A synthesis of homelessness in South Africa: A rapid critical appraisal

ABSTRACT

Homelessness in South Africa requires collaborative research. Social, political, cultural and economic factors that cause homelessness must be considered to develop a common definition and understanding of homelessness. South Africa’s social complexity and diversity make it challenging to construct available results into one conceptual framework, and this in turn complicates national policy implementation and role allocation. By means of a rapid critical appraisal of literature on homelessness in South Africa, researchers provide evidence to direct and structure contextual research pertaining to homelessness. Four themes transpire: conceptualisation; demography; roles and responsibilities of key stakeholders; and strategic policy and research issues. Researchers caution that the findings are not generalisable, because of the rapid nature of the appraisal and possible selection bias of the literature. Measures of validity were used to ensure that the study accomplished its purpose through the key results and to ensure that the results are a true reflection of available evidence.

KEYWORDS:

• Conceptual framework
• defining
• homelessness
• rapid critical appraisal
• South African context

View correction statement:

Corrigendum

1. Introduction

Homelessness is a worldwide phenomenon, with an estimated homeless population ranging from 100 million to one billion or more, depending on how homelessness is defined (HABITAT, 1996). Homelessness has been identified as a global issue that requires collaborative research to explore the best approaches to address the needs of low-income and homeless people, to determine ‘real world’ global health services and policy questions, and to assess how this research can be applied in developing policies and service practices with people who are living in poverty or who are homeless (Richter et al., 2012). Kemeny & Lowe (1998) argue that certain social, political, cultural and economic factors lead to distinct national housing systems. One such system is the homelessness issue.

In South Africa, the homeless population comprises a large percentage of people with low or no income who are increasingly migrating to the larger cities in search of work, which in turn increases the housing problem in these densely populated areas. Schively (2007) indicated that about 16% of the population was living in inadequate housing such as informal settlements and dwellings in 2004. Defining homelessness is a complex matter, because authors use different descriptions such as ‘ … not having a place to stay’ (Sanchez, 2010:102) or ‘ … a continuum, ranging from people who may be at risk of becoming homeless to those who currently have absolutely no shelter of their own and live and sleep “on the streets”’ (Seager & Tamasane, 2010:64).

In a meaningful search of available literature applicable to a South African context, a rapid critical appraisal on available literature in the field of homelessness was conducted to gain a better understanding of homelessness in South Africa.

2. Problem statement

The social complexity and diversity of the South African society make it challenging to establish one common conceptual framework which can accommodate the nature and different types of homelessness in South Africa. The interrelated cultural, economic, political and social factors which influence homelessness make it a matter of urgency for key role players and stakeholders to develop a common definition and understanding of this concept.

Although some research has been done on homelessness in South Africa, there is no conversion of all of the results into a useful and operational conceptual framework for homelessness in South Africa. Even with regards to defining the typologies of homelessness, there is diversity and often disagreement amongst researchers. This makes it difficult to implement the national policy framework on homelessness as well as to allocate the roles and responsibilities to the appropriate role players and stakeholders.

This rapid critical appraisal forms part of a comprehensive situation analysis of homelessness in South Africa that furthermore provides evidence to direct and structure research and research instruments pertaining to homelessness in a South African context.

3. Purpose and objectives

The purpose of the study was to conduct a systematic review through a rapid critical appraisal to conceptualise and define homelessness in South Africa. The critical appraisal methodology was used to compile a synthesis of key results, to describe some lessons that were learned and to make recommendations towards the understanding of homelessness in South Africa.

The objectives of the study were as follows:

• to conceptualise and define the nature, types and characteristics of homelessness in South Africa;

• to develop a demographic profile of homelessness;

• to identify key policy and research issues pertaining to homelessness;

• to map the relevant stakeholders and role players to address homeless in South Africa; and

• to describe some lessons that were learned and make recommendations in accordance with these.

4. Methodology

4.1. Critical appraisal

The critical appraisal research method focuses on finding and utilising research results to support professional decisions using a systematic and meta-analytic process based on appropriate and valid evidence (Makela & Witt, 2005). This involves, among others, filtering through a large body of research with minimum criteria; using adequate reporting of details on purpose, objectives, data collection and analysis, adhering to the technical requirements for the type of research methodology and study elements; indicating methodological soundness and paradigm adequacy; and reflecting on research responsiveness of data and theoretical consistency in methodology and findings or results (Jackson et al., n.d.).

