Population Ageing and participation of the elderly in the labour force of South Africa

Vieillissement et participation à l’activité économique en Afrique du Sud

Gabriel TATI, Population Studies Programme, University of the Western Cape Afrique du Sud

Résumé

Ce papier a pour but d’apporter un éclairage statistique sur les formes de participation des personnes âgées dans la main d’œuvre. Le contexte d’analyse est celui du milieu urbain de l’Afrique du Sud. Les données de recensement et d’enquêtes utilisées montrent que dans l’ensemble les personnes âgées ont un taux d’activité relativement fort surtout aux âges compris entre 60 et 70 ans. Bien qu’une décroissance est observée au-delà de 70 ans, il n’en reste pas moins que certaines personnes âgées continuent à travailler à un âge qui frôle le centenaire. Le taux d’activité est particulièrement élevé pour les personnes de race noire. Toutes races comprises, les hommes sont plus actifs que les femmes jusqu’à l’âge de 75 ans au-delà duquel les femmes dominent dans l’activité économique. Les statuts dans l’emploi les plus représentés sont ceux des travailleurs domestiques, des travailleurs indépendants et des salariés occasionnels. Le statut dans l’emploi est influencé par la race avec les noirs, hommes et femmes, fortement représentés dans le travail domestique. Il n’y a aucun blanc travaillant comme domestique. Ce constat reflète en quelque sorte les inégalités qui persistent dans les relations interraciales en Afrique du Sud. L’analyse fait aussi ressortir la dominance des emplois précaires, non qualifiés (en dépit de la longue expérience professionnelle accumulée dans le temps) surtout au sein des personnes de race noire et des couleurs. Dans l’ensemble les personnes âgées génèrent un faible revenu de leur travail lorsqu’on compare à la moyenne nationale. Les conditions de travail des personnes âgées suggèrent une position de vulnéra¬bilité sur le marché du travail. Sur la base de ces résultats empiriques, le papier propose des directions à prendre en matière de politique sociale notamment sur les mesures de protection sociale envers les personnes âgées.

Mots-clés : vieillissement, Afrique du Sud, Main-d’œuvre, activité économique, protection sociale, population.

Introduction

The paper examines the participation of the elderly in the labour market of South Africa. The concept of active ageing is central to this paper due to the fact that it has been increasingly used in recent policy agendas to address not only the needs for better support but also their willingness to continue contributing to economic life (World Health Organisation, 2002 cited in Kinsella and Phillips, 2005 ; Stloukal, 2007 ; Timonen, 2008). With this in mind, the paper stems from the recognition that many old people in South Africa, irrespective of racial group, prefer for various reasons to delay their withdrawal from the labour market and remain economically active in order to supplement the old age pension they receive from the state. The analytical focus is placed on the urban setting in recognition of the fact that South Africa is dominantly urban with a relatively strong industrial base and issues around retirement at legislated age or withdrawal from the labour force apply conveniently to the urban setting because of the structured forms of its economy.

First it is important to provide a succinct account on the stage reached world wide in the importance of issues around ageing. A starting point would be the recognition that the policy debates on ageing have gone a long way from the Draft Declaration on Rights of the Elderly in 1948 proposed by Argentina to the Madrid Plan of Action on Ageing (known as MIPAA) adopted in 2002. The Madrid Declaration on the aged persons placed the social issues related to ageing amongst the most pressing challenges in the near future. The Plan calls upon governments to morally and politically commit themselves to the incorporation of ageing in all social and economic development policies, including poverty reduction strategies (Timonen, 2008 ; Stloukal, 2007). From the angle of demographic change, ageing is now a global process as it affects both industrialized and less developed nations (Ferreira, 1992 ; Kinsella and Phillips, 2005). World wide, it has been recorded not only a substantial increase in the number of older people in relation to younger people within the population but also persistent needs of income among the elderly (Timonen, 2008). These are the two aspects of ageing examined in this paper. At the continental level and in line with the social policy focus that comes with the MIPAA, the African Union Heads of States and Governments adopted in 2002 the African Union Policy Framework and Plan of Action on Ageing. Among other innovative recommendations, MIPAA emphasizes the right and potential of older people to participate actively in economic and social development. Specifically, the African Union Plan of Action provides recommendations on how to improve the quality of life and conditions of the elderly in the area of employment and income security (Stloukal, 2007). Consumption in old age is financed, among other sources, through own labour income and a possible intervention in that direction is to promote labour force participation through adequate work force polices, access to production inputs, credit, etc (Timonen, 2008 ; Stloukal, 2007). Thus, the so-often classification of old people as exclusively consumers and socially dependent group on pension funds or other public transfers may not be a relevant policy tool in sustaining this family role in the context of South Africa. Interestingly, some recent research works on the ageing population in developing countries suggest that support from children does not substitute for elderly parents’ need to work (Cameron and Cobb-Clark, 2005). In South Africa, government is increasingly looking at way of linking old age pension to self-employment (Presidential State of the Nation, 2011). While several other similar actions have been initiated since the Madrid Assembly, research and policy responses are lagging behind the need to strengthen economic security in old age. Referring specifically to the urban context of South Africa, one must admit that limited information is available on the patterns of the elderly participation in the labour force. Very little is known about the circumstances under which the labour market provides employment opportunities to old people. Hence it is of interest to examine the forms of participation amongst the aged persons. In addition to providing trends in the estimates of heads of households among the elderly across the nine provinces of the country, the article analyses the socio-demographic dimensions of old-age in the labour force using various sources of data. Identifying the position of these old people within the labour market might provide critical referents to the development of sensible social policy responses. Four questions are of importance in the analysis to follow : what types of jobs are likely to be taken up by older workers ? In which sectors are they actively involved ? To what extent does employment status differ according to race and gender ? What is the profile of participation with age advancing beyond 60 ?

