Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

When looking at the complex societal and interrelated factors that impact a child’s access to the most fundamental rights—the right to be educated, to be safe, and to be healthy—one cannot overlook the children in sub-Saharan Africa and the current orphan crisis due to HIV/AIDS.

By the year 2015, 20 million children will have lost at least one parent to HIV/AIDS in sub-Saharan Africa alone (UNAIDS/UNICEF, 2011). The burden of caring for orphans in sub-Saharan Africa has followed four primary models, listed in order of prevalence: (1) care by the existing family structure; (2) care by neighborhoods, allowing children to remain in familiar environments; (3) care through enterprise-­centered collectives, formed by a consortia of government, nongovernmental, and private sector organizations, and modeled on traditional extended family (i.e., foster villages); and, finally, (4) institutional care (Sewpaul, 2001). Nontraditional methods of caring for orphans, such as institutional care and collectives, have not been viewed as viable solutions or as socially acceptable. As child care has historically been thought of in Africa as a social task performed by an entire extended family, adoption and care by strangers is deemed a last resort (Levine et al., 1994; Madhavan, 2004).

I have lived and worked much of the last 12 years in sub-Saharan Africa and have become increasingly interested in both the defended and undefended nature of the children of Africa. While high levels of HIV and the rising number of orphans have placed enormous stress on communities worldwide, this may be less true of Africa than any other region of the world, due to the dominance of the extended family over the nuclear family and the African cultural complex of socially distributed child care (Caldwell, 1997; Weisner, Bradley, & Kilbride, 1997). Weisner et al. (1997) described socially distributed child care as a set of loosely interwoven ecological characteristics, beliefs, and practices that coexist and contribute to one another. Some key components follow. Child caretaking often occurs as a part of indirect chains of support in which one child assists another, who assists another. Children look to other children for support as often as or more often than they look to adults. Care often occurs in the context of other domestic work. Aggression, teasing, and dominance coincide with nurturance and support—all from the same people. Dominance increases with age. Food and other material goods are used to threaten, control, soothe, and comfort. Mothers provide support and nurturance to children as much by securing others to support their children as by supporting their children directly (i.e., child fosterage).

Most orphans are being cared for through a culturally specific child care practice of child fosterage. In Africa, child fosterage has been described as a social welfare system revolving around kinship and defined as the rearing of a child by someone other than the biological parent (Bledsoe & Brandon, 1992; Brown, 2011). What makes fosterage unique is the semipermanent yet adjustable nature of the relationship, one of the most distinct elements of African families (Bledsoe, Ewbank, & Isiugo-Abanihe, 1988). In countries without institutional systems to better the welfare of children and the resources of families, fosterage serves a crucial role and may be a key component in African communities’ response to HIV/AIDS.

But what does it mean to be an undefended child? Some scholars have argued that the nature of being raised away from one’s biological parent places children at a disadvantage (Anderson, 2005; Foster, 2000). But such definitions have been tempered by more ethnographic accounts of children thriving in extended kin care. Thus, one cannot define being undefended simply as living without birth parents or not going to school; a dynamic mix of factors works in collaboration to defend the rights of the child.

From an academic standpoint, the study of child fosterage bridges the disciplines of anthropology and developmental psychology. For years, anthropologists have documented child rearing in diverse cultures and described child care patterns in Africa (Levine et al., 1994; Whiting, 1963; Whiting & Whiting, 1975). Developmental psychologists have stressed the importance of how we are raised, moving in and out of paradigms that stress biological or social influences as fundamental (Shonkoff & Phillips, 2000). In anthropological literature, much is written about child fosterage. Ethnographers began documenting fosterage, using different terms as early as 1937 (Herskovitz, 1937). Influential work by Goody (1973) and Bledsoe (1990a, 1990b) in Africa began to shed light on the variations of fosterage in Africa. Since then, scholars have studied motivations to foster (Isiugo-Abanihe, 1985; Payne-Price, 1981; Pennington, 1991) and predictors of fostering (McDaniel & Zulu, 1996; Vandermeersch, 2002). Fewer studies have examined the outcomes (Oni, 1995; Verhoff & Morelli, 2007). The research findings are contradictory regarding the effects of fostering and care by the extended family structure as the primary way of caring for orphans. Some scholars report uniformly negative effects for fostering, finding that children experience less education, more work, and less well-being (Bicego, Rutstein, & Johnson, 2003; Bledsoe, 1990a, 1990b). Other scholars regard the fostering system as stressed beyond repair (Foster, 2000). Still others hold that fostering is culturally appropriate, with built-in protective factors, and argue that orphans and non-orphans do not show significant disparity in developmental outcomes (Ankrah, 1993; Monasch & Boerma, 2004). These debates elicit questions regarding the family unit, the mother, and the individual child. Understanding how families negotiate fosterage in times of crisis may help explain some of the disparate results in treatment between fostered children, orphans, and biological children and individual developmental outcomes.

