“LIFE IS NOT PAP AND VLEIS”: POVERTY IN CHILD-HEADED

INTRODUCTION Richter (2004:9) suggests that child-headed households (CHHs) are much more vulnerable economically than adult-headed households. “Income in orphan households has been found to be 20-30% lower than in non-orphan households” (Richter, 2004:9). This is confirmed by Donald and Clacherty (2005). Children in child-headed households survive on about one-third of the resources (money as well as contributions in kind, such as gifts or food, etc.) available to adult-headed households. This is because these children lack the presence of parents, they have limited means of generating an income and they are unable to effectively sustain their households (Donald & Clacherty, 2005:24).


INTRODUCTION
suggests that child-headed households (CHHs) are much more vulnerable economically than adult-headed households. "Income in orphan households has been found to be 20-30% lower than in non-orphan households" (Richter, 2004:9). This is confirmed by Donald and Clacherty (2005). Children in child-headed households survive on about one-third of the resources (money as well as contributions in kind, such as gifts or food, etc.) available to adult-headed households. This is because these children lack the presence of parents, they have limited means of generating an income and they are unable to effectively sustain their households (Donald & Clacherty, 2005:24).
A research project was therefore commissioned by the Gauteng Department of Social Development (DSD) to determine the prevalence of child-headed households in Gauteng in order to establish a database and to ensure access to aid programmes by needy child-headed households. The authors of this article formed part of the research team that conducted this study on behalf of the DSD, under the auspices of Chiastolite Professional Services.
The objective of this paper is to describe the socio-economic conditions of child-headed households in Gauteng, so as to measure the extent of poverty among these children. In order to do this, the study includes an assessment of the size and sources of the monthly income of child-headed households in Gauteng. The contribution by the different sources of income to total monthly income is also analysed. Special attention is paid to the contribution of income generated by household members themselves, as well as to social grants.
The data on income received are then further analysed in terms of poverty criteria, so as to determine whether the monthly income of child-headed households conforms to the minimum income necessary for material survival. The monthly income is compared to the different amounts required for material survival by different family sizes in the applicable urban centres in South Africa.

Literature review
Different organisations and communities adopted different definitions of what constitutes a child-headed household. Many organisations recognise a household as being child-headed if the head is still attending school, regardless of the child"s age (Gauteng DSD Report, 2008:xiii). The official definition contained in the Children"s Act No. 38 of 2005 defines a child-headed household as a household headed by a person under the age of 18. It includes situations in which these children need to take care of a terminally ill adult living in the same home (Republic of South Africa, 2005).
In this research project the definition used by the Gauteng Department of Social Development was adopted. In this definition a CHH is described as being composed of "orphaned, abandoned or neglected children who live in a household in which the oldest member is under the age of 18 with no adult supervision and support" (Gauteng DSD Report, 2008:5). A child-headed household is therefore a household consisting only of children, with no adult living under the same roof. The older children in such a household accept the responsibility of supporting themselves and their siblings in terms of the basic needs of the household. Children who have access to adult supervision, or who were looking after elderly or sick adults, were therefore excluded from this research project.
Information on the extent of child-headed households in South Africa and in the rest of Africa is limited. There is no reliable data for child-headed households in South Africa. It is difficult to capture data on these households for a variety of reasons.
Different definitions of child-headed households are used by different institutions (Gauteng DSD Report, 2008:24). Child-headed households are frequently merely temporary households and often exist for only a short period of time (Meintjes & Giese, 2006). In the context of the legal definition, households headed by children aged 17 will only be recognised as a child-headed household for a short period of time, because when that child turns 18, the household is no longer regarded as a child-headed household. Welfare organisations indicated to the research team that, as soon as a household is identified as being child-headed, the children are legally placed in a place of safety or in a foster home (Gauteng DSD Report, 2008:24). There is also extensive evidence that the majority of orphans are indeed taken into the care of their extended families and are then not regarded as constituting child-headed households (Gauteng DSD Report, 2008:26). The available statistics on child-headed households should be viewed against the background outlined below.
According to the South African Institute of Race Relations (SAIRR, 2009:1), the number of children in child-headed households in South Africa increased by 25% from 118 000 in 2002 to 148 000 in 2007. Of the total number of 148 000 children in child-headed households, 146 000 were black. Between 2002 and 2007 some 2 000 coloured child-headed households, which had previously not existed, emerged. Very few, if any, child-headed households were Indian or white (SAIRR, 2009:1). Richter and Desmond (2008) estimated that in 2005 approximately 0.7% of the population, or 120 000 children between the ages of 0 and 17, were living in child-headed households, while 0.6% of all households in the country were child-headed households (Richter & Desmond, 2008). A total of 70% of child-headed households in South Africa comprised only one child, while 90% were headed by children older than 15 but younger than 18 years (Richter & Desmond, 2008). In the same year the majority of children in child-headed households were male, with 64% headed by males and 78% of children living alone being male (Richter & Desmond, 2008). The percentage of children in child-headed households who are male increased from 59% in 2002 to 64% in 2007 (Richter & Desmond, 2008;SAIRR, 2009:1).
From these data it is clear that the prevalence of CHHs is increasing and that not all abandoned children will necessarily be absorbed by their extended families. This necessitates the development of support measures to assist these CHHs. Hulley (2006) and Smart (2003) emphasise financial assistance and protection of children"s rights as the most urgent socioeconomic needs of CHHs. Failure to provide this increases the children"s socio-economic vulnerability and exposes them to exploitation in the form of child abuse, child labour and even human trafficking.
The root cause of children"s socio-economic vulnerability stems from widespread poverty, and specifically child poverty, in South Africa.

