Urbanised Ageing and Strategic Welfare Space in a Namibian Former Township

ABSTRACT

The number of African older people who live permanently in urban areas is growing. This qualitative ethnographic study explores how older people employ welfare strategies, often involving members of the extended family in mutual care and support. These welfare strategies are emplaced; in this case, in different housing types in a former township in Namibia – Kuisebmond in Walvis Bay. Older people stay in former township houses, in backyard shacks or other rentals, or at an old-age home. Government welfare that was adjusted to family needs appeared in similar shapes in these housing types, such as access to better schools. Older people were both caregivers and receivers of care in these efforts. Taking care of grandchildren while their parents migrated for work was a mutuality of informal support that was highly beneficial to all involved. The non-contributory pensions facilitated many strategies by alleviating risks. Access to high quality housing and government healthcare made urban living a feasible alternative that challenged rural living. The study concludes that housing is a strategic welfare space where formal and informal welfare are optimised in various ways. Older individuals contribute to a large extent to the adjustment, maintenance, and development of these joint spaces.

KEYWORDS:

• ageing
• township
• welfare
• housing
• strategic welfare space
• Namibia

The idea that ageing in an African context is a rural phenomenon is challenged by the fact that a growing group of older people now live permanently in urban settings (Apt 2000; Indongo & Sakaria 2016; Nord & Byerley 2020). Images of older people’s welfare in urban areas diverge. Some claim that disintegrating family support to older people is particularly urgent and severe in urban areas (Aboderin 2004; Apt 2000), while others suggest that the traditional multigenerational welfare system still functions reasonably well (Ananias 2014; Nord 2021; Oduaran & Oduaran 2010). State-provided welfare ranges from being virtually non-existent to modestly provided in different sub-Saharan African countries. Namibia – which is the case in this study – is a state with comparatively generous government welfare provision and satisfactory outcomes (Clement 2020). However, the family, often comprising several generations, is of great importance in the production of welfare to complement or take the place of government welfare (Abu Sharkh & Gough 2010; Winterfeldt & Fox 2002). This has implications for older people in their roles as caregivers and as receivers of welfare. It is expected that this co-provision of welfare will give insight into how government- and family contributions of welfare influence each other in an urban context.

This paper thus explores how older people fare in a Namibian urban context by relying on and using government- and family-provided welfare. Household compositions and housing are important parameters for intergenerational relations in the extended family (Nord 2021). Housing may impact other welfare system components (Kemeny 2001). Of the five welfare pillars suggested by Peter Malpass (2004), the paper analyses how education, healthcare, social security and personal social services are received and optimised in housing by older people and their families. Much of existing research has focused the commodified aspect of housing, home ownership and its relation to the welfare system, for instance as a complement to pensions. This research has primarily been carried out in European countries (for example Kemeny 2001; Malpass 2004; Stamsø 2010), in Asia (for example Zhou & Ronald 2017), and in Africa (for example Hoekstra & Marais 2016). Very few other aspects have been in focus, for example welfare and generational cohabitation (Arundel & Ronald 2016). Housing in Namibia is often a common resource of substantial value to the family; in this paper however, the researchers take an interest in other aspects than the commodification of housing and welfare. Central questions for the study are: How do family and government welfare contributions intersect and merge in the Namibian urban situation? What family coping strategies are prompted? What are the spatial implications of these strategies?

The paper is structured thus: part one presents Namibian welfare regime and housing provision, and part two is an analysis of three types of housing alternatives in the case study. The paper concludes with a discussion of welfare strategies, their distribution in space, and their policy implications.

Informal security welfare regime – the Namibian case

In general, welfare regimes in African countries are either ‘insecurity regimes’ with weak or no state provision, in which there is dependence on international aid and a heavy reliance on family for welfare, or ‘informal security regimes’ in which the state provision of welfare is modest or low, and family contributions are significant (Gough, Wood, Barrientos, Bevan, Davis & Room 2004). Human Development Index scores are relatively high in the latter regimes. A recent comparative analysis of social protection in sub-Saharan Africa identified Namibia as an ‘informal security regime’ and placed the country in the highest ranked of four clusters, with above-average results in all welfare variables measured (Clement 2020).

Welfare systems in sub-Saharan Africa bear the legacy of former colonial administrations and economic structures. Namibia is an example of a settler- and labour reserve economy in which government welfare provision was initially delivered to a group of white settlers and for the control of Black labour’s urbanisation. This comparatively generous welfare programme was made available to all citizens upon independence (Mkandawire 2020). The government welfare system that Namibia inherited was inequitable and of a highly disparate quality, based on citizens’ ethnic lines (Ananias & Lightfoot 2012; Devereux & Lund 2010). The legacy of the settler- and labour reserve economy remains as an economic structure. Namibia is the world’s second most unequal country, with a Gini coefficient of 59.1.¹ This extremely imbalanced economy includes high income earners that provide a tax base for welfare provision (Niño-Zarazúa, Barrientos, Hulme & Hickey 2010). This is double-sided, however: The economic situation for Black people who were the low-income group during apartheid has not improved to any great extent. Furthermore, while income distribution is severely unequal, there is also an urban/rural divide in access to social services in which urban areas are furnished with better social security measures and welfare services (Mkandawire 2020).

Namibian government support to older people

Present-day Namibia has special government welfare commitments to older people’s care and wellbeing. There are for instance no healthcare consultation fees for those aged 60 and above (Van Rooy, Mufune, & Amadhila 2015). Government social care ambitions comprise the training of a cadre of social workers at a university level, including social gerontology studies. They are later employed in hospitals over the country (Ananias & Lightfoot 2012). Social security incorporates tax-financed cash transfers (Levine, Van der Berg & Yu 2011). Of particular significance for older people is the non-contributory pension that now amounts to N$1200 per month, on par with minimum wage for domestic workers (MPESW 2019). It is suggested that pensions have dispersing effects beyond the direct economic support to older people themselves (Posel, Fairburn, & Lund 2006). Older people often share their pension with younger family members (Duflo 2003). The distributive strategies generated by these cash transfers are of importance to this paper.

