non-kinship programs. (Figure 2.1) For example, in 1995, an African American child was about ten times more likely to be in an AFDC/Relative setting than any other child, both in Cook County and across the rest of the state (2.2 percent versus 0.2 percent). In comparison, an African American child was only about five to seven times as likely to be in an AFDC/Parent setting (49.5 versus 10.0 percent in Cook, 41.3 versus 6.2 percent in rest of state).
Race has the greatest effect on the Kin/FC group. In 1995, over 5 percent of African American children in Cook County were in kinship foster care, while less than one-fourth of 1 percent of non African Americans were in kin foster care settings, a ratio of 20:1. The race ratio for Kin/FC in all downstate areas combined was 11:1.
For AFDC/Relative and FC/Other, there is no evidence that the higher levels in Cook County are anything more than a reflection of the racial differences observed above being applied to the racial composition of the state. In both cases, the Cook County prevalence is similar to the downstate prevalence within each racial category. The AFDC/Parent category shows a consistent tendency towards prevalence levels about one-fourth higher in Cook County, independent of race. However, the Kin/FC category, which showed the strongest race component, also shows a strong region effect independent of race. African American children in Cook County are three times more likely (5.0 versus 1.7 percent) to live in kinship foster care than African American children elsewhere in the state. Similarly, other-race children in Cook County are twice as likely (.24 versus .11 percent) to live in kinship foster care as other-race children elsewhere in the state.
In summary, it appears that AFDC/Relative cases (informal kinship) are directly influenced by race, with incidence among African American about ten times higher than other groups combined. Kin/FC in Illinois shows an even stronger racial component, which is intensified by an independent tendency for Kin/FC levels to be higher in Cook County. As a result, kinship foster care is very much a Chicago and African American dominated phenomenon in Illinois, the net result being that by 1995, 86 percent of kinship foster care placements were in Cook County and 86 percent of kinship foster placements were African American children. (See Figure 4.2).
Imputing characteristics for the "unobserved" informal kinship population.
Table 4.3 follows the format of Table 4.2, but for a much shorter list of characteristics and different comparison groups. For the most part, this Illinois analysis is based on the premise that the Illinois AFDC/Relative population is representative of all informal foster care in the state. With Table 4.3, we start to indicate where the biases in such an assumption may reside. This table starts with the 1990 census tabulations for the numbers of children living in households with relative caregivers and no parent present. By subtracting the AFDC/Relative population and the Kin/FC population from this total, we are left with counts of the "residual" informal kinship population.
This "unobserved" group contains 57 percent of the kinship care population of Illinois for 1990. We assume the percentage would become smaller by 1995 (because of the rapid growth of Kin/FC), but cannot guess whether the Kin/FC cases were likely to have been drawn from this "unobserved" group or from children living in own-child of parent relationships. Only a slight majority of the Non-AFDC informal population resides in Cook County -- this group includes only one-half of Cook County kinship care children and over two-thirds of kinship care children from the remainder of the state. Finally, as referred to above, this group contains an even smaller percentage of children ages 0-5 than the AFDC/Relative group.
We can observe that this "unobserved" informal kinship group includes children who tend to be older and less "urban" than the AFDC/Relative group. Because they are not program participants, we can assume that in the aggregate this population has at least a somewhat greater access to financial support. We can make no inferences about race, gender, or more detailed age characteristics.
Household Composition and Caretaker Characteristics: AFDC/Relative versus AFDC/Parent
Compiling Illinois AFDC data at the individual level offered the opportunity to compare certain characteristics of the AFDC/Relative population to the AFDC/Parent population. Although neither group represents the broader kinship and own-parent child populations, they comprise a significant segment of each of these, and clearly represent segments of the kinship and own-parent populations with which the public sector is involved.
Table 4.4 presents counts and percent distributions for characteristics of the households these children live in, and the key caretakers in those households, for 1990 and 1995. This household or caregiver data information is counted for each child, so a household with two children of the appropriate type will be counted twice. Certain "child-only" cases that were included in Tables 4.2 and 4.3 have had to be excluded from this table because the detailed descriptive information was not available in their records.
