Conclusions and Recommendations

In the preceding chapters, the impact and consequences of the HIV/AIDS epidemic on small farmers in Eastern Africa, as well as on the estate sector in Zambia and to an extent in Tanzania, have been considered. It was found that some factors appear to be true for all the systems studied, notably that HIV/AIDS furthers the process of impoverishment. Linked to this is the impact which the epidemic has on labour availability. This impact can have grave consequences for farming systems, which are heavily reliant on human labour power, either seasonal or perennial. AIDS-related sickness and death reduce the number of hands available to do both household and farm work having profound effects on the domestic farm labour economy. This results in productivity declines with cash incomes likely to fall.

Obviously the extent to which the farming system is affected depends very much on the particular circumstances which pertain for each situation. Moreover, since farming systems are not static, and at the same time epidemic impact is not constant, with rapid increases in morbidity and mortality being most pronounced in the earlier stages of the epidemic, the consequences vary not only from one context to another, and from also one stage of the epidemic to another.

In Uganda, three different stages of the epidemic as it affects three different farming systems were observed (and supported by population data collected in the course of the study). These were pre-impact, early impact and full impact. Examples of each stage were seen in Rakai, North Mubende and South Iganga respectively. Broadly speaking, it seems that in Uganda, affected households follow a path of regressive decline in terms of the content of their production. This is most marked in the case of Gwanda, where, using comparative material from 1989 (Barnett and Blaikie, 1992), it was observed that there has been a marked shift to more basic and less varied food and other crop production. This change was more pronounced among the poorer households. The focus of these adjustments is in the domestic-farm labour economy, and has serious implications for the lives of women and children. In Uganda, the areas which are more vulnerable to the epidemic are often those which have farming systems which are particularly susceptible to labour loss.

In Zambia, as far as can be ascertained on very limited disease incidence data, the farming systems which are most vulnerable to labour loss are not those which are most vulnerable to the epidemic. epidemic. At the same time, while it may be generally true that the most labour-vulnerable farming systems in Zambia are not immediately susceptible to the epidemic, this should not be taken as grounds for complacency in a country with high levels of seroprevalence and a tradition of large-scale internal and international labour migration. The case studies from Teta and

Chipese indicate that in these communities, the epidemic is already affecting quite large numbers of households. The effects of this differ between patrilineal and matrilineal peoples, the latter, as has been noted above, being much more immediately vulnerable to the effects of labour loss.

In Tanzania where oxen are quite widely used and where there are complex exchanges of labour for use of oxen, one adverse effect of HIV/AIDS may be to increase the depth of socio-economic differentiation between ox-owning and non-ox-owning households. However, a very positive aspect of this is that, in Tanzania, insofar as oxen increase productivity and thus earnings, rural youths appear less likely to leave farming in search of incomes and excitement elsewhere. This is in contrast to the situation in Uganda and is important in relation both to the spread of the epidemic and to labour availability as epidemic-related shortages may be less likely to be exacerbated by labour migration than for example in Uganda.

An important observation, in the light of the differences in the findings, is that it is important to have seroprevalence information from as many sites as possible given that rural data are vital for a deeper understanding of the problem and that currently their availability is very limited.

The recommendations will not exclusively concentrate on the agricultural sector, they also will include nutritional and health aspects, as the impact of HIV/AIDS has multi-sectoral dimensions and has to be tackled multi-sectoral.

6.2 The impact of HIV/AIDS on programmes and projects

Rising mortality and morbidity as a result of AIDS are likely to have an impact on project performance since health is a precondition for development. One of the most serious threats to any project is absenteeism on the part of both beneficiaries and project staff, which can be caused by mortality and morbidity, attendance to funerals and various other exigencies of daily life. Absenteeism from project activities may set back the progress of projects in countries severely affected by the disease. Discrimination at the workplace of HIV-positive staff may further interfere with the work performance of all staff.

The impact of AIDS can manifest itself in various ways. It may result in labour shortages forcing farm households to shift from cash to subsistence crops when food security is being threatened. Cash crops which require a long investment period may not be suitable for families afflicted by AIDS that are in need of quick returns to cover immediate medical, funeral or orphan-related expenses.

Livestock activities might be jeopardized by family members selling off their animals to finance medical care for AIDS patients. In addition, if the person in charge of the livestock dies, family members are often unable to manage the livestock due to the loss of skills and relevant experience.