When addressing the impact of HIV on children, donors and governments in Africa have focused on orphan. Although definitions vary, an orphan is widely defined as a child under the age of 15 who has lost one or both biological parents (UNICEF, 2004).

Using UNICEF’s definition of an orphan as a child under age 15 who has lost one or both parents, over five million children in sub-Saharan Africa became orphans in 2003 alone. By 2010, sub-Saharan Africa will be home to an estimated 32 million orphaned children, more than two-thirds of whom will have lost one or both parents to AIDS.

Monasch and Boerma (2004) looked at national surveys from 31 countries in sub-Saharan Africa to assess, among other things, the impact of orphanhood on education and health. While school attendance varied by country and region, the odds of orphans attending school were less than for non-orphans. Double orphans have a greater disadvantage than maternal or paternal orphans. And, orphans have a lower level of education, regardless of overall attendance. Recent work in rural Zimbabwe showed that maternal orphans, but not paternal or double orphans, have lower primary school completion rates. For paternal and double orphans, a higher rate of school competition is attributed to the presence of an adult female in the household (Nyamukapa & Gregson, 2005).

This study addressed two primary questions. First, are Owambo orphans lower in developmental markers than non-orphans? Second, what is the Owambo cultural logic of fostering during times of crisis?

Methods

The Owambo Context

This study utilized the 2000 Namibian Demographic and Health Survey (DHS) collected from a random sample of all ethnic groups in Namibia and a qualitative case study. The four northern regions of Namibia (Oshana, Ohangwena, Omusati, and Oshikoto) are the traditional homeland for Owambo-speaking people and the setting for this study (Fig. 13.1).

Fig. 13.1
figure 00131

Four North Owambo-speaking regions of Namibia included in the household sample

Full size image

Among Oshiwambo-speaking people, seven Owambo language groups exist today. Owambo societies during the precolonial and colonial periods were predominantly matrilineal agropastoralist societies demarcated from each other by large areas of forest and savanna (Salokoski, 1998).

Kinship is an organizing principle in Namibia, holding more importance than class and playing a critical role in decisions regarding socially distributed child rearing (Hayes, 1998). What class does in advanced capitalistic societies like the United States kinship does in Namibia—it shapes peer relationships, choices about marriage, and choices with whom one can be raised. Kinship in Namibia is complex, with some groups classified as bilateral or unilineal; and among the Herero speakers, a double-descent system exists. Matrilineal descent systems are found among Owambo speakers (Hayes, 1998). Lebert (2005) studied inheritance of land, cattle, millet, and children among Owambos in the northern region of Namibia and described a traditionally matrilineal system. Upon the death of a man, the first order of inheritance is his oldest brother. If there is no oldest brother, then inheritance goes to his oldest sister’s oldest son, followed by his sister’s daughter’s oldest son. If a man has no siblings, inheritance goes to the oldest living male descendent of his mother’s sister. Upon the death of a woman, her children receive the inheritance—girls receive ornaments and jewelry, and boys receive the cattle.

Children traditionally belong to their mother’s family, and men do not pass on their matrilineal membership to children. The mother’s brother often plays a pivotal role in the care of the children, including providing care through fosterage. Even within this system, however, there is significant variation and complexity. Recent adaptations in the matrilineal inheritance system are due to the impoverished state of widows, as the husband’s matrilineal kin traditionally have rights to all of the wife’s possessions after his death. Similarly, the rising number of orphans has forced both paternal and maternal kin to raise children.

