In the same way that census and administrative foster care data can be combined to allow examination of child living arrangements separately in four states, data from these sources can produce similar information for smaller geographically defined places within each of these states. As was the case with decomposition of formal and informal kinship along substantive lines (e.g. age groupings above), disaggregation of these populations requires that the same criteria be available for both census and the Archive data. The census data can be mapped to many different local-area levels, but the geographic information currently available in the four-state foster care data is organized at the county level. Thus, while census data limited substantive decomposition to two age groups, the Archive data provided by the state agencies limits the geographic decomposition to counties.
The initial plan for this analysis was to systematically compare the formal and informal kinship care levels for counties within each state to search for patterns in their variability that could help us gain insights into cross-state regularities and within-state patterns. Other census-based areal indicators of such factors as ethnic distribution, poverty levels, employment, etc., were to be employed in this investigation. The living arrangement information for counties in each of these four states is presented as Table 3.3. Within each state, the counties are ordered by child population, from largest to smallest. Because some of the counties have very small populations, care must be taken in interpreting the rates and percentages for these places.
Initial inspection of county kinship care patterns revealed one overarching finding: not only are formal and informal kinship care distributed unevenly across places within these four states, as we might have expected, but the practice of kinship foster care is almost exclusively limited to the primary urban areas in most of these states.(14) This is most clearly the case in New York, where over 95 percent of all kinship foster care placements involve children from New York City. In Illinois, the use of kinship foster care is also highly localized, with 69 percent of the state's formal kinship placements in Chicago. Although Cook County, including Chicago, had over 6,800 kinship foster cases in 1990, the county with the next-largest frequency was St. Clair County (East St. Louis, IL) with only 213 cases.
Of the four states observed, the only one showing significant levels of kinship foster care in areas away from the primary urban place is California. Los Angeles County, containing 30 percent of the state's children, generates almost half (47 percent) of the kinship foster care placements. But, formal kinship care remains fairly common in a number of the other larger urban counties -- San Diego, Sacramento, Alameda (Oakland), and others. The prevalence of kinship foster care in this group of counties ranges from about one-half to two-thirds of the 6.1 per thousand level observed in Los Angeles. San Francisco County has the highest rate of kinship care prevalence in the state--at 11.8 children per thousand, it is almost double the rate in Los Angeles.
Although the lack of variance in formal kinship care precluded a full ecological analysis of counties for the four states, we have attached a few county-base indicators, including race/ethnicity, which would have been used for such an analysis, in Appendix 3. Most of the California counties with higher kinship foster care prevalence also have the largest percentage of African American children (Alameda, San Francisco, Contra Costa, Sacramento, and Los Angeles), and African American children are indeed over-represented in their formal kinship caseloads. Only one California county with a high African American population, Solano County, has very low kinship foster care rates. California is also characterized by a large Hispanic population. The levels of informal kinship care tend to be higher in counties with larger proportions of Hispanic persons-- such as Imperial, Tulare, Fresno and Los Angeles Counties.
Outside of California, county-based analysis of formal versus informal kinship patterns is not instructive because almost all of the areal variation in formal kinship is explained by location in the primary urban place. Instead of forcing an implausible method on these data, the analysis has been simplified to examine the differences observed between these primary urban places and the remainder of each state. Summary state totals and subtotals for "primary urban place" and "balance of state" are presented at the bottom of each state subtable in Table 3.3.
Figure 3.4 presents age and care-type prevalence rates separately for the primary urban place and the "balance"of each state. The relationships shown in these graphs confirm the earlier finding that formal kinship care levels are either extremely small or virtually nonexistent in the balance of each state. Even in Missouri, a state with very low levels of kinship foster care, this population is concentrated in St. Louis. Only California shows any substantial amount of formal foster care outside of the primary urban place, and levels for the rest of the state are still much lower than those observed for Los Angeles County. A second observation is that levels of informal kinship care are consistently higher in the primary urban places than in the balance of each state. A final observation can be made by examining only the upper chart in Figure 3.4 -- the graph for the primary urban counties-- as these are the only places where both formal and informal kinship arrangements occur with any regularity. In all four large cities and within each age group, there is apparently a strong inverse relation between the levels of informal and formal kinship care. That is, when informal kinship is relatively high, formal kinship is relatively low, and vice versa. This is most apparent in the 0-5 year old age group. For example, St. Louis, Missouri shows the highest level of informal kinship (24 per thousand) and the lowest levels of formal kinship (2 per thousand) for young children. Conversely, 0-5 year olds in New York City have the lowest observed levels of informal kinship (12 per thousand) along with the highest levels of formal kinship (19 per thousand). Los Angeles and Chicago are each mid-range for both types of kinship care. A similar relationship, though not quite so strong, exists for the 6-17 year olds.
This inverse relation between the prevalence of formal and informal kinship care in the cities has potentially important implications. Although observations from four places do not provide overwhelming evidence, there is clear suggestion here of a possible substitutability relationship between formal and informal kinship care. Overall kinship care rates for both types (formal and informal) combined are remarkably similar across the four cities: for 0-5 year olds they vary between 26 and 31 per thousand; and for 6-17 year olds, they vary between 41 and 46 per thousand. What is different is how these cases have been sorted between informal arrangements and formal foster care placements in different cities. The fact that a formal response is most likely to be invoked in New York City and least likely in St. Louis is objectively clear in these numbers. What is new is the tentative hypothesis that the response, and not the underlying condition, differs between places. This finding suggests that the children being placed in kinship care arrangements are fundamentally similar in these cities, and that differing public actions (due to policy, casework practices, etc.) create the variation in informal versus formal kinship levels. Factors such as local court decisions, agency placement priorities and payment guidelines, and federal reimbursement claiming strategies might be the more productive basis for understanding levels of kinship foster care, per se, than underlying social causes.