Over one-third of AFDC/Relative children live in households that also include own-children of the key caregiver. In contrast, almost two-thirds of the AFDC/Relative children live in households where the key caregiver has no own-children present. Thus, a significant segment of these related children are being blended into existing parent-child families. However, the majority of children in relative grants either cause a "new" family unit to be formed, or initiate "successor" family groupings created after an earlier generation of children has already left the household.
Almost half of the AFDC/Relative households include two or more adults, compared with only one-fourth of AFDC/Parent households. In both types of household, the key adults are preponderantly female, although one in ten is male in a relative-grant household and one in twenty in a parent-grant household. A relative child's caregiver is more likely to have been married at one time than a parent-grant caregiver, but is no more likely to be currently married with a spouse present.(22) All these characteristics suggest that some AFDC/Relative households might have access to more social and possibly financial resources than the typical AFDC/parent household. This makes intuitive sense because while both programs were developed to support the child, the AFDC/Relative case is often defined by the resource limits of the non-caregiver parent, in contrast to the AFDC/Parent case which is typically defined by resource limits of the caregiver parent.
The characteristic that most clearly differentiates AFDC/Relative and AFDC/Parent cases, though, is the age of the primary caregiver. (See Figure 4.3). For AFDC/Parent cases, 89 percent of the key adults were younger than 40 years of age, and only 1 percent were 50 years of age or above. In contrast, for AFDC/Relative cases just over one-third (37 percent) of the key adults were younger than 40, about 40 percent were over 50, and 16 percent were over 60. This is explained primarily by the fact that almost four out of every five relative-child caregivers are the child's grandparents. The aging of some members of this relative caregiver population (median age in the mid-40s,) clearly limits the social and possibly financial resources available for the caring of children in some households.
Movements of children between living arrangements over time.
Records for individual children receiving AFDC grants in Illinois have been matched to records from the Foster Care tracking system through probabilistic record-linkage procedures. This process identifies those children who have had contact with both public systems, with the result that any child's foster care events can be joined directly with that child's AFDC events to create a combined "welfare-career" history. The population studied includes all children who were AFDC recipients and/or foster children any time between June 1990 and June 1995.
Using this linked file, we examined the interrelationship between the AFDC and foster care populations of Illinois, and identified and analyzed the transitions of children that do (or do not) move between contacts with these two systems over time. In this analysis, we are particularly interested in information that helps to characterize the formal and informal kinship populations. In the previous section, we considered some of the characteristics of these populations. Here, we can start to describe where the children in each type of kinship care come from, how likely they are to shift program auspices or care arrangements, where they go when they leave kinship care, and whether the same children tend to become involved in both informal and formal kinship living situations.
The movements among living arrangements are examined by considering annual transitions between living arrangement categories as identified by the yearly June cross-sections compiled from 1990 through 1995. As with the previous analysis, each child can be classified by one of the four program categories -- AFDC/Parent, AFDC/Relative, Kin/FC, and FC/Other. For the dynamic analysis, we also classify inactive participants into one of three non-program categories -- Not Yet Born, Aged Out (18+), or Out of Scope. "Out of Scope" is a residual category, invoked when none of the other six statuses apply. In many cases, this reflects a positive situation; such as when a child is living with his or her own parent(s) in economic self-sufficiency. But, the "out-of-scope" does not necessarily imply a positive setting, it just means that the child is currently not involved in either of the two programs being tracked.
Data describing annual transitions of children between these categories is presented in some detail in the Appendix to this section. For clarity, most of the information presented here is based on a pooled average of the five separate June-to-June transition periods. For the most part, data pooling has the effect of stabilizing and simplifying the results, without distorting them.(23)
Table 4.5 presents the basic transition matrix for the AFDC/Foster Care categories. Each cell represents the average number of children living under the arrangement described on the left (row label) in one June, and, who then lived in the arrangement described at the top (column label) in the subsequent June. The cells along the diagonal, shaded for easy recognition, contain cases of net non-transition -- children living in the same class of living arrangement in both the initial and the subsequent June. Each cell off of the diagonal represents a particular group of "movers" and each cell on the diagonal represents a certain type of "stayers.(24) We should read Table 4.5 with statements like "there was an average annual movement of 317 children from AFDC/Relative homes to Kin/FC placements," or, "of 430,955 children who receive AFDC in their parent's care in one June, 328,945 are still active as AFDC/Parent cases in the following June."