While the importance of housing for welfare in later life is emphasised in research on sub-Saharan Africa, this is often limited to general terms (Aboderin, Kano, & Owii 2017; Apt 2000; Okojie 1988). There are a few elaborations on the topic in an African context. Exceptions are studies of informal housing and ageing (Makore 2018), home and care for older people (Hoffman & Pype 2016), and housing designed specifically for older people (Kotze 2006; Madungwe, Mupfumira, & Chindedza 2011; Mupedziswa 1998; Nyanguru 1987; Pype 2016). There are even fewer research projects about urban ageing in an African context. Some studies have indicated the special harsh conditions for older people living in informal settlements (Aboderin 2004; Aboderin et al. 2017; Apt 2000; Makore 2018).

Family contribution

A point of departure for this paper is that African families use a number of strategies to cope with everyday welfare needs (Frayne 2004; Nord & Byerley 2020; Oduaran & Oduaran 2010). This is valid even if they live in an informal security welfare regime like that in Namibia, where government welfare is available. What the family as a collective can achieve is essential to these strategies (Papadopoulos & Roumpakis 2019). Dimensions of collective ownership and mutual care and support responsibilities are traditionally strong in African extended families, and it has been proposed that they form the basis for the development of social security as regards for instance legal and constitutional rights (Tshoose 2009).

It has been argued that there is a need for an Africanisation of government social protection and welfare, by which they are adapted to the social realities of Africans such as the belonging to and dependence on extended families (Mpedi 2018). This is not without significance in an informal security regime such as Namibia, as formal and informal systems may overlap and merge in certain ways. Elements of indigenous welfare practices are important foundations that can be augmented through integration with formal systems (Patel, Kaseke, & Midgley 2012). An example in Namibia is child fostering, a long-standing tradition of sharing the responsibility for children that has acquired new implications in the era of HIV and AIDS, as the burden of care for orphans is placed on older people (Brown 2011; Kalomo & Besthorn 2018). The family acts in a context of power relations vis-à-vis the state (Papadopoulos & Roumpakis 2019). A central issue in this study is to what extent the Namibian extended families have relegated responsibility for welfare sectors for which they were traditionally responsible to the state, and to what degree they contribute by choices and actions. The proposal that older individuals can make a considerable contribution to community and family welfare is essential to the study (Wiles & Jayasinha 2013). This family welfare contribution constitutes a significant gap in research (Roumpakis 2020). The paper will contribute to filling this gap by an ethnographic approach that comes closer to people’s experiences, considerations and strategies in practice in urban living (Baker & McGuirk 2017). Housing is an important aspect of this.

Housing and welfare

The provision of good quality housing has long been a government measure to improve citizens’ wellbeing and health and to achieve policy goals in other welfare areas (Andersson 2002). In line with this, this study is interested in the significance of housing to people’s welfare strategies and to urban living in later life.

Housing provision in Namibia

The intersections between urban living, housing and the other welfare pillars are explored in the case of the former apartheid township Kuisebmond in Walvis Bay, whose huge port and industries make it a dynamic and economically important place for Southern Africa, not only Namibia (Melber 2015). Kuisebmond was originally a segregated area for Black workers with small housing units provided by the apartheid government for nuclear family tenants. Older people were not allowed to live in these dwellings but were instead relegated to rural areas. These living patterns have now changed, and older people live permanently in Kuisebmond (Nord 2021; Nord & Byerley 2020). However, since Kuisebmond’s residents still remain in economically and socially disadvantaged groups, they rely on services in the abovementioned welfare areas (compare Malpass 2004).

The extreme inequality in distributed income in Namibian society is mirrored in the housing situation in Walvis Bay. Expensive houses on the south-western edge of the town are the permanent residences or vacation homes of the affluent (Figure 1). Kuisebmond is on the opposite – northern – side of the town. Here, the average income is very low, and the majority of the dwellings are small and of low quality (Figure 2).² This stratification of the population mirrors in a most tangible way the massive social and economic inequalities that are embedded in the Namibian Gini-coefficient.

Figure 1. Affluent housing in the southern part of Walvis Bay. Photo by the author.

Figure 2. Street with original township houses in Kuisebmond. Photo by the author.

Kuisebmond has an urban structure, a street grid with detached, small single-family housing units on separate plots constructed by the apartheid government in phases from the 1960s onwards, and then by the Namibian government and the Municipality of Walvis Bay after the town was incorporated into Namibia in 1994 following the democratisation of South Africa. Present-day government housing provision for lower-income strata is the responsibility of the state-owned company Namibia National Housing Enterprise through the subsidised Mass-Housing Development Programme (Figure 3). Self-help projects have followed a similar urban form. The government subsidises the Shack-Dwellers Federation’s saving scheme for micro-loans and a self-help construction programme. There are very few formal rental dwellings (Chiripanhura 2018).

Figure 3. Typical Mass-housing dwellings and urban layout. Photo by the author.

