This analysis is based on data from the March Current Population Surveys (CPS) for the 12 years between 1983 and 1994. "Children in kin care" or "kin-care children" refers to never-married children who do not live with their parents, but live with other adult relatives. Whether these children are in formal foster care arrangements cannot be ascertained from the CPS data. "Children" refers to all individuals aged 17 or younger. "Caregivers" refers to both parents and the other relatives who take care of children. "Kin-caregivers" refers to the adult relatives, other than parents, who take care of children. We assume kin-care children are being cared for by the head of the family they are in and, if such a person exists, the head's spouse. We assume foster children are being cared for by the head of the household and, if such a person exists, the head's spouse. (For more detail on definitions, refer to Appendix I at the end of Volume II.)
In families headed by both a husband and a wife, "educational attainment" refers to the better educated caregiver. "Labor force status" refers to the caregiver most attached to the labor force, with "attachment" being greatest for those who were employed, followed by those who were unemployed, then those not in the labor force.
The original plan was to base analysis on four pooled 3-year groups--1983-85, 1986-88, 1989-91, and 1992-94--and to report the average annual values of indicators for each period. This 3-year averaging is necessary because the number of children in kin care in a single year's CPS sample is relatively small and therefore could provide unreliable estimates. However, the CPS data from 1993 and earlier were not comparable to the data collected in 1994. Between 1993 and 1994, the percentage of children in kin care jumped from 2.2 percent to 3.1 percent, an increase of 42.4 percent. Over the 12 years studied, the next largest percentage change in the prevalence of kin care was a 10 percent increase between 1986 and 1987, from 2.1 percent to 2.3 percent. It appears that the marked increase is due to changes in CPS methodology, rather than dramatic increases in the prevalence of kin care. In 1994, several new changes were instituted. The sampling frame used for the 1994 CPS was the first based on the 1990 census, there was a switch from paper questionnaires to computer-assisted interviewing, it became easier for interviewers to code unusual types of living arrangements, and data were processed somewhat differently.
The improved measurement of kin care produced a reporting quandary. The latest, and probably best, data were from the 1994 CPS, but including these data in trend estimates would create the false impression that the annual average number of children in kin care is skyrocketing. We arrived at a two-pronged compromise approach. In analyzing changes over time in the percentage and number of children in kin care, we limited analysis to years for which data on household and family relationships were collected and processed consistently, 1983-93. For analyzing the characteristicsof children in kin care and their families, we extended our analysis to include data from the 1994 CPS. In general, including the 1994 data did not change the results significantly; where there were changes, the addition of the 1994 data usually strengthened already observed trends.
Although our trend analysis stops in 1993, it appears that the
changes introduced in the 1994 CPS improved the identification of
children in kin care. Our current best estimate of the number of
children in kin care is 2,150,000 in 1994, 3.1 percent of all
children.
Introduction
Between 1983-85 and 1992-93, the number of children in the
United States grew modestly, from about 62,532,000 to 66,639,000,
an increase of 6.6 percent (see Table 1.1).
Over the same period, the number of children in kinship care
increased slightly faster, from about 1,282,000 to 1,390,000, an
8.4 percent increase. This increase was due to increased
prevalence of kinship care among African American children, which
changed from 5.2 percent to 6.1 percent between 1983-85 and
1992-93. There was no increase for non-Hispanic whites and only a
very small increase for Hispanic whites, from 2.4 percent to 2.7
percent.
In this first section of the report, we will discuss:
Who is in Kin Care and How Has It Changed?
Race and ethnicity. Kinship care is most prevalent among African Americans and least prevalent among non-Hispanic whites (see Table 1.2). In 1992-93, 6.1 percent of all African American children, but only 1.1 percent of non-Hispanic white children, were being cared for by a relative other than a parent. The prevalence for Hispanic whites, 2.7 percent in 1992-93, is higher than for non-Hispanic whites, but still substantially below the prevalence for blacks. (Sample sizes for other racial/ethnic groups are small.) Between 1983-85 and 1992-93, the prevalence of kinship care for blacks increased substantially, from 5.2 percent to 6.1 percent. For other groups, prevalence remained constant.
Age. The prevalence of kin care increases with age, from 1.3 percent for children aged 0 to 4 to 3.6 percent for children aged 15 to 17. The sharpest increase in prevalence is between age group 10-14 and age group 15-17; between these ages, the prevalence of kinship care increases from 2.3 percent to 3.6 percent. Over the period studied, there are no major changes in the prevalence of kinship care by age.