Namibian Demographic and Health Survey

The quantitative portion of the study utilized the 2000 Namibian DHS collected from a random sample of all ethnic groups in Namibia and focused on all children under age 19 in all households identified in the Owambo homeland (N = 5,949). The four northern regions of Namibia (Oshana, Ohangwena, Omusati, and Oshikoto) are the traditional homeland for Owambo-speaking people. All children whose mothers identified as Owambo were utilized (N = 4,030). For the qualitative portion, Oshiwambo families (N = 4) were also selected (see Fig. 13.3). The DHS is a nationally and regionally representative survey conducted by the United States Agency for International Development (USAID).

Foster children were defined in the DHS survey as children with both parents alive but residing with neither parent. An orphan was defined as a child under age 19 whose mother or father (or both) has died. A child whose mother is deceased was described in the data as a maternal orphan. A child whose father is deceased was described as a paternal orphan. If both are deceased, the child was referred to as a double orphan.

Variables

Child’s Residence Status. The DHS survey afforded two ways of discerning a child’s fosterage status. First, DHS data on Child’s Residence Status (fosterage and orphan prevalence) was collected by asking for each child the following: “Is [name]’s mother still alive?” and “Is [name]’s father still alive?” If the parent was still living, the household head was asked whether the parent currently resides in the reference household. Each head of household reported on all members of a household and their Relationship to the Head of Household (spouse, son/daughter, grandchild, sister/brother, other relative, foster/adopted child, non-relative). Second, all individual females aged 15–49 were asked about birth histories. Information on each child was recorded, including with whom the child currently lived, and was recorded as either “lives with respondent” or “lives elsewhere.”

Developmental Markers. The nutritional status of children under age 5 was measured by three standard indices of physical growth: height for age (which can be used as an index of stunting), weight for age (which can be used as an index of underweight status), and weight for height (which can be used as an index of wasting). Height was recorded in centimeters, and weight was recorded in kilograms. These indices were calculated using the Center for Disease Control (CDC) Standard Deviation-derived Growth Reference Curves derived from the National Center for Health Statistics (NCHS/CDC) Reference Populations, which sample international populations. Education was measured by two variables for children over age 5. First, attendance status was measured by asking the head of household if the child attended school in the previous year and in the current year. Attendance status was coded yes or no. Second, Education in Single Years was measured as a continuous variable.

Qualitative Data Collection

Multiple Case Study. A multiple case study of four Owambo families was conducted to better understand the cultural logic of fostering. In-depth interviews were conducted in Namibia, Africa, between September and November 2006 with four female heads of household.

Two sampling techniques were utilized. A purposeful sample, designed to intentionally select individuals to understand the central phenomena, was used. All families interviewed were interconnected through the practice of child fosterage. Maximum variation sampling, in which the research samples’ cases or individuals differed on some characteristics, was also used. Participants were initially recruited through contacts with women I knew well. I began with a female head of household whom I have known for over 10 years. From the information obtained about the children in the house, the second household was selected. If the family had fostered out a child, the recipient family was contacted and asked to participate in a similar interview. This process was repeated. The fourth family was not connected through child fosterage or kinship but was identified by the first family as a potential outlier. In line with recommendations for the number of cases to select in a multiple case study, four families were interviewed, as the goal of a case study is to gain in-depth understanding, not generalizability (Creswell, 2003; Stake, 1995).

An initial visit was made to ask if the family would be interested in participating, and if so, additional meeting times were scheduled. I utilized some elements of postmodern anthropology theory when conducting the case studies. In an attempt to be self-reflexive and remove the dominant position of the researcher, I disclosed to participants my own experiences of being adopted in the United States, being a mother myself, and my own experiences of living in Namibia previously and returning to do research on child fosterage. I encouraged participants, if questions arose, to stop and ask about my own experience.