Around 1994, the small township dwellings were sold to the residents at favourable prices. These sales prompted considerable alterations to the township houses and the addition of shacks. Urbanities in the global south are often messy, unplanned, overcrowded and characterised by massive slums and ‘grey spaces’ (Avni & Yiftachel 2014). Kuisebmond does not entirely fit that description; the planned urban form of a former apartheid township lingers on (Müller-Friedman 2008). However, Kuisebmond’s orderly and structured appearance largely conceals other orders of small-scale informalities. The concept of ‘grey space’ is highly relevant here, as it indicates a space where binaries such as formal and informal are blurred (Avni & Yiftachel 2014). Grey spaces in Kuisebmond are alterations to the formal former township houses and the informal backyard shacks that are constructed around these with informal leasehold rights (Figure 4). These additions resist the apartheid nuclear family structure that the apartheid government tried to impose. Of the 60 000 total residents in Kuisebmond, an estimated 29 000 are backyard shack-dwellers (Walvis Bay Municipality 2014). It is claimed that backyarding is an effect of the severe housing shortage and the lack of affordable formal housing available (Chiripanhura 2018). Backyard shacks are illegal and dangerous, but nevertheless ‘accepted’ by the authorities for the lack of alternatives (compare Avni & Yiftachel 2014).

Figure 4. Grey spaces: Backyard shacks behind a former township house. Photo by the author.

Recent research shows that the alterations to former township dwellings and the addition of shacks are to a large extent a family matter. Namibian extended families have turned former township housing into ‘family houses’, intergenerational spaces in which a variety of family combinations appear (Nord 2021). Furthermore, these family relations are underpinned by highly diversified connections and exchanges between rural- and urban-dwelling people (Greiner 2011; Nord & Byerley 2020). Extended families and individuals are the greatest contributors to housing in Kuisebmond, as substantiated by the number of informal dwellings and self-help housing units, which by far outnumber the government’s contributions after independence. This is a most conspicuous contribution compensating for inadequate government contributions.

This is thus the urban situation in which ageing is located in this study. There is no housing policy that specifically targets older people in Namibia, but government- or municipal old-age homes can be found in some urban areas (Indongo & Sakaria 2016). There is a municipality-run old-age home in Kuisebmond that offers housing for independent ageing and is not a care home (Figure 5).

Figure 5. Between the houses in Kuisebmond Old Age Home. Photo by the author.

Understanding welfare in space

Housing in the urban area of Kuisebmond is where welfare services are received and moulded by family and individuals. This is akin to the processes that form a policy assemblage: Government policies are inserted in a dynamic field of shaping and re-shaping processes formed by heterogenous components, in this case government welfare provision, extended families and spaces (Savage 2020). Family welfare strategies are an important agent that may alter government provision and services by forging or resisting policy designs. Their emplacement in a local context of a topography of urban grey spaces further increases these dynamics. Glenn Savage (2020) observes that this does not mean that ‘policies are never formed […] or that policies cannot maintain periods of stability […] Instead, it means that […] from the moment any policy is formed, it is already subject to forms of disruption, challenge and multiple interpretations’ (Savage 2020, 326). It is essential to investigate the localisation of welfare in space by transforming and adapting welfare services in order to understand the particular strategies in a former township such as Kuisebmond. Here, space is relational and dependent on contingent global and local factors where politics and policy are locally shaped (Massey 2005). This theoretical approach provides an understanding of how housing influences the four other welfare pillars in this urban context.

Methods and materials

A case study

The study is a qualitative case study in Kuisebmond, Walvis Bay. Case study research is recommended for urban studies that aim to develop theory and inform policy (Yin 1994). Robert Yin indicates the ‘how’ and ‘why’ questions as pertinent to case studies and puts meaning-making in the centre. Interpretations and strategies of welfare of formerly disadvantaged groups are in focus in the ‘real-life context’ of a former apartheid township (Yin 1994, 13). Kuisebmond was purposefully and strategically chosen for the case study for several reasons, including that there are permanently-living older residents there (Patton 1987). While the percentage of Namibians aged 60 + has been a constant 7.1 percent for many years, the proportion of urban-living older people increased from 17 percent in 1991 to 23 percent in 2011 (Indongo & Sakaria 2016; NSA 2011). In 2017, there were 1350 individuals older than 60 in Kuisebmond,³ corresponding to an estimated 16 percent of the 60 + population in Erongo District (NSA 2011).⁴ Because of constant in-migration, the population of Walvis Bay is young.

Participants in the study were mainly recruited with support from the Erongo Elderly Association, two church clubs for older people, and the Shack-Dwellers Federation in Walvis Bay. Chairpersons of the respective associations and clubs provided names of people who could be approached for participation in the study. They were contacted in their own language and invited to participate by the interpreters engaged in the project. Most of them accepted. Snowball techniques also provided the project with participants via various sources such as other participants, local NGOs and civil servants.

Data collection methods

The case study methodology aims at investigating a case in detail and may include a number of data collection methods. Ethnographic data collection methods such as interviews and observations were supplemented with photographs and document studies (Baker & McGuirk 2017). A total of 37 interviews were carried out with 43 older people who were interviewed individually or in couples. An ‘older person’ was not a pre-defined category but was sought for in the empirical material. Twenty-three women and 20 men between 49 and 95 years of age were interviewed. Six couples were interviewed together. An identification of being older was found from the age 55 and up among the participants. An interpreter was used in most of these interviews. Key informants were also interviewed: two social workers, one officer from the town planning office, two activists in the Erongo Elderly Association, one community carer, and the matron in the old-age home. Participation was voluntary and the interviewees were ensured anonymity. Interviews aimed at capturing reasons for living in an urban area, the everyday lives of older people, their use of spaces and personal history. They included themes such as everyday habits, housing issues and cohabitation. An interview guide with questions was used to ensure that comparisons could be made between interviewees. Interviews were carried out in the individuals’ homes or workplaces. They were recorded and transcribed verbatim by the interpreters. Observations were carried out in the homes by walking with the interviewee(s) indoors and outdoors. Photos, sketches and fieldnotes were used to document the homes. Information about Kuisebmond was collected via observations, drawings and interviews. Observations were made of church services, club meetings and street life in blocks where many older people lived.