The viability of agricultural credit schemes may be at risk as a result of HIV/AIDS for three reasons: a) increased mortality may raise the number of defaults; b) AIDS-affected families may be forced to liquidate their assets in order to repay the credit or else have their assets seized, thereby ending up worse off than before they incurred the credit; and c) AlDS-afflicted families may have to spend part or all of the credit to finance medical care for family members suffering from AIDS, rather than use the funds for investment.

Agricultural research priorities might shift in view of AIDS to focus on the special needs of farm household with fewer working adults. Projects supporting agricultural extension services need to ensure that forms of labour-substitution, technical advice and credit services are made available to AIDS-affected farm families. Such projects should review the impact of HIV/AIDS both in terms of increased mortality among agricultural extension staff, but also in terms of the reduction of the work time as a result of the increase in funeral attendance.

In countries or regions where AIDS is claiming the lives of skilled labour and the labour market is limited in this labour segment, consideration of the impact of AIDS may be critical to the success and sustainability of investment projects. Recruitment, personnel replacement, training strategies and employment benefits (medical, pension funds, etc) may have to be revised accordingly.

For any training activity, HIV/AIDS may have to be taken into account both in terms of replacement/re-training provisions and strategies, but also in terms of revising training curricula.

6.3 General recommendations at the planning and policy level

Given that a principal finding of this study is that the impact of HIV/AIDS is uneven between and within countries, it follows that policy responses must be developed in relation to the situation as it is observed in particular regions, districts and communities and thus that local involvement in policy and project development is essential.

Even in countries like Tanzania with high national rates of seroprevalence and cumulative AIDS cases, it is quite difficult to observe the impact of the epidemic in most rural communities. This is because the epidemic manifests quite marked regional and local variations and the impact may vary quite dramatically over small distances. This has implications for the method to be used in any future work on this problem and points once again to the necessity for community involvement in diagnosis and programming if resources are to be used most effectively.

In communities where the impact is just being felt it may be necessary to think in terms of rapid response so as to ensure that

- labour-economising strategies are immediately developed for the specific needs of this type of community through the extension services, research stations and NGOs. This will require considerable inter-agency coordination;
- extension messages contain explicit HIV/AIDS components which should include both general HIV/AIDS advice and education and also sensitization to the known impacts of HIV/AIDS on rural livelihoods and production.

However, in communities such as South Iganga in Uganda the main response should contain medium-term efforts to develop labour economising strategies which are suitable for the local production system plus the inclusion of HIV/AIDS education in the extension service message.

Although these specific points derive from the situation in Uganda, they may be applied to all three countries. Thus, a zoning approach may be adopted using the vulnerability mapping data (see below) from this report or from other sources in other countries. Such an approach would classify areas of each country by a particular type of programme which is relevant.

6.4 Determining epidemic impact on labour availability

The Uganda material, in particular, demonstrates very clearly the difficulty and importance of distinguishing epidemic impact on labour availability from other background effects such as local labour market operation. This observation should be brought to the attention of anybody dealing with responses at the local and district level. At the same time, the most important programme and policy responses will be those which enable people to cope with the results of labour loss.

A major result of HIV/AIDS impact is increased pressure on the household labour economy. The most important programme and policy responses will be those which enable people to cope with the results of labour loss. Recommendations include working with local people, to develop responses which facilitate labour economising responses in:

- production - for example by exploring novel inter-crops or the introduction of new varieties or technologies;
- marketing - for example by assisting women's mobility both spatial and social; and
- caring, for example by the development of and assistance to support groups among members of affected households.

This study reveals that members of communities which have been longer exposed to HIV/AIDS may be more open to innovations than they previously were. People are forced to react to the crisis which is caused by AIDS.

6.5 The use of vulnerability mapping

Earlier work in Uganda used available data to map the relative vulnerability to labour loss of different farming systems. This method has been shown to be as expected a broad but imprecise indicator in the case of Uganda. In the present study, a similar exercise was completed for Zambia, thus providing the basis for broad vulnerability mapping in that country. There were insufficient data to develop the basis for a vulnerability map of Tanzania, although existing reports may provide some guide to this problem and further analysis may be possible.

In general however, it is recommended that an HIV/AIDS farming system/rural livelihood vulnerability mapping system be developed for a country. This would serve three purposes. Firstly, it would enable broad identification of risk by administrative area, agro-ecological zone, farming systems and livelihood strategy. Secondly, it would provide a framework for the development of area-specific policies and thirdly it would facilitate the monitoring of impact and policy outcomes.

