A summary of this committee meeting is not yet available.
FINANCE SELECT COMMITTEE
10 June 2003
ACTUARIAL SOCIETY OF SOUTH AFRICA & HSRC ON HIV/AIDS FINDINGS; OLD MUTUAL ON DEVELOPMENT AND DEMUTUALISATION
Chairperson: Ms D Mahlangu (ANC)
Actuarial Society of South Africa (ASSA) presentation
Human Sciences Research Council presentation by Mark Colvin
Human Sciences Research Council presentation by Olive Shisana:
(download the following documents file1, file2 and open file1 and extract)
Old Mutual presentation
The meeting was called as a result of references made by the Financial & Fiscal Commission in making their recommendations on how nationally-raised revenue should be equitably distributed to all spheres of Government. These hearings were being held with a view to soliciting the detailed analysis derived by the Actuarial Society of SA in research which the Fiscal and Financial Commission (FFC) had used in modelling their proposal around the equitable distribution of revenue. The HSRC had also been called upon to make a presentation on studies they had completed.
Actuarial Society of South Africa (ASSA) Presentation
The presentation was introduced by Mr Joubert Ferreira, who said that the Actuarial Society of SA represents actuaries in SA. There are 500 practising actuaries, and about 1000 students studying to become actuaries in the country. The role of ASSA is twofold:
- to guard over the professional conduct of the society's members
- to ensure that the expertise of its members is developed as circumstances change
Part of developing knowledge involves engaging in research, which is where modelling HIV/AIDS comes in. Actuaries use their mathematical statistical and financial expertise to build models to help predict and manage the future where events are uncertain, especially in the long-term. Obviously that is based on the data that is acquired from the past. That was typically the approach that was followed with modelling of HIV/AIDS.
Prof. Rob Dorrington, professor of Actuarial Science at UCT, and Director of the Centre for Actuarial Research, continued the presentation looking at :
- the ASSA AIDS Committee and the suite of models
- features of the model used
- the process by which the models were fitted to the provinces, called calibration
- a discussion on the differences between models versus surveys
- a comparison of the new ASSA 2001 model that is being developed with the HSRC results by sex and age.
The Committee was told that the ASSA 2000 lite model modelled the country as one whole population group, while the ASSA 2000 full model scientifically modelled the country's four population groups.
The HIV/AIDS model depicted projections of heterosexual behaviour, allowing them to estimate the number of people becoming infected with HIV through heterosexual activity, and hence the impact that it was going to have on the country. In order to do this, the population was divided by these risk categories: age, behaviour, previous social disadvantage, and geographic location.
Prof. Dorrington pointed out that in their modelling, ASSA makes the assumption that sexual activity between the various population groups, and between provinces is not a significant factor in determining the spread of the epidemic.
The fitting process of the model (calibration) requires setting as many parameters as possible from independent estimates; setting other, not particularly significant, assumptions; and the remaining assumptions are set in order to replicate known data of the prevalence of the epidemic, such as antenatal prevalence or mortality rates.
Prof. Dorrington presented various calibration models depicting prevalence trends of HIV/AIDS in the country (see document). The model projected mortality rates amongst South Africans quite accurately, depicting, amongst South African females, a definite change in mortality rates over time, with continual fluctuations. Other slides depicted HIV prevalence for the nine provinces. Some of the slides had to be adjusted for bias, and fitted well for five of the provinces. Kwazulu-Natal and Mpumalanga Province showed the highest HIV prevalence. The models for the North West Province and Limpopo Province were considerably inaccurate, and it as obvious there were some discrepancies in these calculations.
The Committee was told that ASSA was involved in modelling, not surveying. As a result, they did not produce empirical data, but extracted from empirical data in order to predict future conditions. Conclusions to be drawn from models are limited to the extent that modelling involves a great many simplifications and assumptions. However, they can give useful indications of errors in surveys.