Interviews lasted between 2 and 4 hours. Three of the interviews were conducted over multiple days. The fourth interview was conducted during a daylong visit. All interviews were tape-recorded and translated if needed. Participants were invited to speak in their language of choice, either Oshiwambo or English. I encouraged participants to switch between languages if it made explanation easier. All interviews were conducted in English, with passages in Oshiwambo. Participants were paid 50 Namibian dollars for participation. This is equivalent to seven US dollars.

Qualitative Data Analysis. A case study is a type of ethnographic design (Stake, 1995) and is an exploration of a “bounded system” or a case over time, through detailed, in-depth data collection involving multiple sources of information and rich in context (Merriam, 1998). In this study, the multiple cases served the purpose of “illuminating a particular issue” (Creswell, 2003)—namely, what is the cultural logic of fostering during crisis?

Analysis of the case study followed Stake’s (1995) technique and occurred in three phases. I performed the initial coding and thematic analysis. First, a detailed description of each family was created from filed notes, observations, and information gathered during the interviews. All interviews and field notes were typed for analysis. Second, thematic analysis was performed on two levels: within each case and across the cases (Stake, 1995). Thematic analysis of the interview data was done by (1) initially reading the data for overall understanding and writing preliminary notes, (2) partitioning segments of text and labeling them with codes, (3) aggregating similar codes together to develop themes, (4) connecting and interrelating the themes, and (5) constructing a narrative. The interconnected themes were of particular interest to the study, as the relationships that exist between families are subjects of the central questions of the study. The final phase of analysis was to integrate the cases and themes and report on the “lessons learned” or assertions put forth from the study (Stake, 1995).

Results

Information analysis about households, mothers, and children in Namibia revealed a demographic picture of family life in Namibia. Owamboland has the highest rates of children in Namibia who do not live with either parent but whose parents are both alive. Rates range from 29.4 to 36.9 % in the four northern regions and from 11.5 to 28 % in the rest of the country. Rates of living with both parents are much lower, ranging from 13.6 to 18.4 % in the northern regions and from 23.0 to 48.2 % in the rest of the country. Table 13.1 shows demographic information about Owambo children, mothers, and households.

Table 13.1 Demographic information on Owambo children, mothers, and households in Namibia
Full size table

Table 13.2 shows the height and weight of Owambo children. Owambo children under 5 years old averaged 10.8 kg and 84.5 cm. Height and weight percentiles were calculated for all children under 5 years of age using the CDC Standard Deviation-­derived Growth Reference Curves derived from the NCHS/CDC Reference Populations. Owambo children registered in the 19th percentile on average for weight and in the 25th percentile for height. Their weight-for-height average percentile was 29.9, meaning children on average compared to other children internationally are in the 29th percentile on weight-for-height measures.

Table 13.2 Mean level of education and height and weight of Owambo children in Namibia
Full size table

Univariate statistics on all outcome variables used in the study are presented in Table 13.2.

From the household interviews, each person’s relationship to the head of the household was gathered. Table 13.3 shows the number of Owambo children and their various relationships to the head of household. Biological children made up 36.8 % of the children under 19 years old living in households, 41.5 % were grandchildren to the head of household, 13.9 % were other relatives, 2 % were adopted or fostered children, and 4 % were not related to the head of household.

Table 13.3 Distribution of Owambo children’s kinship relationship to head of household
Full size table

The relationship that maternal and double orphans had to the head of household was explored. The majority of maternal orphans (N = 160) lived with a grandparent (64 %), although it is not known if it was the maternal or paternal grandparent; 17 % were living with other relatives; and 7 % were living with the father. For double orphans (N = 44), 68 % lived with grandparents, 16 % lived with other relatives, and 9 % lived with non-relatives. Table 13.4 displays the results.

Table 13.4 Relationship to the head of household of Owambo children who are maternal and double orphans
Full size table

To test the main research question—Within the fosterage system, does being orphaned affect developmental markers of health and education?—a series of 2 × 2 factorial ANCOVAs, controlling for age of the child, were used to examine adjusted mean differences in education and health between male and female orphans and non-orphans. Maternal orphans under age 5 (N = 18) and between ages 5 and 19 (N = 161) were defined as having mother not alive but father still living. Double orphans under age 5 (N = 2) and between ages 5 and 19 (N = 44) were defined as having neither mother nor father alive.