The process of vulnerability mapping as a means of understanding how farming and rural livelihood systems are affected by labour loss may also give rise to the development of an early warning system. This would consist of a hierarchy of information - broad classification as has been achieved in Uganda and Zambia, more detailed information from district-level agricultural and other administrative sources, and finally, limited examination of the nature of the impact in specific communities using RRA/PRA methods. The usefulness of such an approach was in part demonstrated in the case work for Tanzania, where, with the exception of Kagera region, the rural impact is so far slight but seroprevalence rates indicate that it will increase in the medium-term. Development of an early warning system should facilitate focused responses when the epidemic impact becomes manifest.

Once a hierarchy has been established three levels of responses could be used.

Level 1: The creation of a broad map of rural vulnerability to the impact of HIV/AIDS. Experience suggests that the information for this first stage of vulnerability mapping description and ranking of farming and rural livelihood systems in terms of their sensitivity to labour loss -may be obtained from country resources in departments of agriculture and from research stations as well as from other sources. This is then combined with seroprevalence and cumulative AIDS case data obtained from National AIDS Control Programme sentinel surveillance surveys.

Level 2: Limited verification of the broad vulnerability mapping through small-scale studies (e.g. RRA/PRA) in order to indicate whether the broad mapping is providing useful guidance, undertaking a detailed description of the situations in specific areas and identifying the types of local coping response and demands for assistance.

Level 3: Development of assistance programmes for areas which exhibit different levels of vulnerability and are at different stages of the epidemic impact.

Moreover, if this approach is to be successful, it will be desirable to develop adequate HIV/AIDS impact training programmes at all levels. These would include:

- general training in HIV/AIDS impact issues for politicians and departmental heads;
- specific training in HIV/AIDS impact issues for middle level administrators in all departments engaged with rural and agricultural development; and
- community-based participatory analysis of the situation and needs in specific communities.

6.6 The targeting of assistance

The main burden of the impact at the household level falls on the labour economy of the household. This has considerable implications for the targeting of assistance on women, children and in some cases widowers who are raising children. It is difficult to see how to target these groups, thus there are few response strategies which can be targeted specifically towards the HIV/AIDS-affected and or afflicted. The issue is poverty and the response has to be to deal with a likely increase in poverty rather than to assume that a high level of targeting is possible. Policy responses should integrate all aspects of poverty relief, the domestic and farm labour, the educational needs of orphans and other young people, food security and income generation.

Using existing coping strategies

Communities and households have developed their own coping mechanism. The reorganisation of cropping schedules, selection of cultivars, rearrangement of domestic/farm labour interfaces are all examples. Some of the accumulated experience of coping should be documented and members of "experienced" communities should be encouraged to share their experience with communities at the beginning of the coping process or which have not yet been affected. This sharing could be achieved both via the extension services, but also through the facilitation of direct contacts between members of different communities.

Cost recovery

Cost recovery of both school and medical costs have become more widespread in recent years as part of liberalisation programmes. These place additional burdens on all households and in particular on affected and afflicted households. Relief in these two areas could make an important contribution to both current and future welfare of rural households.

6.7 Donor funding and the role of NGOs

In Uganda where the long-term effects of an HIV/AIDS epidemic are observable, NGOs have been very active. The difficulty is in knowing how sustainable are their efforts. Accordingly, this is a question which donors should address very clearly when making funding decisions.

Funding decisions should also take into account two other considerations. Firstly, the social and economic impact of HIV/AIDS is, apart from its specifically medical aspects, to increase poverty. This means that specific HIV/AIDS targeting and project design is difficult in a context of general poverty. Policy responses should integrate all aspects of poverty relief, the domestic-farm labour interface, educational needs of orphans and other young people, food security and income generation.

Secondly, whether in some communities (for example Gwanda) efforts to find and fund sustainable projects may be in vain and the need for medium-term relief should be considered. There is a conventional belief that development projects and programmes are preferable to relief projects and programmes. The main rationale for this view is usually that the former are sustainable and longterm, the latter are emergency measures, short-term and unsustainable. While it is certainly the case that in general long-term sustainability should be a policy goal, there is a danger that the provision of long-term development projects and programmes, may be used as a way of avoiding the cost of long-term relief. However, long-term, sustainable development projects require certain macro-economic conditions and a degree of realism if they are to be truly sustainable and appropriate.