Analysis

Interviews, fieldnotes and visual documentation were subject to analyses that would reveal a policy assemblage in which various actors, human and material, contribute to the emergence of welfare strategies. In line with the proposed methodology by Tom Baker and Pauline McGuirk (2017), the tracing of sites and situations for welfare was significant to this analysis. The theoretical framework thus posited policy assemblages in urban grey spaces defined as urban space together with agglomerations of formal and informal welfare (Avni & Yiftachel 2014; Savage 2020). The analysis focused on the intersections between housing types and the other forms of welfare suggested by Malpass (2005): education, healthcare, social security, and personal social services. These intersections were explored by an open analysis technique that revealed the relationship between urban spaces, family patterns and habits, intergenerational exchanges, and informal as well as formal welfare activities that moulded and adapted social policy and services (Baker & McGuirk 2017).

Results and discussion

Older people in Kuisebmond live in three principal forms of housing: formal housing, an old-age home, or informal housing solutions such as backyard shacks or other rentals. Of the 1350 individuals over 60 in Kuisebmond, the majority live in formal housing. Only 28 individuals (eight women and 20 men) lived in the old-age home at the time of the fieldwork, and about 150 lived in the estimated total of 8700 shacks in Kuisebmond (Walvis Bay Municipality 2014).⁵

Formal housing in Kuisebmond in the study refers to former township housing and to self-help housing constructed after 1994. Fifteen households in the study lived in formal housing; this comprised men and women between 59 and 95 years of age (see Table 1 for details). With one exception, they were all retired and received the non-contributory pension – either Namibian, South African, or both. One man was about to retire and was also expecting an occupational pension.

Table 1. Details of owner participants in the study.

House owner	Year of birth and age at interview	Arrived in Walvis Bay	Transformations	Cohabiting family in the main house	Number of backyard shacks and inhabitants
Imago (widow)	1920 95	1962	extended	four grandchildren one grand grandchild	One unused
Theodore & Taati	1932 85 1947 70	1959	extended	none	None
Laine (widow)	1936 81	1955	not transformed	none	Several • son, daughter-in-law, grandchildren • other renters
Johannes (widower)	1939 78	1956	not transformed	none	Four • two grandchildren, one with girlfriend • two unrelated renters
Zacharias & Nelago	1940 77 1950 67	1960 1972	extended	one grandchild	Two • unrelated renters
Mary (widow)	1940 77	Born in WB	extended	one son one niece	None
Magda (married)	1941 75	1950	extended	husband one son	None
Jacob & Gertrude	1944 73 1947 70	1965	extended	grandchildren	None
Margaret (married)	1947 70	1958	not transformed	husband, son brother’s son cousin’s granddaughter	Four • nephew • cousin’s son and his son • two unrelated women and their families
Halina (divorced)	1950 67	2003	self-built	two daughters two sons two grandchildren	None
Gabriel (not retired, married)	1953 64	1983	self-built and extended	wife two daughters two grandchildren	One • nephew
Evelina (unmarried)	1955 63	1963	extended	one daughter with her daughter daughter-in-law	Two • one daughter, • one granddaughter with her daughter.
Victoria (married)	1955 62	1973	rebuilt	husband six grandchildren brother’s grandchild	Six • not related renters
Amanda (married)	1958 59	1990 ca	rebuilt	husband one daughter one niece	Eleven • not related renters
Augusta (married)	1959 58	1992	not transformed	husband two children	Two • a very old mother • one child

Most of the interviewees in the formal housing group had purchased their township houses around 1994. People who arrived in Kuisebmond after that had to find lodging in other ways. Two had acquired their formal houses in Tutaleni, a self-help housing project. However, most latecomers lived in shacks or other rentals. As a group, they were younger than those living in ordinary housing – between 49 and 65. Most of them were still active in the labour force (see Table 2 for details). The housing situation for this group included various housing solutions. Six households in the study lived in backyard shacks, and three lived in other rentals. One woman lived in a reconstructed garage in a backyarding situation of sorts on her daughter’s premises. Although these housing options were very small in size, there was cohabitation with other family members in all but two instances: One man lived alone in a flat, and one woman lived in a shack.

Table 2. Details of renter participants in the study.

Renter/s	Year of birth & age at interview	Arrival in Kuisebmond	Type of rental	Cohabitants
Adam (married)	1953 64	1975	Rental flat municipality landlord	none
Samuel (married)	1953 64	1975	Rental single quarter – one bedroom employer landlord	one adult son
Virginia (retired, divorced)	1953 64	1985	Rental garage daughter landlord	▪ In the main house: one daughter, her husband, her four children and one foster child one younger daughter and her child ▪ a son in a backyard shack.
Barnabas (separated)	1954 63	1974/1984	Rented house	one daughter 10 shacks ▪ one son ▪ unrelated renters
Selma (divorced)	1959 58	1994	Backyard shack unrelated landlord	three grandchildren one niece cousin’s child
Janet (married)	1958 58	1994	Backyard shack niece landlord	husband one daughter one granddaughter
Frederic (married)	1963 55	2005	Backyard shack cousin landlord	one adult son
Aune (unmarried)	1962 55	1993	Backyard shack unrelated landlord	none
Immanuel & Nelago	1965 53 1968 49	1996	Backyard shack unrelated landlord	one daughter, 14 years old

The first section in the following analysis reveals how welfare is received and optimised in relation to the different housing options. The next section regards the strategies that appeared at the intersection of urban living, access to good quality housing and welfare services.

