The political context of AIDS-related stigma and knowledge in a South African township community

Abstract

The purpose of this study was to examine the presentation of AIDS-related stigma and knowledge within the political context of the South African government’s response to the AIDS epidemic. It was during the 2000 - 2004 period that key government officials publicly challenged the orthodox views of HIV/AIDS, with the South African president, Thabo Mbeki, actively positing the primary role of poverty and other socio-economic stressors in the progression of the AIDS epidemic. This discursive position had real-time effects for AIDS policy-making and ultimately delayed the implementation of a national antiretroviral (ARV) rollout programme. Consequently this position was criticised by commentators in the media and elsewhere for contributing to an already widespread climate of AIDS stigmatisation and misinformation. To shed more light on these claims we conducted a survey in 2005 in Atteridgeville, a South African township, and compared results with those of a similar survey conducted shortly after ARV medications became available in 2004. Results indicated a reduction in AIDS stigma levels across the 1-year period, and that those participants who endorsed contentious political views (such as those expressed by key government officials) were more likely to have a higher level of AIDS-related stigma than those who disagreed. Nevertheless, this study cautions against drawing a causal relationship between the South African government’s position and AIDS-stigmatising attitudes, and suggests that further political and social factors be accounted for in an attempt to gain a fuller understanding of this seemingly complex relationship.

INTRODUCTION

Over the last two decades the subject of AIDS-related stigma has received a substantial amount of attention in the academic literature, largely because the harmful effects of stigmatisation are often cited as major obstacles in the global fight against HIV/AIDS (Aggleton, Wood & Malcolm, 2005; Bond, Chase & Aggleton, 2002; Holzemer & Uys, 2004). In the South African context, addressing AIDS-related stigma has become all the more urgent given the almost unmanageable scale of the epidemic. Currently, the rate of HIV infection in South Africa is amongst the highest in the world: by the close of 2005 some 5.5 million people were estimated to be living with the disease; over 1 million had already died from AIDS-related causes, and another 6 million are forecast to die by the year 2010 (UNAIDS, 2006).

AIDS-related stigma can be defined as the ‘social psychological processes through which people are discredited when they are perceived to be infected with HIV … regardless of whether they actually are infected and of whether they manifest symptoms of AIDS or AIDS-related complex’ (Herek & Glunt, 1988, pp. 886-887). There are a number of social and cultural factors that imbue AIDS-related stigma with its specific character. These factors are often related to the degree or type(s) of knowledge individuals or communities possess about AIDS. In rural and urban African communities for example, social constructions of HIV are commonly guided by cultural, traditional and religious beliefs or myths about disease and illness (Aggleton, 2000; Aggleton & Chase, 2001; ICRW, 2002; Kalichman & Simbayi, 2004). In many cases, HIV-positive persons are typically blamed for, or seen to be deserving of their status; they may be thought of as cursed or victims of witchcraft, or they may be seen as immoral or sinful and deserving of some due punishment delivered by a transcendental, moral being or deity (Kopelman, 2002; van Niekerk, 2001). In southern Africa, high rates of illiteracy and significant levels of misinformation about HIV/AIDS are known to further reinforce these types of beliefs or myths (Stadler, 2003; van der Vliet, 2004).

While these social and cultural factors may contribute to the consolidation and progression of AIDS-related stigma and misinformation, the effect rendered by the contributions of political discourses, policies and roleplayers is given greater attention in this study. Government policies and laws have in the past directly promoted AIDS discrimination and stigmatisation, whether, for example, by denying HIV-positive foreigners border entry or immigrant status, by legal deportation following the disclosure of HIV-positive status, by implementing mandatory screening and testing for HIV, or by enforcing compulsory notification of AIDS records. In some cases, political leaders have openly stigmatised certain population groups (such as homosexuals, sex workers and drug users) for carrying and spreading the disease (Poku, 2001). In South Africa, public health care policies during the apartheid era systematically discriminated against non-whites (particularly black Africans), resulting in a legacy that continues to affect the decisions made by persons at greatest risk of HIV infection (Kalichman & Simbayi, 2003a).

The relationship between political factors, stigma and knowledge is of interest in this study, particularly in the light of the post-apartheid government’s response to the crisis of AIDS. The South African president, Thabo Mbeki, has been a key figure in this story, primarily through attempting to debate the orthodox perspective of HIV/AIDS. Since 2000, Mbeki has raised a number of questions with regards to the HIV contagion model of AIDS, the accuracy of AIDS statistics, the scale of the AIDS epidemic, and the allocation of state health resources to AIDS and other leading causes of death in South Africa (Mbeki, 2000). At the opening session of the 13th International AIDS Conference, Mbeki dismissed the view that the ‘world’s biggest killer (AIDS) and the greatest cause of ill health and suffering across the globe, including South Africa’ could be ‘blamed on a single virus’ (Mbeki, 2001a). According to Mbeki, proponents of this view failed to fully consider the impact of poverty, low levels of literacy, gender and social inequality, poor sanitation, unemployment, and infectious diseases on the progression of AIDS in the African context (African National Congress [ANC], 2001a; Mbeki, 2001b). In due course, Mbeki’s views orientated the South African government towards rejecting the ‘notion’ that an antiretroviral (ARV) programme was the only way of dealing with HIV/AIDS (Netshitenzhe in ‘TAC misread Mbeki’, 2004). As a result, the government’s health care policies remained devoted to the task of addressing general medical conditions such as tuberculosis, malnutrition, malaria and cholera, rather than expending ‘limited resources on the purchase of antiretrovirals’ (ANC, 2001b).