Transition rates "from" a status: Table 4.6a converts the counts from Table 4.5 into annual transition rates: the proportion of children that start in their prior status and that end up in their subsequent status. Close examination of Table 4.6a suggest that this annual "transition" matrix is indeed dominated by "stayers." The stationary tendency of these living arrangement groups is apparent because the proportions in each cell along the diagonal is over .700, meaning that over 70 percent of these children can be expected to end the year in the same type of living arrangement where they started the year; 70.5 percent of AFDC/Relative children and 79.6 percent of Kin/FC children are "stayers" in the average year. Movement between these statuses is infrequent: the most likely transitions observed are to "age out" and "out-of-scope." The largest transitions between program categories are .070 from AFDC/Relative to AFDC/Parent and .079 from FC/Other to FC/Relative, with both types being shifts within the same agency.
The transition rate of children from AFDC/Relative to Kin/FC is twice as large as that from AFDC/Parent to Kin/FC. This suggests that living in kinship arrangements outside of the foster care system increases the likelihood that the child will move to foster care kinship placements. But the magnitude of these transitions -- each less than 2 percent/year -- is much smaller than might have been expected based on arguments posing that a process of "inappropriate" substitution of Foster Care for AFDC has fueled the growth of kinship foster care in Illinois. This evidence, based as it is on annual net transitions, cannot convincingly deny the substitution argument, particularly if the hypothesized living arrangement status changes from "other" to "AFDC Kin" to "Foster care kin" would be expected to occur very quickly. In the absence of fully longitudinal event data, it suggests that this pattern of event processes is probably rather uncommon. Clearly there has not been any widespread movement of long-term AFDC/Relative cases into Kin/FC.
As might be expected given the high proportion of "stayers," the evidence suggests that kinship placements -- especially formal foster care placements -- are unlikely to lead to reunifications with own-parents of the child. The "observed" reunifications here are transitions to AFDC/Parent, while the "unobserved" reunifications are an unknown subset of the "out-of-scope" category. An average of 7 percent of AFDC/Relative cases shift to AFDC/Parent annually, compared to only just over 3 percent of Kin/FC cases. About twice that many from each group shift into the out-of-scope group, which includes own-parents not on AFDC, kin placements not on AFDC, moves out of Illinois, or placement in any program not tracked here (mental health, detention, etc).
The same information is shown in Table 4.6b, focusing on changes by looking only at the proportionate distributions of "movers" across the destination status. Of those children who leave AFDC/Parent, over 9 in 10 either age out (.102) or move out (.825) of the domain of programs being tracked. Formal and informal kinship groups are somewhat more intertwined with other program categories. About one-third of AFDC/Relative "movers" and over one-half of Kin/FC "movers" shift to other program settings.
Composition by source: Table 4.7a presents the same basic transition information once again, but reverses the viewpoint of Table 4.6a. Instead of looking at rates of transition forward from one status to another, this table looks backward in time, decomposing the population of each status by where its incumbents lived the previous June. The "aged out" category is logically replaced by a "not yet born" category to represent infants who enter one of these programs during their first year. As with Tables 4.5 and 4.6a, this table is dominated by the stationary cases -- the "stayers" along the diagonal. Two-thirds to three-quarters of the children in each program group had been in the same living arrangement status during the previous June. This table also suggests that even though a small proportion of the children in these care relationships moved from one of the other categories within the year, the AFDC/Parent population is clearly the most significant "feeder" to the other three programs.
The dynamics of these changes are more easily viewed in Table 4.7b, which presents the same composition by previous status information, but only for those children -- the "movers" --who made a transition in the previous year. Looking across the first row, we see that 43.2 percent of the children who moved into AFDC/Relative and 50.3 percent of the children who moved into Kin/FC lived in an AFDC/Parent arrangement the previous June. In contrast, there is very little net movement observed between the informal and formal kinship groups themselves -- only 1.2 percent of AFDC/Relative children moved from Kin/FC and 4.6 percent of Kin/FC cases moved from AFDC/Relative.