A critical appraisal provides evidence-based arguments which, in terms of an implied research question, claim or propose evidence to support the argument or case. This evidence can be deduced from existing evidence (published and non-published), to provide a set of assumptions or recommendations which will inform and validate decision-making (Marston & Watts, 2003). To this end, a critical appraisal is essential to put research into practice, as it scrutinises existing knowledge to determine within a short period of time in which areas there is existing research and where there are gaps in research. This allows researchers to conduct reliable, up-to-date research with control over the reach and scope of studies.

Nine steps were followed to conduct the critical appraisal, as described by Burls (2009) and UCL Institute of Child Health (2011). Steps one to three entail the formulation of a clear research problem statement and of specific research questions, to ensure that the objectives which are set forth will be accomplished. The fourth step is the development of criteria for the critical appraisal by developing a standardised tool to ensure consistent reviews of the selected and appropriate literature by a group of researchers. The research team has adopted a rapid approach to the appraisal by limiting evidence to 30 pieces and conducting it over a very short period of time. Pieces of evidence were selected during step five for the scope of the assessment from a database compiled by the University of South Africa Library for literature dated between 2000 and 2011, research articles published in accredited journals for both national and international content and for content focusing mainly on the nature and types of homelessness.

The appraisal of pieces of evidence in step six was conducted independently by two teams of researchers at the University of South Africa (UNISA) and the North-West University (NWU) at separate locations. During step seven, these teams used descriptive data analysis on all appraisals to retrieve key results and to identify pieces of evidence and themes for the synthesis. From these findings a basic descriptive data analysis, including frequencies and averages, was conducted to summarise the key results of the critical appraisal. This process was also used to identify the pieces of evidence and themes of the synthesis to compile the main themes emerging from the key results as step eight. Each research team had to summarise in more detail the main discussions under their themes in step nine, after which a working session with the researchers was facilitated to identify some lessons that were learned and to formulate recommendations based on the pieces of evidence, as well as on the results of the critical appraisal and the synthesis, in order to create an appraisal report.

4.2. Limitations

There are certain limitations to this appraisal report – with the outcome that the results cannot be generalised – including but not limited to the following:

• Rapid assessment scope of the study. The main focus of the study was to provide a common understanding of the conceptualisation and definitions of homelessness amongst the researchers in a short period of time.

• Bias towards selecting pieces of evidence. There was a level of bias in selecting the pieces of evidence which were used in this appraisal, due to practical reasons and time constraints. A wider selection must be included, such as different formats of published and unpublished pieces of evidence, to reduce bias.

Some measures of validity were applied to ensure that the study accomplishes its purpose and objectives through its key results. The finalisation of the critical appraisal tool was conducted in two separate settings (UNISA and NWU), which contributed towards obtaining a level of objectivity. A trial run was performed with the draft tool to assess its applicability and appropriateness related to the topic. The researchers first formulated the research question before commencing with the study, which allowed the study to have a central focus. All selected sources were obtained from accredited academic journals.

A cross-review was conducted with 20% (six out of 30 pieces) of the evidence to validate the creditability of the critical appraisal process. Key results of the synthesis were also validated as authentic through the inclusion of strategic quotations from different pieces of evidence that were used to compile the synthesis.

Reliability was obtained using the critical appraisal research methodology (see Section 4), which is a well-established method for a systematic review in medical and social sciences for purposes of evidence-based decision-making.

5. Key results from the critical appraisal

The following information was obtained from the 30 pieces of evidence that were critically appraised by means of the critical appraisal tool.

5.1. Biographical information regarding the pieces of evidence

A list of key words was predetermined, based on a pre-reading of 10% of the sample of pieces of evidence by the reviewers. This list of key words assisted the reviewers to decide whether to include a particular piece of evidence in the appraisal or not. All key words were relevant to the selected pieces of evidence. Most of the evidence (an average of 82%) provided information on the nature and types of homelessness, while 12% described the types of role players and support which are needed to address homelessness. Only an average of 6% provided other information.

Seventy-four per cent of the evidence was first-hand (primary) accounts, while 16% was second-hand (secondary) accounts. An average of 53% also provided some guidelines and information on policy development and implementation. An average of 40% of these documents clearly indicated how and where they obtained their evidence.

5.2. Profile of homelessness

It became evident that homelessness is a multi-dimensional and very complex phenomenon. There is no single description amongst the appraised pieces of evidence which commonly describes the characteristics of homelessness.