Measuring child poverty
The level of child poverty in South Africa is extremely high. In 2005 two-thirds (11.9 million) of children in South Africa lived in households that had an income of R1 200 per month or less (Meintjes, Leatt & Berry, 2006:1). The authors hypothesise that many of the child-headed households in Gauteng would also fall into this category of poor families. This hypothesis will be tested by means of an analysis of the income data of child-headed households in Gauteng.
The prevalence of child poverty differs across South Africa. According to the child-headed prevalence study of Statistics South Africa (2008c:69), Limpopo had the highest rate of child poverty in 2005 at 83%. The Eastern Cape Province followed closely at 80%. The following provinces displayed higher rates of child poverty than the national average: KwaZulu-Natal, Mpumalanga and the North West (Statistics South Africa, 2008c:69).
It is difficult to accurately estimate poverty levels, as there is no general or universal measurement of living standards required for satisfying people"s basic needs. The concept of basic needs can also be defined in a number of ways, depending on the quality or standard of clothing, food or other requirements for subsistence purposes (Barker, 2007:112).
Internationally, comparisons of poverty levels usually use a rudimentary income poverty standard, developed by the World Bank. The World Bank uses national poverty lines for 33 countries, with an international line, arrived at as the median of the ten lowest poverty lines in question. The poverty line in question is equal to $1.08 (currently approximately R7 to R8) per person per day, measured in terms of 1993 purchasing power parity (PPP). This was referred to as the "US$1 a day" standard. The actual amount was later adjusted to $1.25 and an upper poverty line of $2 per day is occasionally utilised in middle-income countries (Mohr, 2010:166). As such, it broadly coincides with the poverty lines used in lower-middle-income countries.
The above indicators are useful as indicators of global progress in the reduction of poverty and for cross-country comparisons. They are, however, not necessarily the appropriate criteria to use within a specific country (Statistics South Africa, 2007:5).
The normal way of calculating national poverty lines as a statistical yardstick is to estimate the cost of a minimum basket of goods that would satisfy the minimum daily energy requirements per person over the period of one month. The South African Medical Research Council (MRC) regards the recommended daily energy requirement as 2 261 kilocalories per person (Statistics South Africa, 2007:7). Statistics South Africa estimated that, with the type of food typically available to low-income South Africans, it will cost R211 per person to satisfy this monthly energy requirement. This was calculated in terms of 2000 prices, using the 2000 Income and Expenditure Survey data (Statistics South Africa, 2007:7). Statistics South Africa (2007:10) attempted to approximate the non-food component of a national poverty line. It assumed that these non-food items normally purchased by a household, who spend approximately R211 per capita per month on food, could be regarded as vital. Households usually have to sacrifice expenditure on food to acquire these non-food items. The price tag of such indispensable non-food items amounts to R111 per capita per month.
Totalling these figures (R211 plus R111) provides an estimation of the minimum cost of both essential food items and non-food consumption per capita per month. The resultant poverty line is therefore R322 per capita per month in terms of 2000 prices. This yields a poverty line of R431 per person in terms of 2006 prices (Statistics South Africa, 2007:10). Using the consumer price index (CPI) figures for 2007, and the available CPI figures for 2008, this amount can be expressed as R517.92 per capita per month, in terms of 2008 prices. This figure is more than double the amount of the current child care grant of R240 per month per child.
The research results on the socio-economic conditions and poverty levels of CHHs in Gauteng are investigated and discussed against this background.