The paper is divided into four sections. The first section provides some theoretical perspectives on the involvement of the elderly in economic activity. This section is followed by a description of methods and data used. The third part provides some international perspectives on the ageing dynamics and the growing importance of labour force participation among the elderly in South Africa. An assessment of the different forms of participation in the labour force provides the substance of the fourth section of the paper. A conclusion contains the major recommendations.

Setting the debate beyond the age concern

It is important to observe that in the context of South Africa, the question around the involvement of old people in the labour force find its significance in the recent debates surrounding new social measures to put in place for the protection of the most vulnerable groups. The elderly represent an important segment within the vulnerable social groups as the welfare system is getting hard to sustain efficiently. Views are being expressed regarding the option of pushing up the legislated age of retirement which is at present 60 for active women and 65 for active men, respectively. To some extent, these debates reflect the steps being taken globally around the extension of the retirement age in those countries with increasing demographic deficits in the working ages (Timonen, 2008). South Africa, as is the case in most parts of the Africa continent, is a country where such a measure has more to do with the precarious conditions under which most of the older people live than any demographic considerations. All racial groups considered, it is argued that most South Africans are financially ill prepared when it comes to retirement. Less than 6 percent of South Africans retire with enough income to support themselves (Mail and Guardian, 2009). Research by Momemtum and Sanlam says only 60% of people in formal employment are covered by a pension arrangement, whereas in the informal sector that number drops to 10%. Studies have also established that most workers’ savings or contributions to pension schemes at retirement age fall below the minimum required for a decent life style. Besides, the adding negative effects of HIV/AIDS pandemic to poverty motivate people in old age to work as they take on more responsibility as heads of household to care for family members left behind by the deceased ones (Noumbissi, 2001 ; Legido-Quigley, 2003 ; Madhavan and Schatz, 2007). Although some of these older persons may be receiving pension, they still need to work in order to earn enough and live out of deep poverty. Other reasons that incite aged persons to maintain their participation in the labour force are found in the deficiencies or insufficiencies of all types which affect the welfare system. Such deficiencies provide some impetus to older people to work on their own account or for salaried work, to take care of orphans, and to provide for other means of livelihoods or social support for their household members (Madhavan and Schatz, 2007). Several analysts concur with the view that older people in more developed countries are generally less likely to work than those in less developed countries. In the former, financial incentives for early retirement enable many older workers to leave the labour force making room for younger workers. In developing nations, the situation is quite different. Participation in old age in the labour force is often out of necessity even for urban dwellers (Kinsella and Phillips, 2005). In Turquia for example, most of the persons in old age considerably delay their retirement age from the labour market due to the shortfalls in transfers from their children (Cameron and Cobb-Clark, 2005). In the context of South Africa, inequalities of all types affect particularly household-headed by old people despite a well established state-sponsored welfare system (Legido-Quigley, 2003). The old age pension allocated by the state does certainly reduce the vulnerability of elderly household to income poverty, but it has a very limited effect. It is often argued that this pension cannot be qualified as the sole means of eradicating poverty in old age since its value is inadequate to meet other needs (Lam et al., 2006).

The legislated age of retirement is 60 and 65 years, respectively, for those women and men working for salary in the formal sector in South Africa. Earlier retirement is not however quite common as people reach that age. This is particularly the case of those old persons in the self employment sector who cannot afford to retire earlier because of lack of pension income. Those working for wage or salary may also opt for delayed retirement because of the relatively small income they get from pension. It is in recognition of that complex dilemma faced by old workers that Lam et al. (2006) argued that decision of the elderly about work and retirement are made in the complex set of circumstances-high unemployment and need for multiple sources of income- resulting from the levels of inequality imposed under the system of apartheid and paradoxically unaddressed under the post-1994 dispensation. The authors also concurred with the view that an important feature of the South African old pension system is that receiving the pension is not necessarily incompatible with working. One of the explanations to this duality is that the mean test-based pension for old age does not preclude work and because the rules of the system are often flexibly applied. Under such circumstances where pension does not compete with work, it may not sound reasonable for some people to continue working beyond the pensionable age. However this not the dominant aspect as it was also observed that for individuals working and receiving the pension at the same time, employment rates are much lower among pension recipients than among non-recipients (lam et al., 2006). Some old persons are impelled to work because they head households and do not have any other forms of social support either within the family or from the state. It is therefore important the see the household as space where the old age labour supply fulfills certain inter-generational or family functions.