Three welfare pillars: Personal social services, social security and education

Three welfare pillars could be discerned in formal housing: personal social services, social security, and education, appearing in localised versions as family mutual care, non-contributory pensions and access to schooling. Government services (or the lack thereof) and housing formed a structure that was interpreted into patterns that brought older and younger people together or separated them by choice. These localised welfare interpretations constituted family obligations, optimising strategies and housing adaptations.

The intersection of housing and mutual family care

In many cases, the grey spaces of family houses, which often included both formal housing, shacks, or other rentals, made accommodation of a variety of cohabiting family members possible. No older person in formal housing in the study lived alone. The patterns of cohabiting with older people are complex and might include other cohabiting relatives than the children (Hoffman & Pype 2016), as was the case in this study. These results thus did not confirm the increasing numbers of older people in Namibia who live alone, who are interpreted as a representation of the ongoing disintegration of the indigenous system of family care (compare Indongo & Sakaria 2016).

The most common relatives with whom the informants cohabited were adult children and/or grandchildren. Six households in formal housing had three generations living together, and two had four. Evelina had a daughter and daughter-in-law living inside her former township house, while another daughter, her granddaughter and the granddaughter’s infant daughter lived in shacks in the yard. The oldest woman, Imago, who was 95 years old, lived with four adult grandchildren and one great-grandchild. These family patterns were made possible by the grey spaces of formal housing and informal additions such as shacks, and they paved the way for mutual familial care. Although rental housing options in the study were of substantially lower quality than ordinary housing, they nonetheless repeated to some extent the household patterns in formal housing by gathering two or more generations under one roof. Janet lived with her husband, one adult child and a grandchild. Selma had six minors living with her, three of whom were her own children or grandchildren.

Like Selma, nine interviewees in formal housing – women and men alike – had one, two or several grandchildren or other, more distant, underaged relatives (for instance nieces and nephews) staying with them for whom they had assumed caring- and raising responsibilities, according to traditional family obligations of child fostering (Brown 2011). The ages and genders of these children varied from household to household – there were boys and girls from six years of age to upper adolescents. In a few cases, the grandchildren were young adults. There was also a mutual exchange of help and support in that co-habiting minors helped older people with errands and housework, ‘unless they were too young’, as one interviewee said. Research has revealed that many young people contribute to the care of older people by performing household work, healthcare and personal care duties (Evans 2010). Mutuality appeared in various shapes in minor everyday acts. Johannes mentioned the importance of his two grown-up grandchildren’s help with picking up medication from the pharmacy. One of them was baking bread in his grandfather’s kitchen while his grandfather was being interviewed, although he lived in a shack behind his grandfather’s formal township house.

In nine of the formal housing households, the interviewees’ adult children also lived in the household. The mutual benefits were that the older individual received support from the children, but also that the older homeowner provided housing to adult children in formal housing or in shacks, distributing mutuality in grey spaces. However, cohabitation did not always seem necessary for mutuality of support. Job migration may have been a reason why grandchildren stayed with their grandparents. The older interviewees spoke of the various places around the country where their children lived and the occupations they had. No one in this group mentioned any deceased children or children living outside Namibia, which could well have been possible. The study results indicate that some children had assumed responsibility for support of their aged parents, but from afar. Mary, 77 years old, commenting on the fact that generations lived separately, said ‘but when they [the children] get the chance then they send a bit of this and that’. The children’s absence was thus not necessarily a sign that they neglected their parents, but job migration may have been a prerequisite for their support.

Child fostering and job migration were joint strategies where contributions from different family members coalesced into a policy assemblage in which the non-contributory pension and housing were important components. They made it possible for the older person to let grandchildren live with them and minimised the risk of associated economic commitments if the absent children’s contributions failed to appear.

Another caring aspect of grandchildren constituted giving access to schooling.

Education and urban living – quality optimisation

A common trait for older people living in formal or informal housing was that all of their school-aged grandchildren and children were attending school. Access to urban housing intersected with access to education to generate a policy assemblage where quality optimisation as regards education seemed to be the goal. As an urban area, Walvis Bay offers eight primary schools, six secondary schools and five private schools (Unhabitat 2011). In some cases, the quality of these schools was considered higher than the schools in the rural areas. This may well have been a correct assessment, since rural schools are indicated as facing quality challenges (Julius & Amupanda 2017). Halina stated explicitly that access to a school for her children in which English was the language of education had been the main motivator for the family’s move to Kuisebmond from Fransfontein, where Oshiherero, a local language, was the language used in teaching. The family had thus avoided the very localised version of welfare. Margaret had her cousin’s grandchild staying with her for schooling reasons. She said that ‘his father has a house in the north, but the child’s school is here’. Selma, who lived in a backyard shack with six underaged relatives, said that they all went to school, which was also observed. Access to education, one of the welfare pillars, contributed to a policy assemblage where grandchildren lived with their grandparents regardless of housing quality. However, older people, as owners of formal houses, could easily accommodate the children in decent conditions conducive to successful studies. Formal housing could offer a higher quality study environment than informal housing. Selma for example shared five beds lined up in a row with the six youngsters in the rear of her shack. Sharing beds could seriously disturb sleeping quality, and this cramped situation made homework difficult. Shack-dwellers nevertheless organised space for children’s education. The nuclear couple living in a shack had organised a minimal private space in their shack for their daughter to sleep and study.

The above examples show how housing was used to optimise mutual care practices and welfare services such as education. However, cases arose in the study where these strategies were neither possible nor desirable. The old-age home then appeared as a housing option.