RESEARCH PROCESS AND METHODOLOGY
At the time of the research there was no accurate database for CHHs in Gauteng. As a result of the lack of a comprehensive database of existing CHHs, an exhaustive research process had to be followed. This process consisted of three distinctive stages.

Stage 1: Identification and capturing of CHHs on a database
Stage 1 commenced in February 2008 and was completed at the end of March 2008. The purpose of Stage 1 was to identify CHHs and to compile a reliable contacts database. To trace the CHHs in Gauteng, the province was divided into geographical areas of equal size according to existing street maps. In each area the local community was mobilised via the mass media, community walks, posters and pamphlets. Close liaison with local government structures, such as clinics, schools, welfare organisations and churches, provided the necessary information on child-headed householdseither directly or via service providers and key stakeholders. At the end of Stage 1 more than 6 000 households believed to be CHHs were captured on the database. It later turned out that only a handful of these households were in fact child-headed in terms of the definition used by Gauteng DSD (Gauteng DSD Report, 2008:5).

Stage 2: Verification stage
The verification stage was implemented from April to June 2008. It entailed door-to-door visits by 66 well-trained field-workers to all 6 000 households. The field-workers were recruited by regional coordinators and trained by one trainer to ensure a standardised quality and content training of all field-workers. Each field-worker received a list of contacts for the allocated geographical areas. These lists were produced via the database of contacts that had been compiled during Stage 1. At this stage a brief demographic questionnaire was used to verify whether these households were indeed child-headed in terms of the operational definition required by the Gauteng DSD, which was also used for this research project. The majority of households (76.7%) reported by agencies as being child-headed were in fact adult-headed. These CHHs were often headed by a child older than 18, or had some form of adult supervision. Only 63 of the 6 000 households visited satisfied the definition used for the research project as being child-headed households.
The regional coordinators screened all completed data-collection instruments for completeness and legibility, so as to ensure that the database to be used during Stage 3 was as accurate as possible. Problems identified during this stage, such as missing data and incorrect or inexplicable codes, were followed up with the field-worker concerned and additional training was provided to that field-worker. These actions ensured quality data collection.
Data collected during Stage 2 were captured on a Microsoft Access database (CHHs database) by two fully trained data capturers. Ten percent of the captured data was checked against the original questionnaires to ensure the accuracy of the data capturing. Descriptive statistics were used to draw up a profile of child-headed households in Gauteng. A number of crosstabulations were performed to investigate the relationship between different variables.

Stage 3: Research design
The third and last stage entailed a combined quantitative-qualitative research design. The quantitative and qualitative information was collected through a structured questionnaire. The qualitative research was phenomenological in nature, as it studied the meaning of experiences for the research population (Fouché, 2007:270).
The research instrument was designed to accommodate the in-depth combined qualitative/quantitative design. The data collection instrument was developed by the research team at Chiastolite Professional Services, based on a service request from, and approved by, the Gauteng DSD. The questionnaire went through several versions before the final one, which was used for the research project, was adopted. It was subjected to extensive internal reviews. A pilot study was done by four Unisa students in Social Work. Based on the pilot study, several small adjustments were made to the instruments.
A total of 61 out of the 63 households participated anonymously and voluntarily in this research. The interviews were conducted only with the head of the household. The project coordinator trained a team of seven field-workers. This team comprised selected field-workers who had proved their interviewing and recording skills during Stage 2.
Interviews were not tape-recorded. Regional coordinators screened all completed instruments to ensure that they were comprehensively and legibly completed. The instruments were then submitted for capturing. The data were captured using appropriate software. The qualitative data were analysed according to themes that emerged from the research.