Regarding HSRC surveys, Prof Dorrington stated that their research was invaluable, particularly to the extent that the information could be used in shared research. However, on its limitations, he said that there was a considerable potential for bias, for the following reasons:
- there was a high rate of non-response to surveys,
- by design, these surveys excluded some high-risk populations (prisons, the military and hospitals), and by default other (truck drivers, criminals, etc.)
- there was a perceived unwillingness to share information derived from surveys
Mr T. Ralane (ANC) asked about the methodology used to derive information. He noted that Prof. Dorrington had identified certain institutions as high risk areas, yet these institutions were provided with free condoms, and there was a form of monitoring prisoners and there was no free movement of inmates from one cell to another. He asked how ASSA had come to the conclusion that Kwazulu-Natal and Mpumalanga had the highest prevalence of HIV-infected patients, mentioning that these were mostly rural people. He felt that if this information was solely derived from women attending antenatal clinics, the information would be ill-informed.
Mr Conroy asked to what Prof. Dorrington ascribed the much lower prevalence for HIV in the Western Cape.
Prof. Dorrington replied that ASSA had created a model of the population which generates the number of people infected in such a way that when looking at the prevalence of those who would be attending antenatal clinics, it can be measured against those people actually attending antenatal clinics, and through calibration, then decide they have the right level of infection of the population as a whole.
He acknowledged that prisons and other institutions do provide free condoms, but added that it was fairly well documented that there is sexual activity in prisons, and people do not necessarily stay in cells all by themselves.
The conclusion that people from rural areas have the highest HIV prevalence, is based solely on the antenatal survey, which has a fairly scientific basis, although in Kwazulu-Natal the sample is not particularly scientifically based, but is based in part on a "convenient sample". However, evidence suggests there is not that much difference between other rural areas and the rural areas in Kwazulu-Natal. There is also some evidence based on hospital treatments to suggest that the patterns of prevalence from rural areas are not that different. In South Africa, there may not be too much difference between prevalence in rural and urban areas.
The Western Cape is different from the rest of South Africa in many ways. In the main, the it has a different mix of population groups. The prevalence amongst the white and coloured population groups is significantly lower than that amongst the poorer, previously disadvantaged black population. Many of them come from the Eastern Cape, and so the prevalence in the Western Cape amongst people moving in from other provinces, is not that different from the provinces from which they have come.
Prof. Dorrington stated it was not true that ASSA was modelling only antenatal survey results. Instead, ASSA ascertains which proportion of women attend antenatal clinics, and its their prevalence which is used to check against the antenatal data. The model projects the epidemic in the population as a whole, and draws from that whole, a proportion who attend the antenatal clinic. An assumption is made about the relationship of the prevalence of all women to the women in antenatal clinics. They also make an assumption about the probability of men getting infected, relative to women getting infected. He continued that the model is a heterosexual one, modelling only heterosexual prevalence.
Mr Ralane noted that Prof. Dorrington seemed to suggest that in the Western Cape, there was some transference of the disease because of migrations from the Eastern Cape, and not only as a result of prisons and other institutions in the Western Cape, of which there are many. He informed Prof. Dorrington that the tot system , used as a form of payment by certain farmers, might have significantly influenced people's behavioural patterns to the extent that it might also have contributed to the spread of AIDS. He asked how, in the light of all that, Prof. Dorrington could account for the low HIV prevalence in the Western Cape.
Prof. Dorrington responded that prisons and hospitals would have a higher HIV prevalence than the population from which the sample was drawn, for more obvious reasons. He explained that people sick with AIDS usually end up in hospitals. He agreed that the tot system could account for people indulging in promiscuous sexual activity, but could give no definite answer to the question on why HIV percentages are lower in the Western Cape. There must be some other reason than Mr Ralane's suggestion for the phenomenon.
The Chairperson asked that the calibration of all the races be explained in simple terms.
Prof. Dorrington explained the dots on the model represent antenatal observations and show what is happening in terms of prevalence in the population groups. There is no certainty on prevalence in the Indian population, because so few people from this group attend antenatal clinics. In that case, researchers speak to employers, to obtain information.