6.8 Taking HIV/AIDS into consideration in project design

In countries with pandemic proportions of HIV/AIDS its impact should be taken into account in all steps of a project cycle such as project identification and design, project appraisal, project implementation and monitoring & evaluation.

In the project identification phase, it is important to find out, to what extent HIV/AIDS is prevalent in a project area. Statistical data provide a first indication, but as experience shows, they often differ significantly from the real situation. Therefore the potential impact of AIDS should be included in Participatory Rural Appraisal and Rapid Rural Appraisal exercises. The identification of target groups should be taken serious to ensure that AIDS affected families are not excluded and that their special needs are addressed.

The implications of AIDS need to be considered in cost-benefit analyses on which labour shortages, reduced agricultural yields, absenteeism, etc have their effect. For the project implementation phase the reduced labour availability of both project staff and beneficiaries to participate in project activities should be reflected in the workplan.

The impact of the epidemic on projects needs to be closely monitored in order to make timely adjustments. Due to the special nature of HIV/AIDS the impact is expected to become more serious over the next years. Project strategies and interventions should be revised accordingly.

In view of the above observations, the need for inter-agency and inter-sectoral cooperation in strategy and project formulation is of great importance. This is of some significance for the role which all United Nations agencies will play in the years ahead (and not only in Africa) and may well prove to be a test of the effectiveness of some of those agencies. It also raises the important issue of the relation between United Nations agencies and NGOs, both local and international.

6.9 Specific recommendations at the micro level

6.9.1 Crop production activities
6.9.2 Livestock and small-stock raising
6.9.3 Income-generating activities
6.9.4 Health
6.9.5 Nutrition

In broad terms, for individual farming systems, the following basic types of activities are required:

- improving returns to labour;
- extending the planting period (minimum tillage, early maturing varieties, crops and methods for later planting);
- crop diversification and reducing external input requirements;
- improved livestock management techniques;
- small credit schemes.

6.9.1 Crop production activities

Specific activities or improvements aimed at improving yields, producing more relish crops and helping farmers to adapt to labour shortages might include for crop production, the following:

- advice on possible labour economising methods for cultivating staples in order to delay the effects of labour shortage, for example, how to deal with the weevil problem in bananas;
- advice on how yields of secondary staples, for example cassava and yams, now displacing bananas, may be increased without needing extra labour;
- advice on better storage of secondary staples resulting in an effective increase in production
- access to improved varieties of relish crops, specifically groundnuts, beans, cowpeas and bambara nuts;
- exploring other mixtures and intercrop possibilities;
- use of labour-saving implements e.g. reduced tillage implements for oxen such as rippers and ridgers;
- using minimum tillage and ground cover crops to reduce weeding for hand hoe cultivators;
- consideration of the weed control problem - is there a role for herbicides in some circumstances? Are there alternative mulching methods which do not require large amounts of labour?
- improvements in the marketing of relish crops
- access to small loans for other crops and staples (especially targeting women, see also below);
- promotion of research into indigenous relish crops, processing methods and seed storage;
- improving crop management for intercropping;
- tractor hire; or promotion of draught power; and
- promotion of alternative soil fertility strategies (as opposed to reliance on fertilisers) through the use of organic fertiliser, intercrops and crop rotation.

6.9.2 Livestock and small-stock raising

Specific activities or improvements aimed at livestock production are the following:

- promotion of breeds requiring limited attention;
- advice on feed and health care techniques for cattle, sheep, goats, pigs and poultry;
- in societies where cultural tradition permits, pig keeping may be encouraged as a labour economising replacement for cattle, having advantages both as a source of cash and protein which may be particularly beneficial for female- headed households;
- poultry keeping may be encouraged especially for orphans to raise some cash;
- promotion of non-traditional livestock activities, e.g. bee keeping (apiary);
- finally inheritance customs which may prevent widows from inheriting livestock should be discouraged.

6.9.3 Income-generating activities

Where appropriate, small especially home-based low labour income generating opportunities and petty trading activities may be considered. However, these may well mean that provision needs to be made to ensure access to credit by small farmers, either as individuals or groups. Such activities include sewing, tailoring, bee keeping, pig keeping and the production of handicrafts.

However, it is worth noting that local initiatives and ideas are only feasible in circumstances where there are markets for products. Accordingly, encouragement of any kind of income-generating activity should only be considered on the basis of the right conditions being present. These may include access to the inputs needed for the activity, storage and transport facilities if necessary and reasonable likelihood of the marketability of the product.