Challenges and limitations of the design
One of the key challenges facing this research project was the myriad of different definitions of exactly who constitute(s) a child-headed household. The impact of this was already felt during the planning stage of the research. The researchers were under the impression that organisations, such as welfare agencies, had existing databases on child-headed households. This assumption proved to be wrong. Although a large number of vulnerable families were reported, some of the available data were outdated. In many cases the agreed-upon definition of a child-headed household was not met. Related challenges were encountered in that duplication on existing databases was discovered during the research process (Gauteng DSD Report, 2008:16).
After verification visits by field-workers, many households identified as child-headed households were either removed or family members moved in with them (Gauteng DSD Report, 2008:17). As a result many households that had previously been entered into the database as child-headed households had to be excluded from the survey, as they no longer met the agreed-upon definition or criteria.

Ethical considerations
It was quite clear from the outset that this research project focused on an extremely vulnerable group of people. The research subjects were not only children, but indeed children without parental care. Any research involving minors usually requires the informed consent of parents. In this research project the child heading the household was duly considered to be the head of the household, since he or she carried the responsibilities usually borne by an adult. The head of the household"s consent to participate in the research was therefore accepted (Gauteng DSD Report, 2008:14). It was made clear to respondents that they could terminate the interview at any stage. The principle of voluntary participation was thus adhered to (Louw, 2007:86).
In order to adhere to the "doing no harm" principle of research, field-workers were trained to identify themselves by displaying cards and wearing T-shirts from the research agency (Bless & Higson-Smith, 1995:102;Rubin & Babbie, 1997:59-63). Furthermore, all field-workers had been screened and expertly trained by an experienced trainer before the field-work commenced. This ensured consistency in the handling of the interviews.
The concepts of confidentiality and anonymity together represent a further important ethical principle that can prove to be problematic during survey research (Bless & Higson-Smith, 1995:103). In order to take this ethical principle into account, the sample surveys were done anonymously. Along with confidentiality and anonymity is the issue of deception of respondents. This implies that there should be no deliberate misuse of facts. Therefore, data must be used only for the stated purpose of the research (Louw, 2007:87). This principle was also adhered to at all times during the course of the research design and data-gathering process. No false promises were made to the respondents about how they would benefit from the study"s findings. The authors are therefore confident that the highest level of ethical conduct was adhered to throughout the research project.

Basic demographic characteristics
The 63 households verified as being child-headed in terms of the definition used in the research project comprised 107 children.
Most of the child-headed households surveyed in Gauteng are small. Just less than half (43%) of the child-headed households comprise only a single child, which is far lower than the 70% child-headed households for South Africa as a whole, as indicated by Richter and Desmond (2008) in the literature. A further 39% of child-headed households comprise only two children, and only 18% of the households consist of more than two children. About 48% of household heads are aged 17, while the youngest head is a mere 12 years old. These child-headed households have been child-headed for an average of 1.42 years, with a range of 83 days to 5.47 years, which is in line with the findings as discussed in the literature.
Households comprise primarily family members, and especially siblings. Some child-headed households also include cousins and, in some cases, the children"s own children. Although the overall ratio of boys to girls in the child-headed households is equal, two thirds of household heads are boys, suggesting that families and communities may perceive boys as being more capable of looking after themselves than girls are.
Approximately half the children in this study were orphans and reported that their parents had died. The other half reported that their parents were alive but living elsewhere, suggesting that they might have been abandonedeither permanently or temporarily. The parents in question might well be migrant and domestic workers.
Informal financial and other forms of support from family members are received infrequently. Support from siblings, neighbours, teachers and friends are provided more frequently. The threat of insecurity, poverty and the demands of survival are very real for many of these children. They continually struggle to secure the basic resources required to survive. The size and sources of the income of these child-headed households are discussed and analysed in the next section. Figure 1 displays a standard distribution diagram, featuring the total monthly income of childheaded households. The total monthly income received by the child-headed households surveyed in Gauteng, as estimated by the respondents, ranges from no income to a maximum of R3 000 per month. Based on the data, the average monthly income of child-headed households amounts to R1 121 per month and the median is R1 000 per month. Two thirds of the CHHs received an income below R1200. Almost half or 46% of the CHHs received an income below R1 000 per month. Most alarming is the fact that 19.7% of the CHHs" income falls into the category R0-R199. This represents no less than one in every five CHHs. An amount less than R200 per month is indeed lower than any accepted poverty line or minimum subsistence level. These CHHs are therefore living in abject poverty in terms of both the absolute and relative definitions of poverty (Mohr, 2010). Figure 1 shows that there seem to be fairly clear threshold levels for the monthly income for CHHs in Gauteng. The total monthly income for a CHH in 2008 was either below R1 800 or above R2 000 per month. The chosen intervals of R200 for the distribution therefore seem to indicate definite threshold levels. This needs further investigation and clarification.