Mr Ralane asked if the study had been based on urban or rural women, adding that, since there are not many clinics in rural areas, the former was more likely.
Mr Dorrington responded that the study had been based only on antenatal clinics, in both rural and urban areas. He informed the Committee that there are antenatal clinics in rural areas. Although ASSA's sample had been drawn to include rural clinics, Prof. Dorrington conceded that there could indeed be a bias away from rural clinics in the model.
Mr Conroy noted that obviously, much of the information gathered is based on assumptions. He asked that, if AIDS were declared as a notifiable disease, would the results then be more factually based?
Prof. Dorrington's response to this question was that he did not believe that if AIDS was declared a notifiable disease, better data would be forthcoming.
Ms C. Kganakga, Project Manager: HIV/AIDS, Nelson Mandela Foundation, asked how the model catered for those women who did not attend public institutions, but who rather went to private hospitals.
Prof. Dorrington responded that they were trying to determine if that particular bias was significant. Personally, he sensed that the bias was quite small, and did not help to explain the difference between the model and HSRC results. He further explained that information on women who attend private hospitals, is derived from employers.
On the reliability of the model, Prof. Dorrington stated that this could not really be gauged with certainty. The only measure of reliability is that when a new set of data is received, ASSA can compare it with their model, to see how accurate their predictions had been.
The Chairperson thanked Prof. Dorrington for his useful presentation. She felt that there had to be separate models for women attending public and private facilities. She continued that since the homosexual and gay community is growing, a model on HIV prevalence in the community also had to be prepared. Lastly, she said that the Committee would like to receive Prof. Dorrington's report on the ASSA 2001 model.
Human Sciences Research Council (HSRC) presentation
This presentation was jointly made by Mr Mark Colvin of the Medical Research Council, and Ms Olive Shisana, Principal Investigator.
It was Mr Colvin's contention that information gathered from the use of antenatal data, is not reliable. He informed the Committee that whereas antenatal data is widely used in developing countries to monitor the HIV epidemic, wealthier countries use more sophisticated surveillance tools. While antenatal data is useful for monitoring trends over time, it is not so useful for making population level estimates. This kind of data is likely to overestimate actual HIV prevalence in the younger age group, and provide a distorted picture of the HIV prevalence. Mr Colvin added that with antenatal data, prevalence measured is among pregnant women attending a small proportion of public sector antenatal clinics for 6 to 8 weeks a year, and then extrapolated to all pregnant women. Assumptions are made about women not attending antenatal clinics, and about women using private facilities.
Ms Shisana stated that until the HSRC study on HIV/AIDS had been undertaken, HIV prevalence measures in the country were based on the Department of Health's annual antenatal survey of pregnant women. This particular HSRC study was intended to augment the antenatal survey through a population-based sample of South Africans including men, women, children, all races and ethnic groups, people living in urban areas, rural areas and on farms, as well as hostel residents. It is the first systematically-sampled national community-based survey of the prevalence of HIV in South Africa.
The presentation explained the survey method in great detail, stating that 13 518 individuals were selected for the survey, of which 9 963 agreed to be interviewed. Of these, 8 428 people gave a usable specimen for an HIV test.
The study found that HIV prevalence in the South African population is 11.4%, or 4.5 million people infected. 15.6% of people in the 15-49 age bracket were HIV positive.
With regard to provincial HIV prevalence, the study found that the Free State, Gauteng Province, Mpumalanga Province and Kwazulu-Natal were amongst the highest, and fairly similar. The next group included the Western Cape, North West Province, and Limpopo Province, and these three also compared similarly. The Northern Cape and Eastern Cape showed the lowest HIV prevalence in the country. It was found that Kwazulu-Natal had a lower prevalence than previously thought.