Withdrawal from mutual care – the old age home

George Mpedi (2018) cautions against sentimentalising the traditional family systems, maintaining that they have imperfections and limitations when it comes to carrying the burden of social protection. Laine, who was 81 years old, had experienced first-hand that living with children did not guarantee older people support. Her son, who lived in a backyard shack on her premises, had neglected to pay rent due to alcohol abuse, which brought the family close to losing the house because of outstanding mortgage payments for which Laine was responsible. A complex policy assemblage of formal and informal support outside the family was mobilised to help her. She had received support from the government’s social services and from the Erongo Elderly Association when her situation had become critical, and she could ultimately keep her house. One of the social workers interviewed indicated cases of children abusing their aged parents as situations in which they would intervene (compare Ananias 2014).

A special case where grey spaces of formal and informal housing were combined for the sake of familial care was that of Augusta, who lived in a former township house with her very old mother, 95 years old. This situation coincides with the main premise for social policy, that is, the family is the prime caregiver of older people in an African context (compare Oakley 1998). Augusta and her mother were emplaced in a policy assemblage of formal and informal housing and heavy informal care provision for which she had no formal support. She had recently moved her mother from the nearby town Usakos to be able to care for her daily. Her mother – who was blind, probably suffering from dementia, and bedridden – was lying on a mattress in a shack in the yard, and her daughter fed and changed her nappies. This was obviously not an optimal caring situation, but Augusta’s caring strategy made up for lacking home care services or residential care, which was only available at a high cost in Walvis Bay.

Both of these cases challenge the mainstream view that there is no need for institutionalised care facilities from various angles. A common notion indicated in research is that older people living alone without care is a challenging divergence from traditional family-based obligations (Oakley 1998), and that old-age homes are primarily for childless older people or older people whose families do not take care of them for various reasons (Hungwe 2011; Madungwe et al. 2011; Oakley 1998; Pype 2016). Some of the 28 individuals in the study who lived in the Kuisebmond old-age home modulated the image of living alone in later life, however. Seven of the men in the old-age home in the study were either childless or had lost contact with their children. One man spoke in an interview of his relief when he was given access to housing in the old-age home when his children did not want to take him in. One man avoided my interview questions and instead told a long story about the painful loss he had felt when his son died. One should not underestimate the value of the solution afforded to these men by the old-age home, even if they were a minority. Most residents had children, some of whom even lived in Kuisebmond or elsewhere in Walvis Bay. Occasionally they came to visit, according to observations. This naturalised and defused the contested government- provided housing for older people and turned it into a new housing alternative in a policy assemblage of welfare.

The government-provided old age home contributed to a policy assemblage where withdrawal from mutual care was possible. Some of the older people in the home said that they had chosen to live alone in the home against their children’s will. One woman stated: ‘I do not want to be a babysitter for my grandchildren’. In the interview, she talked about how she had been left alone with the toddlers for several days – a responsibility that she had found too great. Another woman said: ‘I don’t want to be involved in my children’s matrimonial life’. Conflicts between children and older parents that result in a parent moving are not unusual (Pype 2016). Safeguarding of the older person’s private life, such as giving the opportunity to leave an abusive child, as in Laine’s case, might pave the way for old-age homes as a housing option in the future. Another woman living in formal housing with an abusive alcoholic husband of whom she was obviously very scared nonetheless rejected the old-age home as an option. The option to opt out of both care responsibility and care reception was also made possible by the fact that the rents in the old-age home were subsidised and thus affordable with the small pension, which in turn links the non-contributory pension to this policy assemblage. However, according to the matron and the municipality clerk, the old-age home offered independent living for older people. Some of the residents were aware that they would have to go to live with their children if they needed 24-hour care. The old-age home was thus not an option for Augusta’s mother, even though Augusta might have needed support with her heavy caring efforts.

This analysis above of mutual care, care withdrawal, and housing shows how a site – Kuisebmond and its housing alternatives – is an arena for policy assemblages where intergenerational relations receive and shape welfare services. In the next section, we will connect to Savage’s (2020) discussion of relational space, in which spatial relations reach far beyond the borders of Kuisebmond and link with far-away spaces for the same reasons.

The welfare pillar healthcare – housing and urban or rural living?

This section highlights how access to quality housing influences the choice of rural or urban living in later life. Many participants in the study stayed in rentals and lived far from their immediate families. Fredric lived with his adult son, while his new wife resided in the rural north, where his other children also lived. This was also the case for Samuel, who lived in single quarters provided by his employer, sharing a bedroom with his adult son and kitchen facilities with other workers. Adam lived alone in his flat, his wife was in Windhoek, and a daughter lived in Kuisebmond. Almost all of these renters were still of working age but approaching and planning for retirement. In many cases, this included plans to return to a rural area, which largely appeared to be a simple choice with no alternative. Interviews also revealed alternative considerations that shed light on policy assemblages in which housing quality, access to housing, healthcare provision and the choice of rural or urban living were included.

Access to good quality housing

Adam and Janet, two interviewees from different households in the group of renters, had recently changed their minds about moving back to the rural area upon retirement and were considering remaining in Kuisebmond. Access to good quality housing seemed to be a decisive factor, since they had just received plots through their memberships in the Shack-Dwellers Federation, which included access to a low interest bank loan that could finance the construction of a proper house. This constituted a major spatial change that altered the ways in which possibilities emerged and provided a redefinition of their places in the urban grey spaces, particularly for Janet, who lived in a shack. She expressed her gratitude during the interview, stating that now she would only return to Khorixas, her place of origin, to be buried. Obviously, access to proper, good quality housing might challenge the idea of circular migration and a return to rural areas (Nord & Byerley 2020). Housing as a service provided according to social policy could thus cut the ties to far-away places. However, access to good quality healthcare could have a similar effect.