Conceptual framework

At the outset, it must be said that although the field of ageing has attracted, and continues to do so, a huge body of studies, there is not as such a focused theory on the participation of the elderly in the labour market (Timonen, 2008). Diverse social theories have been developed on ageing since the 1940s but they hardly discuss the labour market outcomes among the elderly. This is not unique to the area of labour market analysis. The theoretical deficiency is also observed for the study of other aspects of ageing (Timonen, 2008). Hence, one must work out first a theoretical framework for hypothesis testing from the existing ‘theories’, as Timonen (2008:9) put it, that ‘seek to make sense of the social processes related to ageing’. These theories are somewhat silent and negative about the participation of the elderly in the labour force. For example the arguments developed in the disengagement theory associate ageing with an unavoidable withdrawal from the many roles and duties one has to fulfill during working age life. The theory argues that old persons become inactive and have no obligation to contribute to the production of goods and services in the labour market. Such a view that argues inactivity (or uselessness) in old age is also present in three other major theories of ageing namely the modernization theory, the structured dependency theory, the political economy of ageing (Timonen, 2008). Kinsella and Phillips (2005) provide a critical review of such theories of ageing with special reference to role, activity, continuity, subculture of ageing, modernization, exchange and feminist gerontology theory. Rather than testing a specific theory, the arguments developed in this paper are built around the activity theory that opposes the ‘theories’ mentioned above. The activity theory stems from the construct of active ageing (WHO, 2002 cited in Kinsella and Phillips, 2005) that brings together the concepts of successful ageing and that of productive ageing. These two concepts have different connotations with the second having a bearing on labour force. According to Kinsella and Philips (2005), active ageing encompasses the broad issues of coping and adaptation in later life. Some old people may succeed in adjusting pretty well after reaching that age, others may not. Successful ageing stems from the view that not all old people are typical of the cliché conveyed through the disease burden as a result of chronic health problems and declining cognitive skills assumed to come with ageing. Some adapt by compensating for losses and declines and retaining the potential for function growth. Kinsella and Phillips (2005) view successful ageing as individual’s attitude towards maximizing desired outcomes and minimizing undesired ones. Rowe and Kahn (1997) view it as the intersecting point of three events that are decreased risk of disease and disability, mental and physical wellness and active engagement with life (involvement in society). From a labour market perspective, productive ageing is seen as the ability to contribute directly and indirectly in older age. Such a contribution could be reflected in one’s decision not to withdraw from the labour market beyond the retirement age. Thus, through this construct of active ageing, it is assumed that old people may still continue to participate in the economic activity in order to secure livelihoods that provide lesser exposure to vulnerability in advanced age. Timonen (2008) also used the concept of ‘active ageing’ to assess the extent to which old people in the context of Europe participate in the formal labour market in order to generate income for their families. If for some aged persons, labour force participation is the result of a choice made voluntary, others are forced to work because the pension is insufficient to sustain the household due to high expenses that come with adult children’s unemployment, illness, and fostered/orphaned children for whom the older person has to care (Madhavan and Schatz, 2007). The activity theory is however silent about the answers to the four questions asked in the section above. This is so because it does not explicitly treat the labour market as a contested economic space among agents. A ramification in the economics of labour market is therefore useful in order to consolidate the activity theory within the framework of this paper. Inasmuch as the elderly represent as a whole individuals that are vulnerable to labour market disadvantage (likely to manifest for example in the form of age discrimination), a labour market segmented (here and after LMS) approach to active ageing seems appropriate in this regard. It makes amenable the activity theory which is rather silent on the labour market outcomes predictable for older workers. The LMS argues that the nature of the job structure and the role of institutional forces, such as discrimination, determine the labour market success (Joll et al., 1983 ; Hull 2009). The mechanisms that allocate workers are controlled as is mobility across the labour market segments. The LSM argues that some workers do not end up in upper segments not because they lack the human capital to work there, rather institutional restrictions on job choice mean that some workers have access only to low status, low wage employment. Drawing from this assumption, one of the reasons why older workers may not get in upper jobs is not because they lack ability or productive potential, but rather because of the refusal of employers to accept them in the work place. Older workers are excluded from upper segments because they are regarded (physically) as discarded labour, whatever their personal and professional records in the past. Borrowing from the arguments developed in the LSM, one can say that since mobility across segments is constrained due to age and discrimination, among other attributes, any old worker’s employment or income opportunities are crucially determined by whether he or she can get into a lower or less skill demanding segment at the bottom level in the late stages of labour market participation. The implications for old workers are quite predictable in magnitude. Because of institutional factors, discrimination being one of them, the proportion of the active elderly in the segments at the bottom and in those with lower demands in skills is predicted to be large. An alternative (or counter argument) to this labour market outcome is that a large proportion of older workers in lower segments may reflect previous job choices ; they had these jobs before reaching the retirement age. This is mostly true for black people as they suffered racial prejudice in the labour market under the apartheid regime. Hence, they could not save enough and that is the reason why they cannot withdraw from the labour market in their old ages. Although this argument holds some truth, it does not as such invalidate the prediction above as it too lends support to the institutional forces that constrain individuals to gaining access to better jobs. Thus the concept of active ageing emulated above may differently fit the trajectories of whites as opposed to those of blacks. Based on the above considerations, it is expected to observe a relative participation in the labour force among old people. This participation may take different forms across the industrial sectors, types of occupation and income category. In short, it is specifically hypothesized the following : (i) the elderly are predominantly represented in the low skill segments ; (ii) they are likely to work outside the salaried form of employment, that is as self-employed or employers, (iii) their involvement in the labour market is mostly in the informal sector and finally (iv) the income derived from employment is modest. Some variations are also predicted across the racial groupings and gender.

Methods and Data

Participation is a dichotomous decision because an individual either participates or does not participate in the labour market. Thus drawing from the conventional measurement, labour force participants are those in full-time or part-time employment (including the self-employed), unemployed job seekers and those temporarily absent from work due to sickness. The socio-demographic analyses in the sections to follow are derived from various household records data and documentary sources in the South Africa context or elsewhere. A variety of data sources as the 2001 Population Census, the 2005 General Household Survey, the 2006 Income and Expenditure Survey, the 2007 community surveys and the 2007 Labour Force Survey are used carry out an assessment of the patterns of involvement in the labour market. All together, these data sets provide means of extracting the primary information on persons aged 60 years and above living in the urban area of the nine provinces of South Africa. The household data offer the possibility of carrying out the analyses for the urban stratum. The unit of analysis is any person aged 60 or above living in ordinary household as invariable defined in almost all the national household surveys and population censuses conducted by Statistics South Africa. The definition of what constitute a household has been consistent throughout the operations of data collection operations conducted by Statistics South Africa along the lines suggested by the United Nations recommendations regarding the methodology of population censuses in Africa. In definitional terms, in all the household surveys conducted under Statistics South Africa, a ‘household’ is defined as ‘a group of persons who occupy the same property and eat from the same pot of food’. A head of the household is a person recognized as such by the other members of the household. A household may comprise of one person only. This definition excludes such accommodating institutions as workers hostel, student hostels, prison, hospital and alike.