Mr Ralane felt the president had been correct in calling an indaba on this matter, so that all stakeholders could address these important matters. On the spread of AIDS, he said that for men, having multiple sexual partners was really to stimulate their ego. He noted that none of the presenters had mentioned the link of HIV/AIDS to poverty.
Ms Shisana responded that she was in favour of calling people from all persuasions to argue in debate. On linking HIV/AIDS to poverty, on a micro-level, studies seem to show that there is a link. Poor areas demonstrate a high HIV prevalence. She stated that the measure used to denote poverty, requires greater investigation.
Mr Schneeman thanked the HSRC for their presentation. He asked if there had been any specific reason for choosing the particular areas selected for the survey. He asked if a reason could be identified for the high prevalence in informal settlements. He commented that the prevalence amongst 2 to 14-year olds did not surprise him, as in his own constituency, it was picked up through information received from schools, that the level of sexual activity, particularly in primary schools, was quite high.
Mr Colvin stated that the areas for the survey were randomly selected. The high prevalence in informal settlements needs to be studied. The prevalence of multiple partners in informal settlements is very high, because these are more transient living arrangements. The President had previously raised this issue that people not living in houses would raise HIV prevalence.
On HIV prevalence amongst children, a study would be done in the Free State, with a special focus on child abuse.
Mr Z Kolweni (ANC) noting the HSRC report regarding sexual behaviour of elderly people, stated that the elderly were very conservative. The research did not seem to have consulted with religious groups, which are very important to these conservatives. HIV/AIDS campaigning generally concentrates on sensitising the youth to the dangers of HIV/AIDS, and he urged a greater concentration for these awareness campaigns on elderly people.
Mr Colvin said that there was no link between being religious and a lower HIV prevalence.
Mr M Mokoela (ANC) commented that none of the presenters had dealt with the government's stance on the provision of antiretrovirals to AIDS victims.
Ms R. Mabena, a member of the Mpumalanga provincial legislature, asked about confidentiality in testing for HIV in the survey. She asked if the information for the survey had also been obtained from private hospitals.
Ms Shisana explained that in testing for HIV, a swab was placed in the individual's mouth, which was afterwards inserted into a tube, and sent to the HSRC. Information on test results were keyed onto a database, and electronically sent, so that those performing the research never knew who the people were who had tested positive for HIV.
In the study, no comparisons had been made with data from private hospitals.
Mr Ralane asked if in the survey, it was found that people understood what antiretrovirals are, and if any education had been given on ARVs.
The survey indicated that people do not think that ARVs will cure AIDS, but they did think it would extend the lives of AIDS victims.
The Chairperson raised the problem that many people who visit clinics for AIDS tests, do not return to receive their results. She could not understand the necessity of the researchers' question asking individuals if they would get themselves tested if the test was free of charge. She noted that each year money is allocated for free HIV testing and counselling.
Ms Shisana agreed that funds were made available for free HIV counselling and testing. However the researchers did not only test those people who make use of public facilities, but also those who preferred private services.
The Chairperson said that it would be appreciated if there were some co-ordination between organisations on some of the key challenges faced, with the understanding that each organisation's mandate was different.
Old Mutual Presentation
The Old Mutual presentation was made by Mr Herbert Mkhize, Manager: Corporate Affairs. The presentation focused on transformation and development within the company, and with the company's demutualisation process.
He stated that Old Mutual was keen on becoming the champion of transformation and development. Their commitment to transformation and development transcends legislative compliance and social investment. Mr Mkhize stated the company has taken the view that says, "legislation plus â€¦", adding that law provides the minimum requirement for transformation.
Key priority areas identified in effecting transformation are:
1. Black economic empowerment (BEE).
Mr Mkhize pointed out that Old Mutual's BEE assets now equal R1.9 billion, and is still growing
2. Infrastructural investment
Here, assets, which total R2 billion, are still growing.
3. Social security
The Old Mutual Foundation invests R20 million per annum in rural economic development, community education, and so forth.