Years of education for male and female maternal and double orphans, as well as non-orphans, are summarized in Table 13.5. There was not a significant interaction between gender and maternal orphan status (F(4,2779) = 1.64, MSE = 2.24, p = 0.20). There was no main effect for maternal orphan status (F(4,2779) = 0.519, MSE = 1.15, p = 0.47). There was, however, a significant gender main effect (F(4,2779) = 21.31, MSE = 47.75, p = 0.001), with orphaned girls having significantly more years of education than orphaned boys. The effect size was r = 0.18 (see Fig. 13.2).

Table 13.5 Mean differences in years of education for maternal and double orphans and non-­orphan boys and girls
Full size table
Fig. 13.2
figure 00132

Education in single years for male and female maternal orphans, double orphans, and non-orphans

Full size image

Similar patterns were found with double orphans. There was not a significant interaction between gender and double orphan status (F(4,2779) = 0.09, MSE = 0.212, p = 0.76). There was not a significant main effect for double orphan status (F(4,2779) = 0.64, MSE = 1.43, p = 0.42). However, there was a significant main effect for gender (F(4,2779) = 4.60, MSE = 10.32, p = 0.03), with girls having significantly more years of education than boys. The effect size was small at r = 0.1 (see Fig. 13.2).

Whether male and female maternal and double orphans attended school in the current year was examined, and adjusted mean differences are presented in Table 13.6. There was no significant interaction between gender and maternal orphan status (F(4,2764) = 0.319, MSE = 0.028, p = 0.572). There was no significant main effect for maternal orphan status (F(4,2764) = 0.470, MSE = 0.041, p = 0.493) or gender (F(4,2764) = 1.70, MSE = 0.149, p = 0.192).

Table 13.6 Mean differences of school attendance in current year for maternal and double orphans and non-orphan boys and girls
Full size table

Similar results were again found when comparing school attendance of double orphans to non-orphans. There was no significant interaction in school attendance between double orphan status and gender (F (4,2764) = 0.012, MSE = 0.001, p = 0.912). There was no main effect for gender (F(4,2764) = 0.634, MSE = 0.055, p = 0.426) or for double orphan status (F(4,2764) = 0.285, MSE = 0.025, p = 0.593) (see Fig. 13.3).

Fig. 13.3
figure 00133

School attendance for maternal orphans, double orphans and non-orphan boys and girls

Full size image

Height and weight percentiles were calculated for maternal and double orphan boys and girls and compared with non-orphans. Because of the small number of double orphans (N = 2), only maternal orphans were used in this analysis. Adjusted mean differences are found in Table 13.7. There was no significant interaction in height between maternal orphan status and gender (F(4,1293) = 0.755, MSE = 581.34, p = 0.385). There was no main effect for maternal orphan status (F(4,1293) = 1.85, MSE = 1424.66, p = 0.174). There was no main effect for gender (F(4,1293) = 1.13, MSE = 872.47, p = 0.287) (see Fig. 13.4). Likewise, when examining weight percentiles, there was no significant interaction in weight between maternal orphan status and gender (F(4,1293) = 0.45, MSE = 269.32, p = 0.505). There was no significant main effect for maternal orphan status (F(4,1293) = 0.182, MSE = 109.78, p = 0.67) or for gender (F(4,1293) = 0.93, MSE = 559.69, p = 0.336) (see Fig. 13.4).

Table 13.7 Mean differences in height and weight percentiles for maternal and double orphans and non-orphan boys and girls
Full size table
Fig. 13.4
figure 00134

Height and weight percentile for maternal orphan and non-orphan boys and girls

Full size image

The quantitative analysis reveals that orphans do not seem to be undefended in terms of education and height and weight. While nonsignificant, orphans often have higher raw scores in years of education and probability of attending school. Gender is the one characteristic that may explain the divergent experiences of orphans, with boys having fewer total years of education. In this case, null results are hopeful for Owambo children. Several questions come to mind when incorporating these findings with the existing literature from other communities in Africa where orphans are experiencing significantly worse conditions than non-orphans.