An average of 25% of the pieces of evidence describes the homelessness settings mostly as being urban, inner-city and shelters. Some pieces of evidence (an average of 15%) also describe these settings in terms of different types of informal housing, like informal settlements, living on the side of a building or a wall and temporary housing.

More than half of the pieces of evidence (an average of 54%) describe the different types of homelessness as mostly being adult men who are unemployed and street dwellers. About one-third (32%) of the pieces of evidence indicated that the key role players for addressing homelessness include policy-makers, social networks, government and others. Most of the pieces of evidence (an average of 82%) show that most of the key role players had some kind of knowledge and skills related to homelessness as well as a positive attitude towards homelessness by rendering various types of support services to homelessness people.

Most of the pieces of evidence (82%) provide some information on the reasons why people become homeless, including socio-economic and environmental factors, social displacement and migration. The main support activities described by more than half the pieces of evidence (66%) include the provision of social services, programme responses and development. Only 12% of the activities reported by the pieces of evidence were not completed, as most were ongoing, current and often new interventions or programmes. All of the pieces of evidence focus on the quality of the activities by describing their impacts on homeless communities.

5.3. Categories of needs and support services

The reviewers were asked to list all of the needs described in the different pieces of evidence, after which they coded them further within four main themes of needs. Eighty-three per cent of the pieces of evidence identifies four main needs among homeless people, namely economic, psychosocial, personal and socio-political needs. There were sub-needs under each of these four themes. Sixty per cent of the pieces of evidence describes different types of support activities for homeless communities, including health and social support services, economic support and policy support. They are listed under three main headings:

• Health care and social support activities

  • ○Faith-based support, family support, social networks, social reintegration, social support structures, collaboration between public and private sectors.

  • ○Shelters, day-care centres, outreach centres, community care centres, mother and child support.

  • ○Prevention strategies, personal skills development, proactive and advisory support, food schemes.

  • ○Socio-medical interventions, more targeted interventions, temporary interventions.

• Economic activities

  • ○Employment, job creation.

  • ○Housing, transitional housing, safe houses.

  • ○Poverty alleviation, community mobilisation.

  • ○Funding for programmes.

  • ○Financial support to municipalities and metro-cities.

• Policy activities

  • ○Advocacy, criteria to determine needs, protection of women and children, policy reform.

  • ○Comprehensive rehabilitation, institutions to offer help, social protection, housing.

  • ○Training of officials.

5.4. Overall rating of the pieces of evidence

A five-point score based on appropriateness, applicability and content related to homelessness was used by the reviewers to rate each piece of evidence on aspects pertaining to the applicability of the evidence; elements of capacity-building and training needs; describing benefits of interventions and programmes to communities and homelessness people; evidence on research conducted which may allow some level of replication; and overall impression. Evidence was rated from low (1), to moderate (3) to high (5).

5.4.1. Applicability of evidence

Forty-three per cent of the pieces (13 pieces of evidence) directly addressed the targeted audiences in their discussions, while 30% contained broad information and only 13% contained partial on homelessness but enough to draw upon for the synthesis. Less than half of the pieces (an average of 40%) had sufficient evidence on homelessness and only two pieces did not contain any evidence, although they contained related information.

5.4.2. Evidence on capacity-building and training

Only three pieces of evidence described training activities, like the training of officials on sensitivity towards homelessness and the training of service providers to understand the nature of homelessness.

5.4.3. Benefits of interventions and programmes

Very few (10%) pieces of evidence identified any benefits of different interventions or programmes to support homelessness, like the possibility of rehabilitation and public participation in policy reform.

5.4.4. Support to duplicate/replicate intervention/research

Nearly 63% of the pieces of evidence support decisions to duplicate or replicate the interventions or programmes they are describing.

5.4.5. Overall impression score on the pieces of evidence

An average of 73% of the pieces of evidence were positively rated, while 27% were negatively or moderately rated.

6. Synthesis

6.1. Pieces of evidence

Based on the overall rating, the 10 pieces of evidence as listed in the following were identified as the most applicable and appropriate for the purpose of the study. These were used to summarise the key results into four main themes to achieve the objectives of the study:

• Cross, C & Seager, JR, 2010. Towards identifying the causes of South Africa’s street homelessness: Some policy recommendations. Development Southern Africa 27(1), 143–58.