Children who move into the informal kinship group come almost entirely from the AFDC/Parent (43 percent) and out-of-scope (47 percent) categories. This group has the smallest proportion of newborns (8 percent) and receives very few (less than 2 percent) of its new cases from the foster care groups. Over half the children who move into the formal kinship group (Kin/FC) come from the AFDC/Parent group. Another one-third are from the out-of-scope and newborn groups. In contrast to informal kinship, a substantial (though not large) proportion of the formal kinship cases come from either FC/Other (11 percent) or AFDC/Relative (5 percent). Clearly the children in formal kinship care arrangements are historically more connected to the public support system than children in informal kinship arrangements.
The main dynamic apparent in Tables 4.7a and 4.7b is the size of the impact that the AFDC/Parent population has on the composition of these other program categories. Because the population of children receiving grants through parents is so much larger than any other groups examined here, even relatively small proportional transitions from AFDC/Parent cases result in very substantial proportional flows of children into either AFDC/Relative or either foster care status. In the discussion of Table 4.6a we noted that the likelihood of an individual child making the transition from AFDC/Relative into formal kinship was twice as great as the likelihood of transition from AFDC/Parent into formal kinship. However, because the AFDC/Parent population is so large, the aggregate number of cases coming into Kin/FC from AFDC/Parent is larger, over ten times larger, than that coming from AFDC/Relative arrangements.
The apparent contradiction between lower likelihoods and higher net impact is fully explained by the relative sizes of the base populations, and the numbers presented here provide a direct way of visualizing these relations. One clear implication is that even very small shifts in the pattern of movement of children from the AFDC/Parent living arrangement category produces very large impacts on the flow of cases to the three smaller groups.
Composition of 'mover' and 'stayer' Groups.
The preceding section showed that informal and formal kinship groups demonstrate different patterns of transition -- whether viewed by where they tend to come from before they enter the kinship setting, by what type of living arrangements they move to at the end of their kinship stay, or by their likelihood of remaining in place in the current kinship home. Here, we examine some demographic characteristics of the various kinship transition groups to see what can be learned about how different types of children might be expected to have different career patterns. Table 4 .8 is divided into four panels presenting region, age, and racial percentages for each transition (or non-transition) subgroup of children that move from AFDC/Relative (Panel A), into AFDC/Relative (B), from Kin/FC (C), or into Kin/FC (D). In each sub-table, the leftmost numeric column is the distribution of children who do not move -- the "stayers," which can be used as a reference group against which to look for differences for those children that change living arrangement status.
Age: The children remaining in informal kinship (AFDC/relative homes) from year to year tend to be older than children moving from AFDC/Relative to formal foster care placements (Kin/FC), and older than those who return to own-parent AFDC homes. This can be seen in Panel A, where only 25 percent of the AFDC/Relative "stayers" are ages 0-5, while 35 percent of the AFCDC/Rel->Kin/FC group and 50% of the AFDC/Rel-->AFDC/parent group are ages 0-5. The ages of children moving "out-of-scope" resemble the AFDC/Relative "stayer" group. Only the small group of children moving into FC/Other appears to be systematically composed of older children.
The same pattern of age relationships applies with minor variations to children moving into informal kinship, as well as those moving into and out of formal kinship arrangements (Kin/FC). Overall, the bulk of the movement between these categories involves younger. Older children tend more often to either stay in their current living situation or exit the program domain that we can observe. The one exception is FC/Other (non-kinship foster care) which tends to "send" young children to other programs, but which is unusual in that it "receives" a disproportionate share of older children from the kinship categories (particularly from the Kin/FC group).
The more rapid circulation of younger children (in and out of kinship placements) suggests these children's early years are most likely to be typified by disruptions, uncertainties and change. As the child ages, the living situation tends to stabilize along one or more of many dimensions: clarity about whether or not the birth parent might resume care, an understanding of the willingness and capacity of the relative to maintain caregiving, and evaluation of whether the program niche (formal versus informal) seems workable. It is telling that as kinship relations (of either type) end for some reason during a child's adolescence, the likelihood of entering a non-relative foster care placement increases. This suggests that other settings are less likely to remain feasible as options by this stage in the child's life.