Sex. There is no difference in the prevalence of kinship care for males and females.
Geography. Kinship care is substantially more common in the South (2.9 percent) than in the other three regions, the Northeast (1.9 percent), the West (1.7 percent), or the Midwest (1.5 percent). In 1992-93, kinship care is also somewhat more common in nonmetropolitan than metropolitan areas, 2.5 percent versus 2.0 percent. Over the period studied, prevalence rates did not change substantially for the regions or for metropolitan or nonmetropolitan areas.
Characteristics of children in kinship care: How have they changed, and how do they compare with children in parent care?
In the last section, we discussed how, within various groups, the percentage of children in kinship care changed between 1983-85 and 1992-94--for example, how the percentage of African American children in kinship care has changed. In this section, we discuss a related topic, how the characteristicsof children in kinship care have changed--for instance, how the percentage of kinship care children who are African American has changed.(1)
We compare changes in the characteristics of kin-care children with changes in the characteristics of parent-care children, focusing on changes that affected kin-care children more than parent-care children. We discuss three sets of characteristics: characteristics of the children (see Table 1.3); characteristics of the caretakers (see Table 1.4); poverty and use of services by the caretaking family (see Table 1.5).
Characteristics of children
Race and ethnicity. In the U.S. population overall, the percentage of children who are non-Hispanic whites is decreasing. This decline has been more dramatic for kin-care children than parent-care children (see Figures 1.1 and 1.2). In 1983-85, white non-Hispanic children made up the largest share of kin-care children, 46.6 percent, but this share had dropped to 36.2 percent by 1992-94. Over this period, non-Hispanic African Americans went from being the second-largest to the largest group in kinship care, from 38.1 percent to 44.3 percent. The share of kin-care children who were Hispanic white also increased, from 10.5 percent to 14.0 percent. Among children in parent care, white non-Hispanics continue to dominate, making up 73.1 percent in 1983-85 and 68.7 percent in 1992-94. The percentage of children in parent care who are African American has remained stable, while the percentage of children who are Hispanic whites has increased.
Age. In all four periods, children in kinship care were, on average, older than children in parent care (see Figures 1.3 and 1.4). For instance, 16.7 percent of children in kinship care in 1983-85 were aged 0-4 and 27.1 percent were aged 15-17 while, among children in parent care, 29.0 percent were aged 0-4 and 17.0 percent were aged 15-17. By 1992-94, however, the differences in age distribution had lessened slightly. Between 1983-85 and 1992-94, for both kin-care children and parent-care children, the percentage of children under 10 increased and the percentage of children aged 10 and older decreased, but this increase was greater for children in kinship care--especially among children aged 0-4. By 1992-94, 21.7 percent of children in kinship care were aged 0-4 and 23.5 percent were aged 15-17 while 29.8 percent of children in the care of their parents were aged 0-4 and 14.7 percent were aged 15-17.
Sex. There are slightly more boys than girls in all four periods for parent care and for all periods except 1992-94 for kin care.
Geography. In all four periods, the largest share of kin-care children lived in the South (48.6 percent in 1983-85 and 46.8 percent in 1992-94), with the remainder roughly evenly divided among the other three regions (see Figure 1.5). The largest share of parent-care children also live in the South, although the share is substantially smaller (33.7 percent in 1983-85 and 33.8 percent in 1992-94).
The proportion of both kin-care and parent-care children living in metropolitan areas increased between 1983-85 and 1992-94 (see Figure 1.6). For kin-care children, the percentage in metropolitan areas increased from 59.1 percent to 74.5 percent and for parent-care children, the percentage rose from 63.7 percent to 77.4 percent.
Caregiver characteristics
In 1989-91 and 1992-94, the only two periods for which we could ascertain the relationship between kin-care children and their caregivers, two-thirds of kin-care children were being cared for by one or both grandparents, usually a grandmother.(2)
As we discuss in this section, the caretakers of kin-care children are older and more likely to be single women--especially widows-- than the parents caring for their own children.
Children in kinship care are substantially less likely to be taken care of by both an adult man and an adult woman than children in parent care (see Figure 1.7). Although the proportion of children living in families headed by two adults has declined for parent-care children, the decline has been even sharper for kin-care children. Between 1983-85 and 1992-94, the percentage of parent-care children living in a family headed by a married couple declined from 76.5 percent to 72.4 percent; over the same period, the percentage of kin-care children in this type of family declined from 56.3 percent to 50.3 percent.