FIGURE 1 TOTAL MONTHLY INCOME OF CHILD-HEADED HOUSEHOLDS
A breakdown of the different sources that make up the average household income of R1 121 per month is presented in Table 1. This distinction is necessary in order to obtain more comprehensive and reliable results (Lighthelm, Martins & Van Wyk, 2000:53). The average contribution of grants towards the total average income of R1 121 is R233. Family members contribute on average R384, which represents the second-largest source of income for CHHs. Income generated by the household members themselves is the largest source of income.   not be economically active as part of the labour force, but rather in school increasing their investment in human capital. However, these are not normal families, but households where the head or responsible person for the welfare and survival of the household is in fact a child. The statistics therefore emphasise the desperate state of affairs for these CHHs. Family members contributed 31.9% to the monthly income of CHHs.
The contributions from grants towards the gross monthly income of child-headed households in Gauteng amount to more than triple the contributions made to the gross annual income to households for South Africa as a whole. For the average South African grants represented only 6.1% of gross annual household income during 2005/06, as compared to 19.4% of the gross monthly income of child-headed households in Gauteng during 2008. Social security therefore plays an important role in the income of child-headed households in Gauteng.
Grants available to support children are the Child Support Grant, the Foster Child Grant and the Care Dependency Grant. These grants can play a major role in supporting child-headed households financially, but the rules to access these grants are strict (Maqoko, 2006:40).
The Child Support Grant is meant for children living in poverty in South Africa. Most childheaded households comply with these criteria. The grant is paid to the person responsible for a child"s primary care, whether it is a parent, a relative or a member of the community who is not related to the child. The grant is, however, only available to children younger than 14 years of age. As from 1 October 2008 the Child Support Grant stood at R230 per month, and as from 1 April 2009 it increased to R240 per month (Gauteng DSD Report, 2008:1). Children heading a household find it extremely difficult to access these grants, since they have to be collected by a responsible adult (Sloth-Nielsen, 2004:30-31).
The Foster Child Grant is meant to benefit children who have been formally placed in the care of foster parents by the Children"s Court. The court may grant community-based caregivers permission to care for children in child-headed households. A report from a social worker has to accompany such applications. The grant amounted to R680 per child per month in 2009, which might not cover all the costs of caring for orphans (South African Government Services, 2009a:1). The Foster Care Grant helps somewhat towards alleviating poverty among orphans in child-headed households.
The Care Dependency Grant is meant for children up to the age of 18 years old who suffer from severe disabilities, and who require permanent home-based care. This grant amounted to R1 010 per month as from April 2009 (South African Government Services, 2009b:1). The grant is meant for severely mentally and physically handicapped children (Desmond & Gow, 2002:33;Sloth-Nielsen, 2004:29). The grant is means tested, which implies that the combined annual family income must not exceed R48 000 per annum, after deductions. Children in the terminal stages of AIDS are also eligible for this grant, although there is no formal policy to guide practitioners as to whether and when HIV-positive children may be awarded this grant (Desmond & Gow, 2002:33).
Of great concern is the fact that less than one-third of the eligible child-headed households are supported by social grants. One reason for this might be that children are not recognised by law as heads of households or primary caregivers, and are therefore not eligible to access grant support (Department of Health [DOH],. The difficulty in obtaining appropriate documentation, including birth and death certificates and identification documents, restricts access to social grants. The processing of welfare applications is also a lengthy process (Demmer, 2004:303). Changes to the legal system are required to make social grants more accessible to child-headed households.
Children who do receive grants indicated that the grants helped them to survive. Only one recipient of grants mentioned that the grant was misused. This might not be the only case of misuse, however. It is clear that social grants are an important source of income for childheaded households and they could play a major role in alleviating poverty in child-headed households. What is of great concern, however, is the fact that many adults pocket much of the grant without providing direct care for the children (Gauteng DSD Report, 2008:XV1). Aunts, uncles, teachers and neighbours were some of the people indicated who collected the grants. The direct care, however, came from other sources.
Further analysis of the income earned by child-headed households requires a comparison to acceptable criteria of the minimum income necessary to survive materially. Such an analysis will indicate how many children may not be able to meet their basic needs in order to survive.