• Cross, C, Seager, JR, Erasmus, J, Ward, C & O’Donovan, M, 2010. Skeletons at the feast: A review of street homelessness in South Africa and other world regions. Development Southern Africa 27(1), 5–20.

• Du Toit, JL, 2010. Local metropolitan government responses to homelessness in South Africa. Development Southern Africa 27(1), 111–28.

• Kok, P, Cross, C & Roux, N, 2010. Towards a demographic profile of the street homeless in South Africa. Development Southern Africa 27(1), 21–37.

• Makiwane, M, Tamasane, T & Schneider, M, 2010. Homeless individuals, families and communities: The societal origins of homelessness. Development Southern Africa 27(1), 39–49.

• Naidoo, V, 2010. Government responses to street homelessness in South Africa. Development Southern Africa 27(1), 129–41.

• Noee, RM & Patterson, DA, 2010. The ecology of homelessness. Journal of Human Behaviour in the Social Environment 20, 105–52.

• Sanchez, D, 2010. Civil society responses to homelessness. Development Southern Africa 27(1), 101–10.

• Seager, JR & Tamasane, T, 2010. Health and well-being of the homeless in South African cities and towns. Development Southern Africa 27(1), 63–83.

• Speak, S & Tipple, G, 2006. Perceptions, persecution and pity: The limitations of interventions for homelessness in developing countries. International Journal of Urban and Regional Research 30(1), 172–88.

6.2. Main themes of the synthesis

The following four main themes were identified based on the purpose, objectives and key results of the critical appraisal.

6.2.1. The conceptualisation of homelessness in a South African context

South Africa, like other developing countries, is challenged by homelessness as it is embedded in political, social and economic aspects (Makiwane et al., 2010). Naidoo (2010) notes that the physical condition of being homeless is driven or influenced by a variety of dynamic latent social and economic factors. The mobility and nature of homeless people and the different ways in which homelessness is defined contribute towards the lack of an evidence-based understanding of the homeless population in South Africa. Despite the difficulties to define homelessness, Seager & Tamasane (2010) stress that it is very important to come to a reliable definition in order to make meaningful comparisons between different studies. Homelessness should be regarded as being on a continuum, ranging from people who may be at risk of becoming homeless to those who currently have absolutely no shelter of their own and live and sleep on the street, in parks, in train stations and the like, and who do it mostly for economic reasons (Naidoo, 2010; Seager & Tamasane, 2010; Makiwane et al., 2010). Naidoo (2010) notes that homelessness means to be without a home, to have access to a shelter or to live in circumstances that are totally inappropriate and of bad quality to personal and social needs.

The homeless population is often viewed as an unwanted and vagrant population (Cross et al., 2010). Homeless people are also often stigmatised because the community sees them as people who have lost their self-confidence, dignity and self-respect; who have alcohol and other substance abuse problems; who have risky sexual relationships even resulting in HIV and AIDS; who are prone to stealing; who are poor, unemployed, often in transit, deprived and mentally and physically ill; who come from disintegrated and dysfunctional families; who very probably had an underprivileged childhood; who have had only partial education and who lack skills; and who are accustomed to violence (Naidoo, 2010; Sanchez, 2010; Seager & Tamasane, 2010; Makiwane et al., 2010). Homelessness is often the final stage in a lifelong series of crises and missed opportunities and a gradual disengagement from supportive relationships and a proper social network (Sanchez, 2010).

6.2.2. The demography of homelessness in South Africa and internationally

The demographics of homeless people include variables such as settings (places and types of shelter), personal characteristics of homeless people (age, gender, employment status, stability and family composition) and reasons and factors for homelessness. Researchers report that homelessness is a multifaceted phenomenon (Ofumi in Kok et al., 2010), and obtaining reliable information about this group(s) of people in South Africa is extremely challenging. Homelessness affects all races in South Africa today but the black population makes up the largest component. There is also a significant growth in the numbers of white and coloured people who are becoming homeless (Seager & Tamasane, 2010).

Olufemi, as cited in Makiwane et al. (2010), points out that in South Africa, homelessness affects adults more than children and more men are homeless than women (Kok et al., 2010; Seager & Tamasane, 2010). Fourteen per cent of adults in the study by Kok et al. (2010) indicated that they became homeless before the age of 18. Adult respondents had on average been homeless for 6.4 years. Street children appear to be predominantly older children and become homeless when they can no longer use children’s shelter systems (Kok et al., 2010). In cities it also appears that most children are leaving the streets as they reach adulthood, to be replaced by a new cohort of children. It is not clear where the adults go when they leave the streets. Unlike adults, children cannot obtain their own shack housing and may have nowhere else to stay but on the streets (Du Toit, 2010).