In demographic terms, as mentioned above, old people refer to individuals aged 60 years old and over. Because of some considerations related to retirement from the labour market, the age of 60 serves as an appropriate threshold. It is recognized that this age limit sounds somewhat restrictive as age as lower as 50 or 55 may be considered to classify one person in that age group (Kakwani et al., 2006). The threshold of 60 years is however universally regarded in most countries as the start of retirement from labour force, thus marking a demarcating line with the production system of goods and services. This is likely true in the urban based economy of South Africa. Moreover, since age at retirement differs according to the sex of the individual, the lower limit of 60 is taken to include as well women who have reached the legislated age of retirement. Dwelling on this age limit, population ageing is therefore measured in reference to the number of individuals aged 60 years and over in the population. As old age labour supply are not likely to be determined in isolation from other individual attributes related to the individual’s position in the labour market, the analytical framework makes extensively use of such variables as sex of the person (female and male), income range, work status, occupation type and industrial sector. For the sake of space, it is important to note that the tables included in the text do no systematically disaggregate data by race. A great deal of analytical efforts however has been made throughout the analysis to bring out the differences on trajectories and relationships to the labour force along the racial or ethnic lines lines, especially between older whites and blacks, and perhaps also in the Coloured and Indian groups. By so doing, racial differentials in patterns of participation are brought into the discussion about the active contribution of the elderly to the labour market. To broaden the empirical evidence from official statistics, it has been deemed necessary to supplement the discussion with insights compiled from previous works reported in the academic literature and public policy records.

Ageing and working in social context

Labour market outcomes are not exempt of the demographic changes affecting the overall population dynamics. It is an established fact that South Africa is going through a demographic transition and along with this its population is experiencing a steady ageing trend (Kinsella and Ferreira, 1997 ; Noumbissi and Zuberi, 2001 ; Madhavan and Schatz, 2007). Fertility, a major driver of the transition, is currently declining in all racial groups in South Africa though the levels and pace are not identical. This trend must be however nuanced. As such, South Africa has not yet reached the threshold at which the Total Fertility Rate (TFR) is below 2.1, regarded by demographers as the level for which a population cannot ensure replacement. Overall fertility remains above that threshold due to some uneven developmental changes. At present, the fertility rate is estimated at 3 children per black woman (including persons classified as coloured) and 2.3 for whites and Asians. However mortality has followed a reverse pattern. After a substantial decline up to the mid of 1990s, the levels and trends in mortality started to increase from that date. Naturally, there are wide disparities across the racial groups with whites showing the lowest rates and blacks displaying the highest rates. Increased levels of youth and adult mortality due to AIDS is somewhat halting this transition. The estimates established by the United Nations Development Programme (UNDP) indicated that life expectancy would have been close to 70 in the year 2010 in the absence of AIDS (UNAIDS, 2002 cited in Legido-Quigley, 2003). HIV/AIDS has significantly contributed to the increase in levels of mortality over the past years especially in the young and adult life years. The long term consequence of the HIV/AIDS epidemic on population ageing remains highly uncertain (Kinsella and Ferreira, 1997) but indications at hand suggest that in more recent time ageing has in a certain way to do with some of the HIV/AIDS related structural changes affecting the age pyramid of the South African population in its lower and upper middle parts. The importance of aged persons has increased due to the fact the generations 70 and over and 80 and over have escaped the HIV/AIDS epidemic (Legido-Quigley, 2003). People are not living longer and at the same time the old population is getting more relatively represented. One must overlook the fact that, globally, South Africa is an epicenter of HIV/AIDS infection killing every year thousands of adults in their prime wage earning years. The AIDS pandemic has considerably eroded the principal sources of financial and material support for older people. Over the past years, the impact of the epidemic has certainly become more widely distributed across age ; there are fewer children because of pediatric AIDS mortality and reductions in numbers of potential mothers, and there are fewer adults overall because of reduced infant survival and adult AIDS mortality (Kinsella and Ferreira, 1997).

Overall, the percentage of old persons relative to the total population shows an upward trend. The 1996 population census enumerated a proportion of 6.4 per cent (all ethnic groups included) for the population aged 60 years or more. The elderly distribute across men and women as 5.7 per cent and 8.3 per cent, respectively, in that age group (Noumbissi, 2001). At that date of enumeration, the old population in the age group 65 years and over had a lower proportion around 4 per cent. Five years later, the population census conducted in 2001 put the number of individuals aged 60 years and over at 3280505 for a total population of 44819778 inhabitants (Statistics South Africa, 2003). This, translated in percentage, gives a share of 7.3 per cent to the elderly in the total population. Again, old women come in greater proportion as compared to that of males, 8.7 against 5.8 respectively. The overall percentage of 7.3 is quite similar to the demographic pattern observed in most developing countries but lower to that of Asian or Latin American countries. This however hides differentials across the racial divide, as the percentage of old people is much higher for Whites than the three other major racial groups (Black Africans, Indians and Coloureds). The four major population groups are at different stages of demographic transition. Within Southern Africa, South Africa has the highest proportions of older population followed by Namibia. Looking at the percentage distribution of the age group along the gender divide, the same 2001 census indicates that males and females represent 4.3 percent and 5.6 percent respectively out of the total population in each sub-group. This is so because women tend to outlive men especially from the age of 70. Population ageing in South Africa has been accompanied by declining sex ratios, whereby a larger proportion of females are reaching old age than males. This has caused a higher proportion of elderly households to be headed by women, who have not traditionally had to shoulder such a responsibility (Makiwana and Kwazeri, 2007). Evidently these percentages vary across the racial groups.