Poverty in child-headed households
An analysis of the raw data reveals that no fewer than 16 of the respondents live in a childheaded household in which the average total monthly income is less than the R517.92 per capita per month estimated by Statistics South Africa as the minimum amount required for a person to survive materially. This effectively implies that 26.2% of the respondents cannot even support one person. Table 3 completes the analysis by reflecting the percentage of respondents who are able to support various possible household sizes, using the benchmark of Statistics South Africa as the poverty line.
Less than 40% of the respondent child-headed households in Gauteng are able to support a household of two members. This is an important figure, as the median household size of the respondents is two and the average size of a household is 1.8. The percentage of CHHs able to support their household size decreases significantly to a mere 11.5% if the household size comprises four people, and no respondent will be able to support a household size of six members.

TABLE 3 PERCENTAGE OF RESPONDENTS ABLE TO SUPPORT VARIOUS HOUSEHOLD SIZES, USING STATISTICS SOUTH AFRICA'S POVERTY LINE
Household size (number of persons) % of respondents able to support various possible household sizes with their total monthly income 1 73.8 2 39.3 3 31.2 4 11.5 5 6.6 6 0 (N = 61) Source: Authors" calculations from survey data Martins (2004:4) calculated the various amounts that families of different sizes would need to survive materially in various urban centres in South Africa. Table 4 reflects this yardstick, adjusted by the CPI for different family sizes in Pretoria, Gauteng. This could be regarded as another yardstick to measure the obstacles faced by child-headed households in Gauteng.  Martins (2004:4) and authors" CPI adjustments and calculations On average a typical child-headed household will only be able to support two members if the R517.92 per capita per month figure of Statistics South Africa is used. If the CPI-adjusted estimate of Martins (2004) is used as a benchmark, the situation worsens to such an extent that the average child-headed household is unable to support two persons. In fact, the average income of R1 121 per month constitutes only 80% of what would be required to support a household size of two people. Table 4 also reflects the corresponding ratios of bigger households. For a household size of six people, the ratio drops to 35%. Table 4 provides the results of the percentage of respondents who are able to support various possible household sizes on their total monthly income, using Martins"s (2004) alternative measures as a benchmark. The results are somewhat lower for each category of household size than those in Table 3, which was compiled using the poverty line of Statistics South Africa. The trend, however, is exactly the same. Once a household comprises three or more people, the percentage of respondents who are able to support this household size decreases rapidly. If the average household size is considered, then at least 50% of respondents" households are living below the two poverty line criteria employed in this analysis.

The last column in
By matching the individual household size of each respondent with its reported total monthly income, it is possible to calculate the existing poverty gap in child-headed households in Gauteng. The results indicate that no less than 44.3% of the child-headed households in the survey are living in absolute poverty. The sharp rise in the CPI, because of higher food prices in 2008, explains the significant increase in the criteria of both poverty lines. It is therefore clear that the vast majority of the child-headed households in this survey are indeed extremely vulnerable as a result of the higher cost of living, which always has a greater impact on the poor than it has on the more affluent sections of the community.
There are a number of child-headed households that have little or no income at all. The crucial question is therefore: how do these children survive? This is the topic of discussion in the next section.

HOUSEHOLDS WITH LITTLE OR NO INCOME
The pilot study revealed that several households had reported that they had no source of income. In the survey the households that declared little or no income were asked how they managed to buy the things they needed to survive such as food, transport and clothes. A total of 47 out of the 61 respondents completed this question. Based on a content analysis of the qualitative statements of these respondents, three important themes emerged. These themes describe the main dimensions of how households with little or no income managed to survive.
These themes are presented in Table 5 below in descending order of frequency.