Race and Region: The racial and regional composition of transition groups follows patterns that could, for the most part, be predicted from the overall compositions of the program groups. Both the AFDC/Relative and Kin/FC groups are disproportionally composed of African American children and children who live in Cook County. As was noted above, there is a particularly strong joint effect of race and region for the Kin/FC group, so that almost 80 percent of kin foster care placements were of African American children in Cook County.
Of the children moving from AFDC/Relative placements to Kin/FC, 87 percent were from Cook County and 91 percent were African American. The AFDC/Relative "stayers" were 68 percent Cook and 76 percent African American. Only movers "out-of-scope" were significantly lower, with 56 percent Cook and 66 percent African American. Looking to children in kinship foster care, the "stayers" were 88% in Cook County and 88 percent African American, with all "mover" groups somewhat lower.
Relationship: The relationship between the child and kinship caregiver can be identified only for the AFDC/Relative population. The great majority of these informal kinship caregivers are grandparents (78 percent), with most of the "other" category being aunts. In Panel A of Table 4.8, we can see a slight tendency for grandparent-child living arrangements to stay more intact from year-to-year than arrangements where the child lives with other relatives. The other-relative arrangements are somewhat over-represented in moves to non-kinship foster care and in moves to the unobserved "out-of-scope" statuses.
(16) The "own children" tables can be classified by employment characteristics of parents and subfamily composition, and the "all related children" tables can be classified by poverty status -- but these cannot be directly compared or linked together because one is grouped by "parent" characteristics and the other by "head of household" characteristics.
(17)Anecdotal evidence pointed to the possibility that young mothers and their children are frequently lumped into grant units where the adult recipient is a relative of the mother. A "credible mother" was defined as a minor female, more than 13 years older than the child in question, with a relation to the grantee that would be consistent with the child's relation to the grantee. Thus for a 2-year-old who is the grandchild of the grantee, we would search the list of household members for "daughters" of the grantee between 15 and 17 years of age. There is no way to confirm that this "credible" female is actually the child's mother, and there are many possible scenarios where she would not be. However, two considerations led us to preemptively force the classification of "parent" on this type of case. First, these determinations were strongly corroborated by a data field that indicated whether or not a mother had been present at the child's entrance to the case. We found very few "credible" mothers in cases where the mother had not been present, and we did find "credible" mothers in the majority of cases when a mother had once been present. Second, analysis of these "credible mother" present cases shows that a new mother-child grant unit is often formed within the next few years, suggesting that the anecdotal information is, at least in the aggregate, often correct.
(18)In certain "child-only" cases, the grant support is provided to payees who receive only on behalf of the custodial child and who are not active recipients on the grant themselves. Unfortunately, because of their non-recipient status in the case, most key demographic information for these caretakers is not coded or maintained by the state agency. Thus, these cases are not included in discussions of household and caretaker characteristics. Fortunately, they represent a fairly small share of the Illinois AFDC population during the period of study.
(19) "Other foster care" includes all foster care activity that is not defined as "kinship" foster care, e.g. non-relative family foster care, emergency shelter care, and congregate care placements.
(20) cf. A Report from the Multistate Foster Care Data Archive: Foster Care Dynamics 1988-1992 Chapin Hall Center for Children 1994.
(21) Data for the intermediate years is available, but only the end-years were presented to simplify presentation. No internal patterns were noticed to suggest that the additional years would contribute substantive changes to interpretation.
(22) The marital status tabulations both contain a substantial "unknown" category, and this reporting is based on the partial information remaining.
(23) The main dynamic that is hidden by this pooling of data from five sets of annual transitions has already been described, namely, the extremely rapid growth of kinship foster care in Illinois. Although the size of transitions into Kin/FC do reflect the growth of this group between 1990 and 1995, we have observed no real changes in associated trends or patterns apart from the overall shift in incidence.
(24)" "Stayers", in particular, must be understood as being defined by a ?net' outcome. A certain amount of movement in between the June points of observation is not captured in this analysis, and the fact that a child was in a living arrangement in both Junes does not require that the child did not experience two or more moves in between.