There seems to be a clear distinction between homeless people living in shacks and homeless people who are street dwelling (Cross & Seager, 2010). In the shack-dwelling population as many as 40% of adults hold formal jobs. There are more female residents (about 59%) and most of them survive on child grants, while 24% of the disabled receive disability grants. The majority of homeless people have an average education of eight years. Unlike street homeless people, the people of shack settlements participate in the normal migration society of poverty in South Africa. The shack population is less severely excluded in terms of social policies and may already have certain advantages related to some form of home ownership.

South Africa’s street homeless are a small and a highly specific group, compared with the shack population. They differ in terms of access to employment and government benefits, and a resisting core group has been identified that commonly rejects any assistance. Issues like formal shelters, cheap rental options, homeless people’s drop-in centres and help with transport and formal housing delivery still require much investigation. There are more males (87%) than women and children among the street homeless. Unemployment seems to be one of the migration push factors, as these people have little or no access to reliable income, with only 27% being formally employed, usually as casual workers. The educational levels of the majority of street homeless people are not sufficient to enable them to enter the job market (Weinberg in Kok et al., 2010), with average schooling up to primary school level (i.e. Grade 7). Cross & Seager (2010) observe that social security mainly targets people staying in shacks and other poor households. Social security is seldom accessible to street-dwelling homeless persons. Street homeless people are also not protected by social grants, mainly because they do not have a physical address and are mobile and unstable in numbers.

In addition to the high percentage of homeless people who reported that they were South African citizens, a number of areas were also identified that are frequented by homeless people from across national borders. These areas are located in a migration corridor stretching from South Africa’s north-eastern border with Swaziland, westward across Mpumalanga and southern Limpopo, into Gauteng. Cross-border migration appears to be a very significant contributing factor towards homeless populations in South Africa, because approximately 14% of the respondents in the study by Kok et al. (2010) were citizens of other countries, mainly from Mozambique and Zimbabwe.

6.2.3. Roles and responsibilities of key stakeholders involved in homelessness

Homelessness is a reality in both rural and urban South Africa. It needs to be asked whether the legislative framework captures the complexity of the social and economic circumstances of street homelessness (Naidoo, 2010). National legislation in South Africa does not deal directly and specifically with ‘homelessness’ or the ‘street homeless’ in any single statute, but it does otherwise respond, through a variety of legislation, to the social and economic conditions or circumstances of the street homeless (Naidoo, 2010). Cross & Seager (2010:154) observe that ‘[s]treet homelessness in South Africa is only partly a problem of welfare delivery’, which alludes to the extent and complexity of homelessness, and challenges key stakeholders to offer solutions.

The findings reflect that central government is the main stakeholder in the homelessness phenomenon, as it has political structures in place that set social policy. However, persons in key positions in local society are often those who deal with homelessness, and this may result in a lack of consensus on what to address, and how to go about this, regarding homelessness in South Africa, and uncertainty about whom should be held responsible. Cross et al. (2010:8) note that the policies are ‘driven by empowered actors in local society – and actualised by the fears of businesses and local administrators – rather than coming down from higher levels of government urging the municipalities to act’.

Since 1994, the government agencies in South Africa concerned with homeless persons are the national, provincial and district Department of Housing, Department of Social Development and Department of Health. In the South African health sector it is essential to consider services and stakeholders like provincial hospitals, day hospitals, primary health care clinics, retail pharmacies, faith healers, traditional healers/herbalists, home-based care and private doctors (Seager & Tamasane, 2010:64). The same authors note that prevention ‘is the preferred health intervention strategy, but where there are large populations of sick people, resources must be brought to bear on current problems, while at the same time addressing prevention of further “cases”’.

Faith-based organisations respond to homelessness by ‘dealing with social problems, particularly in difficult socio-economic environments’ and are seen as ‘trustworthy and capable, particularly in harsh environments where the state itself might struggle to operate’ (Sanchez, 2010:103).

Role players such as international development aid agencies, universities and the Human Sciences Research Council not only added valuable information on the extensiveness of homelessness in South Africa (Cross et al., 2010; Cross & Seager, 2010; Seager & Tamasane, 2010), but emphasised the need for further exploration and description on the reality of homelessness in South Africa.