Theme
Frequency % of subsample

Sample statements
Respondents make do with the income available to them and regard it as barely enough, but they are able to survive on it.

24
51.1 I buy very cheap stuff in order to save money to last me the whole month. The money is sufficient as we don"t pay for electricity. We struggle to buy proper clothes for the two of us. Material needs are provided to child-headed households in cash or in kind by family members, churches, neighbours and employers.

17
36.2 We get free food and clothes from the church.
The money given to us by our aunt helps us to buy most things we need. My uncle doesn"t give me money directly, but buys me the things I need. Material needs are provided for by one of the parents who is seemingly not part of the childheaded household. 5 10.6 My mother buys everything for us. Thus, we don"t need to buy clothes for ourselves.
Our father expects us to use whatever he gives us at month-end to buy food and clothes. We don"t pay for electricity. Statement(s) difficult to interpret. Because of the absolute and relative poverty in which they find themselves, orphans in childheaded households face particular challenges and exclusion, which poses a serious threat to their education (Maqoko & Dreyer, 2007:724), and this also applies to abandoned children in CHHs. This is of particular concern, as the structure of the labour market has changed during the past two decades. With a steadily increasing unemployment rate and a continuously declining demand for low-skilled labour in South Africa (Loots, 1998:332), these children have very little chance to escape the cycle of poverty that they find themselves in.
Obviously, these children also experience difficulty in finding food and shelter, which puts them in an extremely vulnerable position with regard to exploitation. They are also at high risk of being sexually abused by relatives and neighbours, and being drawn into the dark world of child labour or child prostitution.
In many cases neighbours and relatives play a positive role in the lives of children in CHHs. A total of 17 (36.2%) members of child-headed households say that many of their material needs are provided for in the form of cash and/or in kind by family members, churches, neighbours and employers. This speaks of communities where individuals are trying, within the restrictions of their own limited resources, to help these children wherever possible. It seems as though, with the help of the informal support systems in the community, most of these children have access to at least two meals per day via school nutrition programmes, neighbours and family, but they do not have food security.
A surprising theme is the number of children stating that their material needs are provided for by one of their parents. The biographical data indicated that 41% of the mothers and 49% of the fathers had abandoned their children. These children can therefore be regarded as "social orphans" (Cornia cited in Jones, 2005:163). There were five respondents in this category, representing 10.6% of the sub-sample. The parents have abandoned their children and do not form part of the same household any longer, but they still support their children in some way or another. They are possibly migrant workers and/or domestic workers. "Our mother comes to check on us occasionally", was the response of one of the respondents in this regard.
One respondent indicated that he/she had to live in the house, as the parent lived in another town in another house and they did not want to lose the house. The narratives indicated that there were children who might stay on their own in urban areas to be able to attend school, while the parents remained behind in the rural areas. They also feel responsible for their parents: "After I finish school, I will fetch my parents from the rural area and work for them"; "My dream has always been to buy a house for my parents". Some CHHs took on the added responsibility of earning an income while still attending school and taking care of themselves and/or their siblings. "I work after school and on weekends to help with money-generating activities"; "When we are not in school, we fix cars belonging to the neighbourhood"; "I drive local taxis on weekends to get money for food". This last statement indicates that these children will engage in illegal activities to earn additional income, as this child was driving local taxis without a driver"s licence, being under the age of 18.
A total of 35.6% of the children"s income was self-generated and placed an additional burden on them. As one of the field-workers reported: "One of the children was going to school in the morning, in the afternoon he fixed cars, and in the evening he had to cook, do his homework and see that the homework of his siblings was done. He expressed a wish for some support, as he could not cope with the stress anymore".
One of the 47 responses was virtually impossible to interpret accurately. The statement in question is presented in Table 6 for the sake of completeness and accuracy, but does not warrant further attention at this point.