Put aside the uncertainty over the long term associated with such measures, the projections established by the United Nations population division suggest that in absolute numbers, old people will likely double in importance. Looking at the estimates, the projected percentage share in older age from 1950 to 2050 has increased by 6.3 points from a baseline estimate of 3.6 in 1950 to 9.9 in 2050 (table 1). By the year 2035, the population aged 60 and over is expected to reach 7.4 million, which will likely represent 6.4 percent of the total population. Looking at these trends, altogether, it is without doubt that ageing, measured with the proportion of persons older than 60 or by the dependency ratio, will increase at a faster rate in the years to come due to the changes of all sorts affecting the population of South Africa. Internally, the magnitude of ageing varies across the provinces, with some accentuation in more economically endowed and urbanized provinces where the metropolitan areas are located. South Africa has its majority of the elderly living in urban areas (Kinsella and Ferreira, 1997). The demographic characteristics of the elderly are also changing due to an unevenly ageing process. As indicated in table1, of the three groups the most rapidly growing age group is that aged 60 and over. Thus one can caution against any temptation to aggregate all the old into a homogenous group. Policies and programmes designed on that basis may no reach the intended social outcomes because of the differing needs and problems.

Table 1 : Percentage in older age (1950-2050)

South Africa is not unique in experiencing the temporal variations observed above. Ubiquity put aside, they are however determined by the specific social context in combination with the peculiar sequencing of events in the course of the demographic transition vary from one country to another (United Nations 2006). Variations are also observed across the socio-economic groups within the country. The geographical diversity of demographic influence may in turn influence the participation of the elderly in the labour force. Within the old, the oldest group, 80 and over, possesses most of the characteristics that make them radically different from the rest of the old. The oldest old are economically inactive, have very few social contacts, bodily instable and diseased. By contrast, the old in the age group 60 and over or 65 and over may still be economically productive and physically healthy. Either group has needs and problems that require specific policies and programmes. Although this age-based distinction is not strictly taken into consideration in South Africa, the social security system in place has been quite responsive in providing support to old people. Some of the instruments of the social policy date back in the early 1900s. This paper is not the place to review or assess their effectiveness. Sagner (2000) provides some historical perspectives that inform on ageing and social policy in South Africa under Apartheid. The author stressed that the support provided through the formal system of social security is not in many circumstances as comprehensive as one would expect it to be. It may be the reason why so many old persons resort to transfers operated through other mechanisms, direct and indirect, such as family or community or evening the labour market (as discussed in this paper). Intra-familial transfers have in the past, as is still the case today in most developing countries, played a major role in protecting old peoples against extreme poverty. They are built around family cohesion and inter-generational solidarity. The decision to participate in the labour force may be the result of a self-evaluation of the gap between what is required for meeting one’s life basic needs and the basket of resources actually available to protect oneself against such a level of poverty. In most cases, this self-evaluation is determined by the position one has in the household.

Differentials in labour market outcomes

Rate of participation

As indicated earlier pre-1994 population data are hard to find due to the statistical deficiency deliberately imposed under the apartheid regime. The post-apartheid era is only 16 years old. In demographic terms, this is less than a required generation for substantial change to be recorded. One must therefore resort to alternative sources to get a glimpse into the trends before that date. Looking into the trends elaborated by the United Nations Population Division (2006), it appears that labour force participation of the older population has declined over the last decades in South Africa. The decline in the level of participation was the most marked in the formal labour market (Kinsella and Ferreira, 1997), and older men were the most affected between the mid 1980s and early 1990s. While this may suggest that older people are less likely to participate in the labour force than they were in the past, this could also indicate that the labour market is providing fewer opportunities or becoming somewhat exclusionary to this group of individuals. As argued in an earlier section, older workers face an increasingly competitive labour market characterized by high unemployment and age discrimination, with limited opportunities for those with poor education and training of which blacks predominate (Lam et al 2006). Looking at the direct estimates derived from the 1996 and 2001 population censuses, a moderate proportion of the elderly reported being employed in 1996, with a higher proportion observed for men for the formal employment. Using the first census, Noumbissi (2001) found that 10 per cent of the elderly in South Africa continue to work in their very old ages. No indication was given by this author about the type of economic activities or occupations these old people were involved in. This percentage is not significantly different from that of 11 per cent indicated by the 2001 population census. This second population census also indicated that the elderly were involved either in formal, informal or farming work in the respective proportions of 73 per cent, 13 per cent and 14 per cent among active old persons. Others had employment but were temporarily absent. From either source of data, it emerges a level of participation in the labour force which is not relatively negligible, the least one can say. Hence they inform on the continued involvement of the elderly in the labour force at later ages. Albeit the moderate trend, variations are observed along the gender lines. The differentials in the participation are possibly related to an increase in female share within the older work force. Overall, the proportion of economically active older men however remains far above than that of older women despite the fact that the declining trend is relatively less marked in the second sub-population.