6.9.4 Health

The large numbers of AlDS-afflicted and affected families in areas of high seroprevalence means that the care of AIDS patients is an important, time-consuming and expensive part of the daily lives of such households. Many of those interviewed in the course of the field studies whose families had been affected by AIDS said that they preferred to have patients die at home.

In Zambia, for example, home-based care programmes will be essential in most rural areas, especially since several of the local hospitals examined, already had constant occupancy rates of over 100 percent. Thus it will not be possible to care for the bulk of HIV/AIDS patients in hospitals, and neither is it particularly desirable, as they are then separated from their kin.

Women in particular reported that they prefer those who are sick to be at home, as it disrupts their production activities less and removes the cost of transport to the hospital. However, caring for patients at home also involves material costs, notably for food and bedding. With regard to food, it was observed that patients often did not like and would not eat vegetables and thus became very hungry. They wanted meat, fish, and eggs. With respect to bedding, blankets are expensive and if they become soiled and torn are costly to replace.

Measures should therefore be taken regarding the home care of HIV/AIDS patients so as to:

- provide for the training of family members in the basic skills of handling and taking care of patients to avoid the risk of infection; and
- provide resources so that households can afford both food and medication, treatment and bedding for the sick.
In all farming systems, regardless of the stage of epidemic impact, AIDS prevention measures, most notably through health education' urgently need to be introduced (or strengthened as the case may be). Such measures include:
- promotion of social activities since many, particularly marginalized rural areas, offer virtually nothing in terms of activities apart from beer drinking and sex;
- door-to-door awareness-raising campaigns are preferable to rallies as a means of disseminating information;
- health education/AIDS teaching should be integrated into work with agricultural extension groups;
- schools should be supported to provide health and AIDS education. This is particularly relevant in areas where the HIV/AIDS epidemic is regarded as a kind of witchcraft or is generally considered to be a taboo subject. In a number of villages in Zambia, for example, people were more reluctant to discuss at any length symptoms of the living, however, knowing that to say someone has AIDS is to condemn that person. There was no occasion when someone said categorically that a living person had AIDS, but there were occasions when it was stated that someone had died of AIDS. One difficulty in dealing with AIDS in Zambia, in urban as well as rural areas, is that often it will be associated with witchcraft. This lack of understanding of what AIDS is simply hinders campaigns to arrest and mitigate its impact;
- support should be given to awareness-raising concerning the use of condoms. The wider distribution of condoms is also required e.g. from stores, depots, grocery stores etc.;
- pregnancies amongst HIV couples should be discouraged; and
- cultural tradition like sexual cleansing in Zambia, should be discouraged or at least men should be encouraged to use condoms.

6.9.5 Nutrition

Findings from the research in most of the farming systems studied cited the declining nutritional status among households affected and or afflicted by HIV/AIDS. Thus efforts should be made to improve the diets of households burdened by having to care for patients or having a high dependency ratio, because of children (and mothers) who have been relocated through divorce and death.

Greater encouragement should be given to:

- the production and use of groundnuts, soyabeans (soya flour) and mashed cowpeas;
- creating awareness of the need for a balanced diet (and therefore for diversified production);
- mothers continuing breastfeeding even when pregnant again;
- improving food preparation (many foods are frequently over cooked);
- involving men in nutrition programmes; and
- promoting child spacing programmes (i.e. to allow sufficient time for a baby to be adequately weaned and to grow before the next child is born).

6.10 Recommendations for specific target groups

Orphans

As noted in previous chapters, a direct effect of the HIV/AIDS epidemic has been to increase the number of orphaned children.
Accordingly, support to orphans and institutions or organizations concerned with the welfare of such children should be assisted or where necessary established. This includes assistance in the payment of school fees, skills training, etc..

Pastoralists

In Tanzania efforts to complete a study of pastoralists were unsuccessful. Little is known about the effect of the epidemic on such livelihood systems (see Chapter 3), these groups are often marginalised, and such a study should be given priority in future work on HIV/AIDS impact.