For children in parent care, there has been an increase in families headed by both single women and single men. There has been an increase in the percentage of kin-care children being cared for by single women--from 38.7 percent in 1983-85 to 43.8 percent in 1992-94--but little increase in the percentage being cared for by single men--from 5.0 percent in 1983-85 to 5.9 percent in 1992-94. Among kin-care children being cared for by a single woman, the largest group is those being cared for by widows (16.9 percent in 1983-85 and 15.2 percent in 1992-94). However, the percentage being cared for by divorced and never-married women is increasing, from 8.5 percent and 4.9 percent (respectively) in 1983-85 to 11.3 percent and 7.8 percent in 1992-94.
One of the most striking differences between children being cared for by their parents and children being cared for by other relatives is the age of their caregivers (see Figure 1.8). Among children who live with their mothers, in all four periods more than 95 percent have mothers under age 50. However, less than half of kin-care children with a female caregiver have a caregiver under age 50. In 1983-85, about half of these children had female caregivers under 50, about a quarter had female caregivers in their fifties, slightly less than 20 percent had female caregivers in their sixties, and more than 5 percent had caregivers aged 70 or older. As we have noted, since 1983-85, the average age of children in kin care has declined, but the opposite is true for their female caregivers. Since between 1983-85 and 1992-94, the percentage of kin-care children with female caregivers whose female caregiver is aged 50 or older increased from 50.5 percent to 56.6 percent. Similar patterns can be observed for fathers and male kin-caregivers.
Parents are better educated than kin-caregivers (see Figure 1.9). For instance, in 1983-85, kin caregivers were substantially more likely than parents to have dropped out of high school (47.9 percent versus 17.0 percent) and were substantially less likely than parents to have graduated from college (8.1 percent versus 23.7 percent). By 1992-94, the educational attainment of both groups of caregivers had increased, although the pattern of improvement differed. For parents, there were declines in the percentage of high school dropouts (from 17.0 percent to 14.2 percent) and high school graduates (from 36.0 percent to 30.8 percent), and increases in the percentage of those with some college (from 23.3 percent to 28.2 percent) and with college diplomas (from 23.7 percent to 26.8 percent). In contrast, among kin-caregivers, there were declines in the percentage of high school dropouts (from 47.9 percent to 42.6 percent) and increases in the number of high school graduates (from 29.9 percent to 32.7 percent), but little change in the number who had some college (from 14.0 percent 16.7 percent) or had college diplomas (no change at 8.1 percent).
Parents are more likely to be in the labor force than kin-caregivers (see Figure 1.10 and Table 1). In 1992-94, 83.0 percent of children in parent care had at least one parent who was employed compared with only 57.5 percent of children in kinship care. Children in kinship care are about three times more likely than children in parent care to live with a caregiver who is not in the labor force, 39.1 percent versus 12.4 percent. We only have information about the activities of caregivers who are not in the labor force for one period, 1983-85. Among those with a caretaker not in the labor force, children in parent care are more likely to have a parent who is either housekeeping or attending school; children in kinship care are more likely to have a caregiver who is either unable to work or "other/retired." These differences appear to reflect the older average age of kin-caregivers. The labor force status of both kin-caregivers and parents changed little over the four periods examined.
Poverty status and use of services
Kin-care children are more likely to be poor than parent-care children (see Figure 1.11). Far more children in kinship care are in families whose income is below the poverty line, 38.8 percent versus 21.4 percent in 1992-94. Far fewer children in kinship care than in parent care are in families whose income puts them above 150 percent of the poverty line, 44.1 percent versus 67.6 percent. Over the four periods studied, there has been little change in poverty status for kin-care and parent-care children.
In all four periods, children in kinship care are substantially more likely to be in families with no earned income than children who live with their parents (see Table 1.5). Furthermore, the percentage of kin-care children in families with no earned income has increased, while there was little change in the percentage of parent-care children in families with no earned income. From 1983-85 to 1992-94, the percentage of kin-care children in families with no earned income rose from 23.8 to 26.3, while the percentage of parent-care children in families with no earned income went from only 8.9 percent to 9.2 percent.