4. Savings and financial education
In the community, the company provides sponsorships totalling R18 million towards financial education and other special projects.
Old Mutual has invested in a R4 million workplace programme, providing seroprevalence testing, awareness and prevention campaigns, counselling, and a healthcare programme.
6. Rural development
8. Investment in South Africa
On infrastructural investment, Mr Mkhize stated that the company was eager to get involved in bankable issues. If, for instance, Government would allow Old Mutual to take over the building of the Wild Coast Road, Government would then be in a position to focus its finances on improving infrastructure in other, needy areas. Although the provision of housing was not the company's "cup of tea", they are involved in the National Housing Finance Corporation, an organisation that specialises in housing, and therefore knows better how to effect the process of housing. Old Mutual is assisting by making finances available.
On HIV/AIDS, Mr Mkhize said that the company had done a study to determine the prevalence of HIV in Old Mutual, in order for them to know how best to deal with the matter. The company is developing medical schemes specific to assisting HIV positive employees.
On demutualisation, in terms of investment, he said the company is aware that almost on a weekly basis, governmental ministers stop over at various international airports. Old Mutual tries to capitalise on this, by arranging meetings between those ministers and potential investors for those brief moments.
Mr Mkhize reported that when the demutualisation process had begun in August 1998, one million shares had become available. Anyone who was a policyholder as at 24 September 1999, was entitled to receiving free shares. Such policyholders are required to come forward and claim their shares. To date, in spite of media campaigns to inform policyholders, about 300 000 untraceable shareholders have not come forward to claim. One person in particular whom the company had not been able to trace, was in possession of about 17 500 shares. Mr Mkhize asked for the assistance of the Portfolio Committee in informing policyholders of their free shares before the cut-off date of July 2004.
Mr Ralane remarked that when the idea for demutualisation was first suggested, there was a huge outcry by trade unions against it. He asked what the present feeling was and what spin-offs had resulted from demutualisation. He inquired about skills development towards BEE. He also asked what Old Mutual's investment was, in terms of municipality infrastructure development. On Mr Mkhize's statement that Old Mutual was eager to deal in bankable issues, he asked what their social responsibility would be, here.
Mr Mkhize acknowledged the great reluctance of trade unions to accept demutualisation at first, fearing losses of jobs. The company went ahead with the process, and are hopeful for the creation of opportunities that might result from its London listing. The company is now looking at what it could do, as a global market player, to market South Africa.
On skills development, Mr Mkhize reported that Old Mutual has its own business school, called the Old Mutual Business School. The programmes, which are as yet not SAQA-accredited, are run jointly with UCT's Graduate School of Business and the London School of Business.
Mr Mkhize confirmed that the company deals in bankable projects. They are in discussion with National Treasury, especially with regard to PPP, with the idea of Old Mutual developing this project at the conceptual stage. Given the expertise the company has, they might package some of the investment projects. On bankable versus social responsibility, the company has separated the two issues. The mandate of the company to deliver on people's policies, necessitates that the company participate in projects that are bankable. In dealing with issues of social responsibility, that would be done under the umbrella of social responsibility. Municipal infrastructure has been the exclusive terrain of the Industrial Development Corporation (IDC), and they have been unwilling for anybody else to become involved there. However, discussions with the IDC are ongoing.
Mr G. Schneeman (ANC) stated that he was not surprised at all the interest in roads, especially toll roads, probably because of the return on investments with the running of the toll gate system. He asked what Old Mutual actually put back into the community.
The Chairperson asked how Old Mutual's bank operated, and to what extent the bank was assisting poor communities.
Mr Mkhize acknowledged that it was important to consider what sort of spin-offs were created for the community with, for instance the construction of toll roads, other than the temporary jobs that are created. He was not truly in a position to inform the Committee of what was being done to plough back into the community. However, it was an important issue that needed to be addressed.
At this stage, the Chairperson thanked all the presenters for their submissions, and for their preparedness to answer all questions, and adjourned the meeting.