The roles and responsibilities of stakeholders as key role players in homelessness should not be underestimated, because they often form tight and supportive social networks. It is, however, advised that these stakeholders should integrate the suggestions discussed by Du Toit (2010).

6.2.4. Strategic policy and research issues related to homelessness

Often there is a sense that homelessness is politicised around policy formulation and implementation: ‘The dichotomy in debate positioning homelessness as either a result of individual or structural factors may have served political and policy objectives, but does not advance an etiological understanding of homelessness’ actual complexity nor does it foster multi-systemic response options’ (Noee & Patterson, 2010:106). This argument highlights the complexity of formulating global and national policies around homelessness.

There seems to be quite a number of issues that need to be addressed before formulating global and national policies around homelessness. Du Toit (2010) notes that although there are policies which address homelessness in South Africa and internationally, these policies need to be consolidated to be effective. He further argues that more should be done in terms of monitoring and evaluating the implementation of these policies and programmes, to determine whether they have an actual impact on homelessness. Cross et al. (2010:8) express concern about the reasons for the lack of policy implementation and urge that ‘before central or local government in South Africa attempts more accommodating approaches, their power to put them into practice may be restricted by the lack of a clear policy consensus and the absence of data on the homeless population’. Du Toit (2010) is also of the opinion that a lack of policy consensus and data also results in serious constraints with regard to implementing national and local policies at municipality levels. Du Toit (2010:121) believes that ‘although municipalities are well prepared at a policy level, they lack preparedness on the levels of resources and implementation, making it difficult to put policies into actions’. He further explains that ‘the policy, legislative and financial support municipalities may or may not receive from national and provincial governments are therefore more important than any other factor when it comes to responding to homelessness’. Cross & Seager (2010) also argue that the focus of policies should be more on prevention and remediation rather than on reaction, as is currently mostly the case. They further indicate that, for South Africa, policy efforts have so far concentrated on putting figures to budgetary needs and that a comprehensive programme of measures to prevent homelessness has not yet been attempted. This entails a closer collaboration between community-based organisations, faith-based organisations and all layers of government. Sanchez (2010) emphasises the importance of acknowledging and ensuring the role of civil society in addressing the needs of homelessness in South Africa and internationally.

Naidoo (2010) considers it essential that social welfare and housing departments should coordinate the sectors’ policies and programmes to ensure a more comprehensive response to the needs of homelessness. Cross & Seager (2010:154) note that ‘to prevent new street homeless, interventions need to address the effects of unemployment on the family and include: reduction and alleviation, prevention and targeted interventions’. They further stipulate that ‘the issue of social welfare support needs to be addressed more fully. The bulk of social grant go to children or their caregivers, the elderly and the disabled’. It should also include a wider scope of homeless people, including male adults, women and children (Cross et al., 2010).

Sanchez (2010:105) stresses the importance of addressing the following issues in any policy formulation:

It is the people in need. Someone living on the streets or in bad conditions. A person who is struggling with issues of unemployment, housing and food. Those who are living on the streets and have nowhere to go for various reasons. Those who struggle and are without. Those who lack legal and economic power and this puts them in a vulnerable position. It is an ongoing negative cycle. People with no physical structure where they can better themselves in terms of their human dignity. For instance, people living in overcrowded mekhu-khus [shacks], since they don’t have possibility of developing.

Noee & Patterson (2010) attempt to distil the collection of social science literature into a coherent and cogent map in their ecological model of homelessness towards a comprehensive and integrated policy approach to homelessness.

It is clear that all policies should work systematically through key issues related to the bio-psychosocial factors pertaining to individuals and social structures that are inducive to an environment for homelessness. This in turn has direct and indirect socio-economic outcomes related to homelessness which may affect both the individual and social structures and may require different social housing interventions and support. This model allows policy-makers to be more comprehensive in their approach as well as to approach homelessness from a multi-sector approach.

However, one of the major constraints on the provision of housing is that the problem often lies with the housing policy of the country. South Africa, for example, does not make any provision for the housing needs of homeless people (Cross et al., 2010). This requires serious consideration with regard to formulating any policy related to homelessness.

7. Conclusions

Although there is a broad definition of homelessness globally as discussed previously, it is problematic to establish a common understanding and definition of homelessness within the social diversity of South Africa. There are often too many inter-related social, political, cultural and economic factors which vary in content and context for describing the nature and types of homelessness in urban, semi-urban and rural settings. Although homelessness pertains to homeless people or people living on the street, the phenomenon is more complex than this because the population to which it refers is also very dynamic and mobile.