Children in kinship care are substantially more likely than children in parent care to be in families receiving government assistance (see Figure 1.12). In 1992-94, kin-care children were more than twice as likely as children living with their families to be in families receiving "public assistance or welfare" (27.0 percent versus 13.3 percent), were almost five times as likely to be in a family in which someone collects supplemental security income--SSI, a welfare program for the elderly and disabled poor (14.5 percent versus 3.0 percent), were twice as likely to be in a family in which the children receive free lunches at school (49.8 percent versus 25.4 percent), and are nearly twice as likely to be in public housing (6.7 percent versus 4.0 percent) or in a household receiving food stamps (31.2 percent versus 18.9 percent). In addition, kin-care children are more than five times more likely to be in a family in which someone receives social security (34.6 percent versus 6.4 percent) and two-and-a-half times more likely to be in a family in which someone receives disability insurance payments (3.6 percent versus 1.4 percent).(3)
The largest differences in program participation are in those programs aimed primarily at the elderly--SSI and social security--which is no doubt the result of the large portion of elderly caretakers.
For kin-care children, family participation in three government programs increased noticeably between 1983-85 and 1992-94: public assistance/welfare (from 21.1 percent to 27.0 percent), children receiving free lunches (42.0 percent to 49.8 percent), and SSI (10.1 percent to 14.5 percent). For parent-care children, family participation in these three programs also increased, although not as dramatically. This increase in use of government services is somewhat unexpected because, between 1983-85 and 1992-94, for kin-care children, there was little change in either the percentage in poverty (from 39.2 percent to 38.8 percent) or the percentage "near poor" (from 15.1 percent to 17.1 percent)--that is, at or above poverty but no more than 150 percent of the poverty line.
Children in kin care and children in foster care: Comparisons for 1992-94
Some children in kin care are also in foster care; that is, they are being cared for by relatives other than their parents, but they are under the authority of the foster care system in their state. Unfortunately, we cannot identify these children. Until 1988, children in foster care were not identified in the CPS. However, in identifying how children are related to others in the household, respondents must choose between identifying them as foster children or as relatives; there is no way to identify them as both. Our intuition--albeit, with no empirical basis--is that, if forced to choose between the two ways of identifying the children, adults who have had relatives placed with them as foster children will identify these children as relatives, rather than foster children. Our population of kin-care children probably, therefore, includes some children who are actually part of the foster care system.
In this section, we compare the characteristics of children identified in the CPS as foster children with children we could identify as kin-care children for the 1992-94 period. As noted, these are not clean comparisons because foster children being cared for by relatives can fall in either group. Foster and kin-care children have in common the unfortunate circumstance that their parents either cannot or will not care for them adequately. If our assumption about how related foster parents identify these children is correct, then the comparisons noted in this section will show how children placed with non-relatives compare with children who are being cared for by relatives, whether placed with these relatives by the foster care system or through other arrangements. Many of the differences in the living arrangements of foster and kin-care children probably arise from the requirement that foster parents meet certain criteria in order to have children placed with them. Informal kin-caregivers do not have to meet these requirements. Furthermore, in some states, relatives who become foster parents face fewer, or less stringent, approval or licensing requirements.
The most striking difference between foster care and kin-care children is one of magnitude: in 1992-93, there were an average of 1.4 million children in kinship care each year, compared with only approximately 200,000 children in foster care reported in the CPS. In 1993, there were actually 440,000 children in foster care nationally.(4)
Children in foster care being cared for by relatives is responsible for some, but not all, of this undercount of foster care children in the CPS. Other reasons for the undercount are not known at this time.
Characteristics of children
Race/ethnicity. Like kin-care children, a much larger proportion of foster children than parent-care children are members of a minority group (see Table 1.6). Equal proportions of foster and kin-care children are African American (42.8 percent versus 44.3 percent), but a much smaller share of foster children are non-African American Hispanics (14.6 percent versus 7.9 percent).(5)
Although an equal percentage of foster and kin-care children are African American, it is instructive to look at the number of African American children in each status: during the 1992-94 period, each year there was an average of 735,274 African American children in kin care, more than seven-and-a-half times the number in foster care, 95,030.(6)
Age. On average, foster children are younger than kin-care children. A much larger proportion of foster care children are aged 4 or younger (34.2 percent versus 21.7 percent), and a much smaller proportion are aged 10 or older (41.4 percent versus 53.2 percent). However, although children 4 and younger make up a higher proportion of children in foster care than in kin care, there are about five times more children aged 0 to 4 in kin care than in foster care, 356,510 versus 73,342.(7)
Sex. In 1992-94, for both foster and kin-care children a slight majority (51.7 percent and 50.6 percent) were females. In contrast, for parent-care children, a slight majority (51.3 percent) are male.