The level of participation of older persons in South Africa is a reflection of how much less prevalent the old-age support systems in the form of pension and retirement programmes may have become. One year before the end of Apartheid in 1994, a study named Project of Statistics on Living Standards and Development found that elderly households (defined as households with at least one elderly member) were poor by most standards (Kinsella and Ferreira, 1997). The study also established that only 30 percent of all surveyed elderly households were living above the poverty line. The participation of older persons, albeit moderate, should not be seen as exceptional to South Africa. In the broader context of Africa, it is a ubiquitous fact. The ambient climate of poverty creates the condition for old persons to remain active in the labour market as they strive for decent living conditions in advanced age. It has been documented that Africa has by far the highest proportion of economically active people among those 60 and older, among the world’s major areas.

By definition the labour force comprises as well the job seekers. Looking at the patterns of unemployment among old people, one can observe that while some old persons successfully gain access to jobs, others remain unemployed in the labour force (actively searching for job). It is an established fact that age tends to act against old people in search for job the labour market. In the context of South Africa, race too plays its part in the search for job. Data however do not show a high rate of unemployment among old people as compared to that actually observed among persons in the age group 25-34. The importance of unemployed old persons declines sharply with age beyond 68. It is likely that some of these people decide to withdraw definitely from the labour force as they get systematically incapacitated. For some, it is not a decision that comes out of choice. Lam et al. (2006) asserted that in contrast to persons of the same age group in European countries, old people in South Africa have a low propensity of withdrawing definitely from the labour market. They estimated that passed age 60, the probability of leaving for good the labour market, conditional on being in the labour force in the previous year, is 10 per cent at age 64. For men, the probability is 30 per cent at age 66. This observation corroborates the view that probabilities found in South Africa are considerably lower in comparison to the patterns in OECD countries (Lam et al., 2006). Compared to whites, old black people in South Africa have a lower propensity of withdrawing from the labour market before the age of 70. An explanation to this difference may be in the weaker incentives to retire due to the features of the retirement systems. Taking considerations of the racial divide, even if apartheid has ended, older black workers may be discriminated against on the basis of race, age and sex. It is however observed that participation by older white people has as well declined through time. While a cause of this decline may lie in improved pension schemes that encourage early retirement, it may at the same time signify a real decline in employment opportunities for older white people. In the current context of affirmative action under the Broad Based Economic Empowerment and the rigidity in wages, their failing productivity with age prices them out of the labour market.

Type of occupations

Occupation provides a qualitative appreciation of the elderly participation in the labour force. In an earlier section, it was hypothesised that old workers are likely to be predominant in low-skill segments of the labour market. The profile of occupations provides some support of that prediction. In relative terms, however, the structure of occupation among the elderly is not markedly different from what one would expect to observe in the actual active population of any African developing country. Using the data from the Labour Force Surveys and the 2001 census which are quite detailed about the different occupations, it is found that old heads of household participate in the labour force in different ways : business ownership, casual work, paid domestic work, unpaid family work, or farm work. Most of them get payment from casual work, suggesting that some of these household heads do not have a stable employment. The involvement in any type of work for payment was mentioned by an overwhelming percentage among the elderly, 73 per cent according to the 2005 LFS. Two years later, the 2007 LFS revealed that people who did work for payment represented 52 per cent. Old people work. Although in a much more minor position, domestic work emerges as a major occupation with nearly 13 per cent. This is most likely to attract old female old persons. Among the domestic workers, the presence of males doing domestic work for wage is quite noticeable, 92 percent against 88 percent females in the age group 60-69 according to the 2007 LFS. In both sexes, the proportion of domestic workers is higher in age group 60-64 than in the rest of age groups. Beyond 70 years, there is a considerable drop in female and male domestic workers though women tend to stay longer in the job than men. Some women are still working at the age of 78 years as domestic workers. For most disadvantaged black and coloured women, they have no other choice but to do domestic work at that age in order to have some financial resources. The most difficult task is to get at the work place every morning as they have to walk long distance from the main bus station. Male domestic workers are mainly found in those outdoor-house activities such as gardening, housekeeping, watering and waste collection. In contrast, female domestic workers are in what one would call indoor-house activities such as kitchen work, cloth washing, child caring and house cleaning. The distribution of domestic workers along the racial lines displays a concentration of this occupation among the most disadvantaged groups namely the blacks and the coloureds. The presence of Indian among the domestic workers is close to nil. Remarkably or perhaps as one could have predicated, there are no whites among the domestic workers within the age 60 and over. The business owners are more represented than domestic workers among old persons. The distribution is even more flattened than those working for wages. A sharp decrease starts to show up only from age 80 and onwards. Men are more represented in the age group 60-64 and 65-69 beyond which women take over. When broken down by race, only two racial groups, black and white, show a substantial ownership of business. The type of business owned may not be of the same nature. The ownership of business in the service and the petty trade sectors are quite ubiquitous among blacks. Their micro businesses generally serve the demands of residents in townships. Being of more sophisticated nature, the business owned by white is more inner-city located. Interestingly, some of these old persons, mostly blacks and coloureds, were involved in farm work for the household, ensuring in this way a comple¬mentary source of income or food for the household. Urban agriculture is very common around the black townships of South Africa. It provides household means of supplementing insufficient income generated from other sources, especially in those locations in the informal settlements with the townships or at the outskirts of the city.