Access to healthcare

Adam’s thoughts and attitude were different, and he was less determined to stay in Kuisebmond than Janet was. He did not see the new plot as a strictly personal opportunity but as a family asset, although not for reuniting his family, which was scattered at the time of the interview. He considered passing his newly acquired right directly to his daughter and moving up north. Although he preferred the countryside, he had doubts about living there. In his opinion, the rural policy assemblage he imagined had serious shortcomings. He was concerned about difficulties accessing the healthcare for which he foresaw a need as he grew older in the near future. He stated that his farm up north was 20 kilometres from the closest hospital, forming an unviable policy assemblage. Similarly, Johannes – who was living in a former township house – indicated access to healthcare as the main reason for staying in Kuisebmond. He said in interview that ‘I am here only because the doctors are near’. He nurtured plans to live with his sister in Otjimbingwe but had to date refrained from making this move. In his case, access to healthcare seemed more important than his comparatively high housing quality. Other research has shown that access to healthcare may be a decisive factor when it comes to individuals choosing to age in a rural or urban area (Makore 2018), sometimes bridging vast distances over international borders (Hunter 2011). Here, links appear between housing and urban, or alternatively rural, welfare provision. A policy assemblage comprising housing and necessary welfare provision in an urban area that is perceived as better can outweigh the advantages of rural living or the disadvantages of urban living in later life.

The choice of where to live is a strategic issue as regards welfare in later life for the older individual personally. However, older persons’ choices appear in another way for authorities that provide various services such as healthcare. Accessing healthcare in rural areas can be very difficult, with long distances to healthcare facilities and high transportation costs (UN 2017; Van Rooy et al. 2015). Housing in rural northern Namibia consists of sparsely scattered farms, and a car, preferably with 4WD, is a necessity even in non-remote areas. Many interviewees came from former ‘native reserves’ in the southern part of Namibia, some of which are poorly serviced and remote. Apart from the major motorways, roads are often sandy dirt tracks. Such conditions may create a dependence on younger family members – if they are available at all – considering the many younger people who move to urban areas (Greiner 2011). Access to medical care was an important aspect that appeared in the study. Most participants in the oldest group were being treated for type-2 diabetes, high blood pressure, or both. Aune showed her prophylactic drug for HIV during the interview. Participants stated in interviews that they often visited the hospital and healthcare centres and needed regular access to a pharmacy. There was one clinic and one pharmacy in Kuisebmond, so the area offered convenient access. Some interviewees, like Johannes, said that children or grandchildren helped them with the purchase of their medication, while others could pick up medication themselves. In rural areas – when living on a farm for instance – this would be very difficult.

Adam, who had expressed doubts about returning to the rural area because of possible difficulties accessing medical care, had good reason: the urban area formed a more convincing policy assemblage because of the more efficient and accessible healthcare. While this was obviously not always an easy decision, it was strategically important. Housing quality and healthcare services constituted a crossroads where access to both high quality housing and healthcare were optimised in the interviewees’ perceptions.

These participants’ considerations as regards access to housing and healthcare show the mobile qualities of a policy assemblage and how it can connect near and remote places (Baker & McGuirk 2017). One component can obviously change an older person’s strategic responses to welfare and housing.

Discussion

Many family welfare strategies highlighted in other research were present in this study, such as child fostering (Brown 2011), job migration (Posel et al. 2006), and young people’s care and support of older people (Evans 2010) These strategies were combined with government welfare services in new interpretations integrated into the urban area of grey spaces (Avni & Yiftachel 2014). The demise of the extended family’s welfare obligations observed elsewhere did not appear clearly in this study (Aboderin 2004; Indongo & Sakaria 2016). Instead, the study’s results suggest that the extended family is still an important agent in producing and consuming welfare. The study also indicates that the family still very actively assumes responsibility for older family members (Ananias 2014; Obrist 2016). However, older people are not merely receivers of care and support in extended family life: they also contribute to a large extent to helping meet others’ needs (Wiles & Jayasinha 2013). One important contribution is their optimisation of spatial resources by creating a strategic welfare space for the benefit of other family members.

Spatial aspects of local policy assemblages

Strategic welfare space is a space where housing and welfare are optimised. Living space in different scales was used by participants in the study as a strategic advantage to optimise welfare in response to government services or the lack thereof. Older people created multi-dimensional spaces that held together local policy assemblages of informal and formal housing and informal and formal welfare in various combinations, all with the aim of getting the maximum out of the system for the benefit of the whole family. Housing and home space provided an arena for intergenerational mutuality (Kaplan, Haider, Cohen, & Turner 2007). The family house was thus an important space in this respect for its ‘greyish’ capacities, but informal housing and other rentals were also used in similar ways. Spatial strategies also aimed at increasing potential advantages to handle risks and adversities in local as well as on larger scales by translocal optimisation in which distributing resources were vital, often human resources (Frayne 2004; Greiner 2011; Nord & Byerley 2020). Different outcomes were expected from these strategic welfare spaces, such as compensation for a lack of formal welfare (for example, non-existent residential care for older people), supplementation of formal welfare (few child day-care options), and increased quality of formal welfare (choice of school). Not least, compensatory strategies were making up for the inadequate government housing policy and provision and a serious housing shortage (Chiripanhura 2018). As owners of a family house, older people in this study helped employment-seeking young relatives access workplaces in Walvis Bay by opening their homes to them (compare Posel et al. 2006). The older individuals in the study played decisive roles at the very centre of strategic welfare spaces by optimising the family use of space, whether this constituted formal or informal housing.