The demographics of homeless people also vary vastly, according to certain socio-economic and socio-political conditions like poverty, lack of housing or shelter, unemployment and social disruptions. Although most homeless people are adult males, there has been an increase in the number of children, women and elderly people who are becoming homeless across South Africa. Children still remain the most vulnerable group amongst homeless people due to their psychosocial needs such as access to care, education and health care. They have a need for adult role models and are often in survival mode.

There is also a culture amongst homeless people which provides a social platform for caring and sharing. Formal and informal employment and housing for these people are often created before any official intervention has been implemented.

Although South Africa has a national policy framework for homelessness, there still exists a substantial amount of confusion about roles and responsibilities. This confusion is not limited to the level of national and local government structures but also occurs amongst different stakeholders like community and faith-based organisations. Homelessness in South Africa is an evolving phenomenon and requires a multi-sector approach towards formulating policies and programmes which will not only address the socio-economic needs of the homeless individual or groups, but also the social structures that are affected, like communities and households.

8. Lessons learned

The conceptualisation of homelessness in South Africa is embedded in political, social and economic aspects, which makes it difficult to formulate a single definition of homelessness. The physical condition of homelessness is thus influenced by the dynamics latent to social, political and economic factors (Naidoo, 2010).

The demography of homelessness in South Africa and internationally reveals that the most affected groups are male adults and children. Children, unlike adults, are not able to own shack housing and end up staying on the streets since they do not have any other place to stay. The lesson learned from this study is that children are the most vulnerable group and that they are often exploited by being forced into begging for money on behalf of others, by people who expect favours (often sexual in nature) for money, and by peers and fellow street children. The lack of legislative polices to create and maintain sustainable supportive and protective infrastructure to address the needs and rights of street children is a matter of concern.

Another lesson learned in this regard is that the movement of people from other countries of origin into South Africa (usually due to political unrest, poverty and unemployment) and the movement of people from rural to urban areas in search of a better future will continue to increase. Therefore, there is a need to find long-term solutions to this problem, as it aggravates the crisis of homelessness in South Africa and contributes to conditions associated with homelessness, such as a burden on social and economic infrastructure, crime, diseases and environmental pollution. Homeless people who are not well educated find it difficult to gain employment. On the other hand, those who are educated sometimes find only temporary employment or become de-skilled after being unemployed for some time; and they are then forced to take menial jobs that cannot sustain them for a long time. This discourages work seekers and makes them feel ashamed and vulnerable to exploitation.

There also seems to be no consensus or coordinated strategy on how to go about addressing issues of homelessness in South Africa. No sector, discipline or institution is prepared to take the initiative and to assume the responsibility to address and coordinate homelessness effectively. Homeless people rarely have an opportunity to give their input and to collaborate on projects that focus on homelessness and homeless people. In most cases, it is people in power in local society who are prepared to deal with homelessness. The lesson from these findings is that intervention plans for homeless people can be jeopardised by negative and false views of homelessness portrayed by the public media or politicians. Roles and responsibilities of homeless people as key role players are underestimated. This puts a huge burden on government sectors such as Social Welfare and Health. A lesson to learn in this regard is the importance of creating enabling legislative, environmental and social opportunities for the homeless to embark on entrepreneurial and money-generating activities and projects without the fear of xenophobic and destructive responses.

9. Recommendations

• An integrated approach is necessary to address the many facets of homelessness in South Africa.

• Although it is recognised that the complexity of homelessness makes it difficult and unaffordable for only one sector or institution to take responsibility for the issue, the cost-effectiveness of efforts to address and provide services to the various forms of homelessness will be enhanced if a coordinating, advocating structure could be assigned the primary responsibility for it.

• Further research on homelessness is recommended because the research outcomes can be useful for identifying intervention strategies to alleviate the problems of homelessness. Active participation by the homeless in exploring and analysing the phenomenon of homelessness is essential for the relevance and sustainability of intervention strategies.

• There is an increased need to monitor and evaluate the implementation of the policies and programmes to determine whether they have an impact on homelessness in South Africa.

• The resilience of homeless people should not be underestimated. Government, researchers in communities and other stakeholders should give them the opportunity to demonstrate their resourcefulness to find solutions to their problems.

Acknowledgment

The authors would like to acknowledge Ms Germari Kruger for editing and technical support.