Women

In many instances women form the backbone of the farming systems. At the same time, their responsibility for the collection as well as the production of food, and their household duties render them particularly vulnerable to the effects of HIV/AIDS morbidity and mortality. Measures which might assist women and especially those who are or become (e.g. by reason of sickness or divorce) heads of households include the following:

- regarding both land and inheritance laws, such legislation should be reviewed so as to clearly stipulate land ownership rights for women. And although laws may not discriminate against women with regard to land rights, traditional or customary laws often do. In this regard, state laws should prevail over customary or religious laws;
- assistance and advice is generally required in the transport, storage (e.g. through the provision of milling facilities for grains in order to produce flour) and in the marketing of products;
- access by women to draught power, farm implements and other inputs;
- provision of adequate extension services especially to women in remote areas;
- encouragement of group formation among widows who currently and in the future will carry much of the burden of coping;
- encouragement of the use of bicycles by women to ease the burden and time required in undertaking a number of farm and household tasks especially for the collection of firewood, water and in the transport and marketing of produce.

6.11 Recommendations relevant to the Estate Sector

6.11.1 HIV/AIDS education and prevention
6.11.2 Improvement of the socio-economic environment
6.11.3 Recruitment policies
6.11.4 Training and staff development
6.11.5 Employment benefits

So far, the impact of HIV/AIDS on the estate examined in this study, the Nakambala Sugar Estate (NSE) has been noticeable and worrying, but not devastating. In general, there are two main responses for HIV/AIDS: HIV prevention and mitigation. In an environment with limited financial resources, these have to be cost-effective.

In the NSE example, medical staff had strongly expressed the need for HIV testing facilities in order to make better diagnosis and hence provide more appropriate treatment. The testing should obviously not be compulsory but be available for individuals wishing to know their status. In addition, the development of HIV infection could be monitored. However, it is absolutely necessary that results be kept confidential.

HIV testing should also be accompanied by counselling and a home-based care system. In the study of NSE, respondents frequently said that they could not endure an HIV positive result. Therefore, psychological support is essential. The opinion of women in the NSE townships, for example, was split between AIDS patients being cared for at home or in the hospitals. Some women expressed the fear that they do not know how to care for patient and therefore think that the care in hospitals is better. Others believe that the conditions in the hospitals are not favourable and prefer to look after patients themselves? especially if they are taught how to treat the patients. The establishment of a home based care system would be one way to mitigate the effect of HIV/AIDS on companies.

6.11.1 HIV/AIDS education and prevention

Companies should become more active in HIV prevention and education to limit the spread of HIV among the workforce. The present HIV education activities are not sufficient to reach all employees. The expense of such a programme is minimal compared with the cost of replacing someone who dies of the illness. In an awareness campaign, other STDs should also be included. High risk groups, such as cane cutters in the NSE example, should receive special attention. As alcohol abuse is often associated with unreasonable risky sexual behaviour, an HIV prevention programme could be combined with an anti-alcohol campaign.

6.11.2 Improvement of the socio-economic environment

Companies should contribute towards creating a socio-economic environment, which makes the spread of HIV more difficult. One way is to empower women by improving their economic status. In companies such as NSE, there are often plenty of positions, which could be occupied by women, ranging from the unskilled to the professional level.

6.11.3 Recruitment policies

Pre-employment testing is not recommended and would be discriminatory. In countries, such as Zambia, where there is a narrow skill base for professionals and high rates of infection, exclusion of HIV positive persons would be a self-defeating policy. An HIV positive person can be productive and beneficial for the company for a considerable number of years. Even if an employee is HIV negative at the time of starting work, there is no guarantee that the employee will not be infected afterwards. The situation is different with applicants who are already sick and stretch the company's budget on medical costs and other benefits right from the beginning of their assignment. Therefore a mandatory chest examination as part of routine medical evaluation for permanent employment is recommended as is the practice in other parts of the world.

6.11.4 Training and staff development

It is debatable whether employees should be screened prior to training, as training is an investment in persons, which will only be paid back over time. If the employee falls sick soon after completing the training, the investment will have been lost. This investment is also lost if the employees leave the company for other reasons. But such cases tend to be in the minority. Where there is a proven positive result, owing to the nature of the HIV/AIDS, it is not known whether the employee is able to work for the next 6 months or the next 10 years. Therefore a company could still realize a return on its investment because afflicted individuals may still have a fairly long work life. The longer the training takes, the less return could be expected. A cost-benefit analysis could compare the costs of the training with the likely return form the individual; such an analysis would also have to take into account the probability of death and disability occurring from other diseases or circumstances (Pangs, l 992).

6.11.5 Employment benefits

Most of the costs which arise due to HIV/AIDS may be attributed to employment benefits. These costs are inevitable and the company has to live with them. But as it is expected that these costs will rise in the near future consideration should be given to whether any financial limit might be set. If this is the case? these limits should be negotiated with trade unions.