Urban vs rural living

A variety of housing options appeared in Kuisebmond along with an urbanisation of traditional support systems (compare Baart 2016; Patel et al. 2012). The family house can be interpreted as an urban version of the traditional multi-generational rural homestead (compare Bolt & Masha 2019). The old-age home is a newcomer that seemed to be – at least partly – an alternative option in the study, thus defying what is traditionally regarded as good care for older people in an African context (Oakley 1998; Pype 2016). Perhaps this innovation might be met with a higher degree of acceptance in an urban setting, as it enables older individuals to extract themselves from both care responsibilities and care demands on the family and thus, to some extent, also from family resources, or to find housing even if family support fails to appear (Hungwe 2011; Madungwe et al. 2011; Oakley 1998; Pype 2016). These various motives mirrored both deliberate choices or more or less coerced solutions to an experienced lack of housing (and care) in later life.

A return to rural living is always a possibility even for an individual who has been living in an urban area for a long time. However, urban living appeared in the study as a considerable contender to rural later life by offering an alternative with access to government-provided services and high-quality housing. Rural provision of welfare is a considerable challenge for the Namibian government, and government welfare services are often better in urban areas (Mkandawire 2020; Van Rooy et al. 2015).

Pensions, urban living and housing

Housing – formal or informal – was important, and non-contributory pensions also seemed to be highly significant for the joint family endeavours in strategic welfare spaces in the study. Devereux (2007) claims that an argument against overly generous pensions is that they might undermine family responsibility for the care of older people. The results of this study suggest that ‘pensions may help consolidate the family, rather than encourage the disintegration of interdependence within families’ (Pelham 2007, 17). Since pensions have intra-familial distributive effects (Duflo 2003), they may have played an important role in facilitating the formation of multigenerational strategic welfare spaces. First and perhaps foremost, non-contributory pensions would allow an older individual to pay for housing where relatives could be received. Secondly, pensions and housing may allow older people to assume caring responsibilities for underaged relatives and allow the middle generation to migrate elsewhere for work (Posel et al. 2006). This could be regarded as an investment and an insurance against risk on the part of the older person, hopefully resulting in long-distance financial support from the children while safeguarding against unmanageable fallouts should their support fail to materialise. Many of the adult children lived elsewhere while the older generation took care of minors. Thirdly, pensions might facilitate urban living due to the simple fact that everyday life demands access to cash in that environment (Aboderin 2004).

Impact on policy

Concurring with other scholars’ suggestions, this study indicates the need to consider policy adjustment in housing as well as welfare provision in order to accommodate families’ shared needs (compare Mpedi 2018; Tshoose 2009). The mass-housing programme currently offers housing to individuals and nuclear families and not extended families, as has been observed in South African research (Bolt & Masha 2019). Mass-housing is not an alternative for many families, as it is too expensive and not flexible enough to accommodate changing familial circumstances. Furthermore, it is not available in rural areas. The rental housing market is very weak (Chiripanhura 2018). The results of a study of family life and welfare strategies like this provide important feedback to the present supply and how it could be adapted to accommodate the needs of extended families.

In line with researchers who have indicated the strong mutual shaping of welfare and housing (Kemeny 2001; Malpass 2004; Ronald 2008), this paper draws the conclusion that housing in this study is embedded in the welfare regime as a strategic welfare space. The roles and contributions of the family and government to housing provision differ in different welfare regimes (Ronald 2008). The variety of family welfare strategies included in the concept of strategic welfare space is an outcome of an informal security regime like that in Namibia, where family welfare contributions are essential although governmental provision is comparatively generous (Abu Sharkh & Gough 2010). In this study, family contributions to housing weighed more heavily than contributions from the government. The family’s informal contributions also improved government welfare by including and adjusting it to strategic welfare space. This study concludes that it is essential to understand the links between formal and informal housing and welfare provision in order to pave the way for families’ strategic welfare spaces.

Acknowledgements

The study was financed by Riksbankens Jubileumsfond under Grant no. P15-0140:1. The author thanks Erongo Elderly Association and the Namibian Shack-Dwellers Federation for support and help carrying out this study. The authors also thank Mrs Paulina Junias and Miss Revival Gâo-o Resendez for assistance, translations and transcriptions of interviews.

Disclosure statement

No conflict of interest has been declared by the authors.

Additional information

Notes on contributors

Catharina Nord

Professor Catharina Nord is an architect SAR/MSA and professor in spatial planning in BTH, Blekinge Institute of Technology. She has a PhD in built environment analysis. She has conducted research on African health issues and spaces for ageing for many years.

Janet Ananias

Dr Janet Ananias is a senior lecturer in social work at the University of Namibia. For her PhD in social work, she looked at ageing in rural and urban Southern Namibia.

Notes

1 https://data.worldbank.org/indicator/SI.POV.GINI?view=map&year=2015

2 The average income per households in the affluent areas in Walvis Bay is between N$15 000 to N$30 000. In formal housing in Kuisebmond it is N$5000-6000 per month, while in informal housing the average is N$1044 (Walvis Bay Municipality 2014).

3 According to the Elderly Association in Erongo.

4 Of the 150 000 people in Erongo, 5.5 percent are older than 60 (approximately 8 250 individuals). The 1 350 in Kuisebmond constitute about 16 percent of these.

5 There is no source providing the exact number of people older than 60 living in backyard shacks in Walvis Bay. An unpublished survey carried out by the municipality in 2014 of 6 500 households living in shacks shows that most of the inhabitants are substantially younger, but this survey is incomplete since it only indicates the number of dependents of the interviewees and not their age. About 100 informants of the total of 6 500 surveyed were older than 60, which gives an extrapolated figure of 150 in the total of 8 700 shacks. This is, however, a highly uncertain figure based only on the assumption that the oldest person is the head of the household. This survey does not include other types of rental options.