How families negotiated fosterage arrangements, specifically when biological parents were ill, was an important theme of the study. Different means of negotiation existed for the donor family and the recipient family. In these negotiations, children were requested by the receiving family or given by the donor family.

Early arrangements were honored when negotiating where children would live, both in and out of times of crisis. For example, if a child had lived with a maternal grandmother since she was age 4, the preference among both maternal and paternal family members at the time of a parent’s death was to leave the child where she was. Ndeleo explains:

It depends where the child has been brought up. When they are young, if the child is with the father then he has a right to the child. If the child stays with the mother, then she has more rights. It depends on who brought up the child.

In times of crisis, living arrangements for children were collectively made before the mother died. In this way, the culturally normative practice of fosterage was evoked to make the transition to a new home easier by preventing children from witnessing the parent’s death. However, siblings were often split. While all homes in the study were involved in caring for orphans, only one family fostered two siblings.

Magdelena, Nangula’s only fostered child, came to live with Nangula when Magdelena’s mother fell ill. Magdelena’s biological mother had previously lived with Nangula and helped her care for her two boys and do domestic chores.

When I came to Eenhana, she (Magdelena’s mother) had a baby and then a second baby and then became sick. She went back to Ombalantu to stay with her parents. She was tested (for HIV) and the first one was positive, a baby boy, and he passed away. She then wrote me a letter and said she is sick and said I can take Magdelena to care for her. My husband was working in Arandis and I tell him, but then another letter came and repeated that I was supposed to take the child. She is now my kid. I can take her with me. Then I go there with my husband and we pick up Magdelena. In only two years, she (mother) passed away.

Two examples were found of children negotiating living arrangements themselves. While this was incongruent with cultural rules, it mimicked the practice of voluntary fosterage. In one instance, a 17-year-old girl (Aune) was a second-­generation orphan who chose to live with the Nakele family. Second-generation orphans are children who have been placed in a home before or at the death of a parent, and then the caregiver of the new home also dies. Usually, these children are more removed from close kin than is typical of fostered children. Therefore, the possibility of being absorbed into a new family diminishes with each death of a caregiver. Meme Nakele explains how Aune came to live with them.

She was not living with her mother when she died. When she was three, right straight after breast feeding, her grandmother took her from the mother and kept her. Her grandmother was working as a cleaner in the hospital when she became ill and died. Then Aune is staying in the location with a distant relative.

Aune encountered conflicts with the people staying in the house with her and was locked out of the house without a place to stay. So she went to the Nakele house.

We thought she was just coming to visit but after one, two, three days we then asked, “Aune when are you leaving?” Aune said, “No, Meme, I am not leaving.” I said, “For what good reason?” I couldn’t say go back so I said, “Wait until Tate comes home and we will sit and talk because you can’t just come and stay with someone without informing them.” I know she is a true orphan and that is why we allow it.

Women explained that at times, “You can do no other than take the child,” as no alternatives exist. Women expressed worry and concern as motivations that prompted these relationships. Being fostered in times of crisis did not have the same advantages to the recipient family as being fostered voluntarily, and children often did not have advocates or adults to hold the recipient family accountable. Hilde, an orphan from the sister of Meme Nakele, had her maternal grandmother intervene and move her to the Nakele house. After her mother’s death, Hilde was first placed with her mother’s sister. The woman’s husband was not happy with all the orphans in the house, and word got back to the maternal grandmother. Meme Nakele described:

So my mother heard this bad talking of Martin (husband) and she came to us and asked us to take Hilde. At the time my sister had taken all the kids to Hamutenya’s house and I said, “Oh, they have too many children and Meme is old.” There is no other place to stay so Meme came to us to see if we can take her.

Grandparents acted with vigilance when they heard of an abusive fostering arrangement, often negotiating other living arrangements for young children without birth parents.