In spite of the predominance of elementary occupations, the occupational deployment of the elderly suggests the presence of a skilled element in it (table 3.a). The labour market segments into which old heads of household find employment show their presence in all types of occupations, with a substantial dominance of employment related to elementary occupation. In both sexes, employment in the category of elementary occupation represents 25 percent of the working elderly. It is only for the occupation ‘domestic workers’ that women slightly take over men in terms of number of workers represented in the sector. For all the other types of occupations, male workers are more represented than female workers. Although the occupational distribution displayed in table 3.b does not suggest marked differences between men and women, more disaggregated statistics by race (not displayed in the table) indicate that old black people tend to be more involved in casual or manual work than whites. It is observed that the proportion of black workers who are completely unskilled and therefore lower paid is significantly greater than for whites. Consequently, this may explain the smaller proportion of blacks in the non-manual occupations. In sum, the profile of occupation suggests a highly segmented elderly labour force with most of in the lower segments. The observed dominance of casual and manual work is an indication of the difficulty they possibly encounter in promoting to the highly paid segments. Among blacks, there is a slight predominance of old workers for occupation in clerks, service and market sales workers and craft workers. The largest proportion, 44 per cent, is found in the elementary occupation. With 59 per cent of their occupations located in the upper segments of the labour market, old white workers are obviously more advantaged than old male workers in terms of higher paying jobs.

Table 3.a : Occupation group of the elderly by sex, LFS 2005

Table 3.b : Occupation group of the elderly by sex, main occupation and race, LFS2007 (for those who did any work for payment)

Income generated from work

As the above observations suggest, the elderly rely on varied forms of work for their economic security. Work therefore provides a major source of income through which consumption is financed. A question of interest would then be around the magnitude of that own labour income. The data on monthly income generated from the participation in economic activity (table 4) show that 42 per cent of the surveyed earned an income below R4500 (at the time of the survey, this was roughly $US500). By the South Africa’s standards, an income of R4500 is insufficient for an effective protection of the household against vulnerability to poverty or social deprivation. For some, this income supplements the old age pension or other forms of retirement savings, making the situation financially better. As illustrated in the table below, although the distribution of disposable income displays a great deal of variation among old people, there is a concentration of earners in the lower ends of the income distribution. For example, the ranges of income for those who did not mention their total salary show earnings that at varying degrees that fall into all the categories. The last too categories-“don’t know” and “refuse”-represent 22 percent of the total respondents for whom income range is not known. (There is not much one can say about this category as clarify¬cation was given).

Table 4 : Monthly income range (for salary not declared)

On average, the small income earned from work suggests a need to design programmatic instruments to capacitate old people in the establishment of income generating activities. At present, such initiatives have been intensively promoted by the government over the past years but they mostly target historically disadvantaged younger people of working age. A wide range of empowerment schemes are pursued but they fall short in integrating old age needs in the labour force participation. One would argue that they were not in the first place intended to target old people. This does not however justify the exclusion of old people from the schemes taking consideration of their demands for work activity to generate additional income. In spite of South Africa having embarked over the recent years on promoting job creation strategies, very few opportunities are offered to old people to participate in such programmes. Such programmes are generally associated with those expanded public works and agricultural schemes which tend to focus on younger workers. Using the 2005 GHS data set where information on skilling can be extracted, it is for example found that country wide less than 1 percent of working household heads of old age took part in such programmes. This corroborates the exclusion above mentioned.

As it emerges from the 2005-2006 Income and Expenditure Survey, salaries and wages provide the main source of income for the elderly. This has to be evaluated against other diversified forms of income sources such as welfare grants, rented property, investments and allowances, which are quite common within the white community. The reliance on salaries and wages although sharply decreasing with age is quite marked in the two age groups 60-64 and 65-69. These two age groups contained respectively 68 per cent and 24 per cent of the old heads of households that mentioned salaries and wages as the main income that brings the most value to the households. Beyond these two age groups, the percentage dropped to about 8. Put aside the percentage gap, there are clear indications that old heads of household have maintained a certain presence in the waged and salaried labour market. Inasmuch as this is a positive labour market outcome, it does not prevent one to think of the vulnerability they face at their work places. In time of poor performance of business, they rank first among the employees to retrench. Hence the importance of looking at their deployment across the industrial sector of the economy where the individual works because some sectors are more exposed to business performance downturn than others. This also provides additional insights into the segmentation of the elderly labour market outcomes.

Industrial deployment

The conventional grouping in nine industrial sectors is used here along the lines of the international nomenclature. From table 5, one can observe a great deal of diversity in the industrial sectors of old workers. Their participation in the labour force is the most reflected in those sectors providing community services (related to social and personal), wholesale and retail trade, agriculture, manufacturing and that of services provided to private households. Interestingly, the profile of participation in any of these industrial sectors is almost the same for both male and female workers. One may look at this as a manifestation of old women getting a strong presence in the labour market. Put aside the sector of community services, male workers are more represented than females. The ratio male to female in most of these sectors is around 1.5 (or 10 males for 5 females). In the construction sector, the ratio is 2 against 1 (or 20 males for 10 females). The industrial profile of the labour force among the elderly provides ample indication of the extent to which individuals aged 60 and above are capable of taking up job opportunities arising in the urban economy in order to sustain the livelihoods of households they are in charge of. The deployment across the industrial sectors should not be taken as if they made the first entry in the sector at advanced age. While some old persons may be just carrying on with the occupation they had long before reaching 60 years old, others may have converted to sectors of easy entry such as the one with activities related to community, social and personal services as they approached that age. It is interesting to observe that agriculture emerges as a sector for which 13 per cent among the 65-69 year olds reported working on a family plot. This is presumably urban agriculture in and around the residential areas, mostly in black townships where old residents make use of each and every small parcel of land to crop in order to supplement their incomes or daily nutritional intakes.