Geographic distribution. Like children in kin care or parent care, foster children are more likely to reside in the South than any other region. Compared with kin-care children, they are more likely to live in the Northeast (20.2 percent versus 16.2 percent) or Midwest (25.2 percent versus 18.5 percent), and are less likely to live in the South (39.3 percent versus 46.8 percent) and West (15.4 percent versus 18.5 percent). Children in foster care are somewhat more likely than children in kin care to live in nonmetropolitan areas (29.4 percent versus 25.5 percent).
Caretaker characteristics
For all the characteristics of caregivers examined--marital status, age, educational attainment, and labor force status--foster parents fall somewhere between kin-caregivers and other parents (see Table 1.7).
Foster children are substantially more likely than kin-care children to be cared for by a married couple (66.6 percent versus 50.3 percent), although still less likely than parent-care children (72.4 percent). A sizable portion of both foster and kin-care children are cared for by widows (14.6 percent and 15.2 percent), only 1.0 percent of parent-care children have widowed mothers.
Foster mothers are older than other mothers, but younger than female kin-caregivers. The largest age group for mothers is 30-39, for foster mothers is 40-49, and for female kin-caregivers is 50-59. The same pattern is observed for foster fathers, other fathers, and male kin-caregivers.
Foster parents are better educated than kin-caregivers, but less educated than other parents. For example, most kin-caregivers (42.6 percent) are high school dropouts, but this group makes up a very small proportion of foster parents (15.2 percent) or other parents (14.2 percent). On the other hand, foster care parents are less likely to be college graduates than other parents (18.3 percent versus 26.8 percent), but are substantially more likely than kin-caregivers (8.1 percent).
Foster parents are more likely to be employed than kin-caregivers (71.6 percent versus 57.5 percent), but less likely than other parents (83.0 percent). Conversely, they are more likely to be out of the labor force than other parents (26.6 percent versus 12.4 percent), but less likely than kin-caregivers (39.1 percent).
Poverty status and use of services
Of the government programs we examined, foster families are less likely than kin-care families to receive benefits from all programs except unemployment compensation (see Table 1.8). The largest differences in receipt of benefits are children's use of free school lunches (31.1 percent versus 49.8 percent), receipt of food stamps (15.0 percent versus 31.2 percent), receipt of public assistance or welfare (11.0 percent versus 27.0 percent), and receipt of social security (23.2 percent versus 34.6 percent). The pattern of receipt of benefits for foster families is similar to the pattern for regular families whose children live with them, although foster families are more likely to receive social security payments (23.3 percent for foster families versus 6.4 percent for other families).
Foster families are substantially less likely to be below the poverty line than kin-care families (23.1 percent versus 38.8 percent). Their poverty status is similar to families whose children are at home (21.4 percent). On the other hand, foster families are similar to kin-care families in that very high proportions have no earned income (22.3 percent compared with 26.3 percent, contrasted with 9.2 percent for intact families). As Table 1.9 illustrates, kin-care and foster families with no earnings are equally likely to be below the poverty line--76.4 percent and 76.3 percent, respectively. The difference in poverty rates for the two groups is entirely due to differentials among families with earnings. Among kin-care families, 25.4 percent of those with earnings are in poverty, compared to only 7.8 percent for foster families. The explanation may be that foster payments--which are not tracked in the CPS--are enough to push most foster families with earnings over the poverty threshold.
1. 0 The characteristics of children in kin care are affected by two factors: the percentage of children in kin care for various groups, and the characteristics of all children. Thus, there will be in increase in the percentage of children in kin care who are black if the percentage of children in kin care increases, if the percentage of all children who are black increases, or both.
2. 0 Until 1988, the CPS did not identify whether children were living with their grandparents unless the children were also living with one or both parents. Children were identified only as "other relative" of the head of family.
3. 0 Coverage by Medicaid could not be evaluated because of inaccuracies in the CPS Medicaid coverage data.
4. 0 Child Welfare League of America (1995). Child Abuse and Neglect: A Look at the States. Washington, D.C.: Child Welfare League of America, Table 2.1.
5. 0 Including black Hispanics in the Hispanic group rather than the black group does not change this finding.
6. 0 Note that, as discussed above, 1994 data are not completely comparable to earlier years for point estimates. When 1994 is excluded and 1992 and 1993 data are used, the yearly average number of all black children in kin care is 650,275 and in foster care 93,325.
7. 0 When 1994 is excluded and 1992 and 1993 data are used, the yearly average number of all children 0-4 years old in kin care is 258,878 and in foster care is 71,719.