Discussion

The questions addressed by this study asked if children fostered in times of crisis (orphans) differed from biological children in education and health markers. There was no difference found in education or health between orphans and biological children. The findings suggest that orphan status, alone, does not have the impact on education and health in Owamboland that it has been found to have in other communities. Past research has found differences in education between orphans and non-orphan foster children in Zimbabwe (Foster, Makufa, Drew, Mashumba, & Kambeu, 1997) and between orphans and non-orphans in Uganda (Kamali et al., 1996). Monasch and Boerma (2004) combined DHS data from all Southern African countries (South Africa, Botswana, Namibia, Zimbabwe, Zambia, Lesotho, and Swaziland) and found that orphans were less likely to attend school than non-­orphans. Both maternal and ­double orphans in this study had nonsignificant but higher rates of school attendance than non-orphans, with female orphans having the most education.

I am not suggesting that Owambo orphans are not suffering and do not need to be considered and defended. The case study revealed instances where children were struggling to secure shelter. However, results from the case study may help illuminate why orphans and non-orphans do not differ in education and height and weight from non-orphans. In the study, care for orphans was collectively negotiated prior to the death of a parent. The indigenous system of child fosterage and care by extended kin is defending Owambo children on basic levels.

Gender and kinship may also play a role in the null findings. Both maternal and double orphans primarily lived with their grandparents. Grandparents may have defended these orphans by providing education and access to food with more vigilance than a more distantly related kin would have. Again, both fostered and biological males were lower in education than were females. This poses questions about the gendered nature of work and how traditional gender roles may shape the experiences of Owambo children.

What remains a crucial question is not whether children orphaned by AIDS have psychological and circumstantial problems, but whether the current system of absorbing, nurturing, and caring for these children is meeting children’s basic developmental needs. There is no doubt orphaned children anywhere experience psychological trauma, but within a system that possesses indigenous practices and knowledge of shared childrearing, do these children have a better prognosis? Among Owambo-speaking people in Namibia, kinship relationships, more than orphan status, affected children’s developmental markers, and these markers interacted with gender almost always to the disadvantage of boys (Brown, 2009). Orphans did not differ from non-orphans in terms of education and health, unlike in other communities across Southern Africa. This study illustrated that it is the interaction between context and developmental practices that was crucial, not the specific practice. For example, because fosterage was a common practice in many African communities, there was not the social stigma attached to this practice that one might see in the United States. On the contrary, fosterage provided a culturally intuitive mechanism for successfully managing parental absence.

Limitations of the Study

Limitations existed for the current study. First, the Namibian DHS data is now nearly 13 years old. A new wave of data collection is scheduled in 2012. This new data will be able to show, more accurately, how the burden of Namibia’s orphans is being addressed by families in Namibia. As the HIV/AIDS epidemic increases, the burden of orphans increases, though at a slower rate. More recent data may provide quite different results.

Second, because ethnicity information was not available for the household survey, the four northern regions (known as Owamboland) were chosen. This sampling procedure excluded the children fostered to the capital of Windhoek. As seen in the case study, fostering to the capital for better educational opportunities was not only common but important. These households were excluded from the study. While sampling the four northern regions allowed a focus on the Owambo ethnic region, it most probably underrepresented the whole picture of child migration and fostering among Owambo speakers in Namibia.

Future Directions

Is the extended family system stressed beyond the point that children’s developmental needs are being met? This becomes a crucial question if we accept that there is a limit to the capacity and acceptability of existing family kin networks, and that many of these children are likely to be emotionally, behaviorally, and physically compromised. The practice of fosterage exists, at least in part, to extend kinship and create cultural and social ties. Thus, this study attempted to understand fosterage not only from a sociobiological perspective but also from a cultural perspective.

Who will care for these children in the future if projections are correct that 20 million children will be orphaned by 2015? I believe that using what we know about child fosterage and the current experience of orphans, a developmentally informed approach could be better utilized when planning and providing support to orphans and vulnerable children. Keeping sibling groups together, not overlooking the experiences of boys, and finding close kin to provide care are important, but so are the quality and dynamics of the relationships between donor and recipient families. Developmental knowledge can help reshape the orphan crisis by connecting culture, the family, and the child. John Caldwell, a demographer who has written extensively about HIV, believes that Africa is uniquely poised to handle the HIV crisis (Caldwell, 1997). This study affirms his beliefs and provides some description of how the cultural ecology of childrearing is defending those children thought to be the most defenseless.