Table 5 : Main industry of activity by sex among the elderly (2005 LFS)

Discussion and some concluding remarks

The foregoing empirical accounts provide an overview of some socio-demographic conditions under which the elderly participate in the labour force in the urban settings of South Africa. The results lend support to the trends documented in some other studies on the elderly in South Africa from a broader perspective. They point however to a contrasted picture for the elderly when one looks at, on the one hand, the position some of them have as persons in the advanced age groups and, on the other hand, the magnitude and quality of their involvement in the labour force. The participation of the elderly in the labour force indicates a sizeable, though relatively moderate, involvement in work for income likely used to supplement old age pension, if any, from the state. This corroborates the view expressed by Lam et al. (2006) that benefiting from pension does not compete with work in old age. Old persons are involved in all the segments of the labour market. A motive of concern however comes from the predominance of jobs taken in the category of elementary occupations. Some of the occupations may classified as “less productive jobs” (Hull, 2009). The explanation to this lower position is not straight forward as time order between age and entry in the job segment cannot be checked out between age and entry in the segment. The positive sign is however the presence, though modest, of old workers in the high skill segments suggesting that their professional competency must be valued in the context of generalized shortage of skills at the national level. The market skills some of them have accumulated may not have deteriorated with time. The data used in this paper do not support the view that older workers in South Africa have fewer consumption demands, fewer family responsibilities, greater access to accumulated savings. The fact that they continue to work beyond the age of 60 is an indication of active ageing in one form or another to generate more income. This does not rule out the possibility that they may suffer lower productivity in market activities due to impaired health. This is likely true in advanced old age for which it is observed a sharp decline in the activity rates. If older workers become inactive or unemployed there is less chance that they will choose to resume employment than younger workers : search for training investments will be unprofitable given their limited remaining active years. The available evidence is broadly consistent with these expectations. Overall, the findings reported in this paper recommend the consolidation of the old age pension by breaking with the means test method to address the needs of the elderly-headed households. Rather a broader range of employment opportunities should be promoted to facilitate the participation of old persons through adequate labour force schemes.

Acknowledgements

The author is indebted to Statistics South Africa (StatSA) for making available the data sets used in this paper. Thanks are offered to Enid Schatz for her valuable comments on an earlier version of this paper. The views expressed in this paper are those of the author alone, and should not be attributed to Statistics South Africa.

References

Cameron, A. Lisa and Cobb-Clark Deborah (2005) Do co residence with financial transfers from children reduce the need for elderly parents to work in developing countries ? Discussion Paper No 508. Centre for Economic Policy Research. Melbourne : The Australian National University.

Ezeh, C. Alex et al. (2000) The situation of older People in Poor Urban Settings : The case of Nairobi, Kenya. In Baney Cohen and Jane Menken (Eds.), Aging in Sub-Saharan Africa : Recommendations for Furthering Research (pp. 189-213). Washington, D.C : The National Academies Press, 189-213.

Ferreira, Monica (1992) Multidimensional Survey of Elderly South Africans, 1990-91 : Key Findings. Cape Town : University of Cape Town.

Hull, Katy (2009) Understanding the relationship between economic growth, employment and poverty reduction. In Organisation for Economic and Cooperation Development (Ed.) Promoting pro-poor growth employment, (69-94) Resource document. OECD http://www.oecd.org/dataoecd/27/7/4... (Accessed 01 May 2010).

Joll, Caroline et al. (1983) Development in labour market analysis. London : George Allen & UNWIN (publishers) Ltd.

Kakwani, Nanak, Son H Hyun and Hinz, Richard (2006) Poverty, Old-age and Social pension in Kenya. Working Paper No 24. Brasilia : International Poverty centre.

Kinsella, Kevin and Ferreira, Monica (1997) Aging Trends.Washington : US Department of Commerce, Bureau of the Census.

Kinsella, Kevin and Phillips, R. David (2005) Global aging : The Challenge of Success. Population Bulletin, 60, 1. Washington : Population reference Bureau.

Lam, David, Leibbrandt, Murray and Ranchhod, Vimal (2006) Labor force withdrawal of the elderly in South Africa. In Barney Cohen and Jane Menken (Eds.) Aging in Sub-Saharan Africa : Recommendations for Furthering Research (pp. 214-249).Washington, D.C : The National Academies Press.

Madhavan, Sangeetha and Schatz, Enid J (2008) Coping with Change : Household structure and composition in rural South Africa, 1992-2003. Scandinavian Journal of Public Health, 35 (3), 85-93.

Makiwane, Monde and Kwizera, Alice, Stella (2007) An investigation of quality of life of the elderly in South Africa, with specific reference to Mpumalanga province. Applied Research in Quality of Life, 10.1007/s11482-007-9022-6.

Noumbissi, Amadou (2003) Poverty among the elderly in South Africa. Resource document http://www.pop.edu/africahh (Accessed 16 August 2007).

Noumbissi, Amadou and Zuberi Tukufu (2001) Household structure and aging in South Africa : A research note. Resource document. http://www.pop.upenn.edu/africahh (Accessed 16 August 2007).

Rowe W. John and Kahn, L. Robert (1998) Successful Aging. New York : The MacArthur Foundation Study.

Sagner, Andreas (2000) Ageing and social policy in South Africa : historical perspectives with particular reference to the Eastern Cape. Journal of Southern African Studies, 26 (4), 1465-893.

Stloukal, Libor (2007) Where do we stand in international action on ageing ? Briefing note prepared for the BSF/FAO Meeting on Ageing, Lilongwe 28-29 November 2007.

Statistics South Africa (2003) Census 2001 in brief. Second edition. Pretoria : Statistics South Africa.

Timonen, Virpi (2008) Ageing Societies : A comparative introduction Berkshire : Open University Press.

United Nations (2006) World Population Ageing, 1950-2050. New York : Department of Economic and Social affairs, Population Division.