CONCLUSIONS
The number of child-headed households in Gauteng, defined as those who have a head who is under the age 18, is very small and typically comprise one to two children. The majority of household heads are aged 17, while the youngest head is 12 years of age. The income available to child-headed households varies between no income to R3 000 per month. The average monthly income of a child-headed household in Gauteng is estimated as amounting to R1 121, and the median is R1 000 per month.
In 2005/06 the individual contributions from grants and money earned by the children"s own work differed significantly for child-headed households when compared to South Africa as a whole. Grants and pensions of parents generated just over 20% of the gross monthly income of these households. In the case of South Africa as a whole grants represented only 6.1% of the gross annual household income during 2005/06. Grants are therefore an important source of income for child-headed households. These grants are, however, very difficult to access, especially by a child as head of the household, as children are not recognised by law as heads of households.
These child-headed households receive financial and other forms of support from families, friends, siblings, neighbours, teachers, friends and community-based organisations. On average households in South Africa generated almost three-quarters of their annual income from work activities during 2005/06. For child-headed households this source of income amounted to 35.6% of their monthly income in 2008 (the year studied here).
Child poverty in South Africa is extremely high and rates differ across South Africa. Childheaded households also fall into this category. An analysis of the data reveals that 26.2% of households cannot support even one person with their total monthly income, while less than 40% would be able to support a household of two to three members. Only 11.5% would be able to support a household size of four people and none of the respondents" households would be able to support a household size of six members. This conclusion holds for both the povertyline criteria used in the analysis.
The situation of child-headed households in provinces other than Gauteng is likely to be worse than those in Gautengor at least not any better. By matching the individual household size of each respondent to its reported total monthly income, it was calculated that no less than 44.3% of the child-headed households in the survey are living in absolute poverty.
The majority of respondents who maintain that they have very little or no income state that they have no other option but to make do with what they have. This puts them in an extremely vulnerable position, particularly given the increased cost of living. These disturbing statistics should compel policymakers to consider the needs of vulnerable children in designing appropriate policy interventions. It must always be kept in mind that these children have definite emotional needs (such as care) in addition to their physical needs for food, shelter, clothing and protection. Any policy intervention must have, as its long-term objective, the eradication of the cycle of poverty and vulnerability in which these children find themselves.
Neighbours and relatives play a positive role in the lives of these children. A significant number of child-headed households say that family members, churches, neighbours and employers meet many of their material needs in the form of cash and/or in kind. A surprising theme is that some children state that their material needs are provided for by one of their parents. These parents therefore help to support their children, but do not form part of the same household.
The migration of caregivers to their workplace during the week or during specific seasons results in many children living in child-headed households during the week.

RECOMMENDATIONS
The fact that many children in CHHs are abandoned by their parents reinforces the need for policy interventions aimed at promoting the preservation of families. It also emphasises the need for applicable statutory action against parents who abandon their children.
Many children in CHHs have an absent parent as a result of the migration of caregivers to their workplace during the week or during certain seasons. These children are frequently responsible for their own food, financial management and school attendance during the week (Tucker, 2008:1). Schools should therefore play a vital supportive role in the delivery of policy interventions aimed at improving the socio-economic circumstances of child-headed households. A further important recommendation flowing from the results is that children heading child-headed households should have direct access to social security grants instead on relying on adults, who may abuse the money.
It is the authors" view that for any policy intervention to be successful, the Departments of Health and Education, in addition to the Department of Social Development, should play a much more active role, and that interdepartmental coordination would have to be improved significantly. The analysis of this important social problem will also benefit from extending the debate to all households with vulnerable children, instead of merely concentrating on childheaded households.
Another critical aspect that warrants further investigation in future surveys is to be found in the expenditure side of child-headed households. It is important to analyse their spending patterns, as this information could provide valuable insight into possible policy interventions to alleviate the socio-economic plight of child-headed households. Spending patterns differ according to race and province (Statistics South Africa, 2008b:10). Follow-up research on the expenditure side of the economic situation in child-headed households should aim to further explore their economic vulnerability in greater detail. The authors also recommend that a similar study be undertaken with regard to foster and extended families who need to absorb biological and "social" orphans into their financially overloaded households.
The dire socio-economic circumstances of child-headed households in Gauteng revealed in this study should indeed serve as an urgent call to action for all who claim to have the wellbeing of the children of this country at heart. Indeed, as one of the respondents commented, "Life is not pap and vleis" for child-headed households.