Mbeki’s views and the government’s response to AIDS were met with widespread condemnation and protest. A number of social commentators attempted to explain Mbeki’s AIDS discourse as a regrettable case of political denialism (Cameron, 2003; Heywood, 2005), as an outdated reaction to past colonialist (or Western) biomedical practices and the racial stereotyping of Africans which emerged from it (Mbali, 2002a; Posel, 2005), or as a misguided attempt to advance a ‘pseudo-scientific’ framework of HIV/AIDS (Geffen, 2006). The South African media were mainly unified in criticising Mbeki for exacerbating an already widespread social context of stigmatisation and misinformation towards those living with HIV/AIDS (see for example, ‘Mbeki AIDS policy’, 2001; ‘HIV/AIDS discrimination’, 2003; Connelly & Macleod, 2003; ‘Mbeki’s AIDS stance’, 2004; ‘Mbeki AIDS facts’, 2004). South Africa’s leading AIDS activist organisation, the Treatment Action Campaign (TAC), also admonished the government for speaking of ‘poverty, transformation and delivery as a national issue rather than HIV’ (Heywood, 2003).

The South African government in turn rejected these criticisms, with one official announcing that much of the debate on AIDS was being grossly misrepresented and oversimplified, particularly by the media, who were trying to bring about the ‘wilful encouragement of hysteria’ (‘ANC still cautious’, 2001; ANC, 2001a). Furthermore, the TAC was also dismissed as a ‘single issue’ campaign which ignored the challenges of implementing a ‘comprehensive approach to fighting HIV/AIDS’ (Dugmore, 2003; Tshabalala-Msimang, 2004). Mbeki himself remarked that there was a determined effort ‘to hide the truth about the direct and immediate relationship between poverty and health … and the diseases of poverty that regularly claim the lives of hundreds of thousands of our people’ (Mbeki, 2002). Despite their commitment to this position, both Mbeki and the government were ultimately compelled to accept the mainstream science of HIV and the socio-economic rights of citizens to access ARV treatments. Following two court rulings and mounting public pressure, the Cabinet announced in August 2003 that medications would be made available at public health facilities countrywide (South African Government, 2003) and in February 2004, HIV-infected individuals with CD4 counts below 200 registered for the first time to access government funded medications.

Although withdrawing from the debate by 2004, Mbeki’s discursive challenge had real-time effects and consequences for AIDS policy-making in South Africa (Butler, 2005). By the close of 2006, for example, a climate of political scepticism towards the orthodox science and treatment of HIV/AIDS was still in existence, despite the commencement of a national ARV rollout programme. This study attempts to examine this political context and the potential impact thereof, particularly with respect to a number of claims made within the public realm that the government’s position had heightened a climate of AIDS stigmatisation and misinformation. The aim of this study was therefore twofold. Firstly, it aimed to assess the differences in AIDS-related stigma and knowledge over a 1-year period following the South African government’s decision to make ARV medications available in early 2004. This study also sought to investigate differences over the same time period in the number of people who had personal experiences of knowing someone with AIDS. It was expected that the eventual rollout of ARV medications would result in a lower level of AIDS-related stigma, a higher level of AIDS knowledge, and a higher level of experience of AIDS because of the institutional acceptance of HIV science and medicine filtering down to the community level. Secondly, it aimed to assess the relationship between the endorsement of contentious political views about AIDS and stigmatising attitudes. It was hypothesised that those participants who agreed with the kinds of political views expressed by Mbeki and the government were more likely to demonstrate a higher level of AIDS-related stigma than those participants who disagreed.

METHODS

Participants

This study uses data collected through two surveys administered in July 2004 and August 2005 in a South African township called Atteridgeville. Established in 1936, Atteridgeville is one of the two main townships situated on the outskirts of the capital city, Pretoria. The inhabitants of Atteridgeville are mainly indigenous black Africans with a household income ranging from R1 000 to R3 200 (approximately $150 - $500 US) per month (City of Tshwane, 2005). Townships are unique features of the South African urban landscape and were first established as self-contained black African localities. These localities were positioned on the fringes of white urban areas and functioned as temporary reservoirs for black migrant labourers during the apartheid era. Investments in permanent housing, infrastructure, education and other essential services were purposefully kept to a minimum, in order to defer the potential mass movement of Africans from the rural areas into the townships. Today, Atteridgeville is a lively post-apartheid township comprised of informal dwellings (such as shacks and other forms of temporary housing) and permanent formal housing. Despite the transition to democracy since 1994, black Africans continue to remain the most disadvantaged of the four population groups (compared with whites, coloureds and Asians); more than 30% of this group (blacks) are unemployed and are at the greatest risk for HIV infection, particularly if living in an informal urban community (Shinsana et al., 2005; Statistics South Africa, 2004).

The 2004 community survey sought to examine the level of AIDS-stigmatising attitudes within Atteridgeville and the relevant factors that affected these attitudes (Visser et al., 2005). The 2005 community survey shared the same research design and methodology, and included a scale consisting of six contentious political views or statements about AIDS. The sampling technique used identified participants by firstly the representative sampling method in which participants were recruited according to the gender and age ratios (based on 2001 census data) of the total Atteridgeville population, and secondly the convenience sampling method in which participants were recruited randomly until the specified gender and age ratio targets were met.

Procedures

Five hundred questionnaires were collected in 2004 and again in 2005. Four public sites were chosen in Atteridgeville for data collection. Two of these sites were situated in the east of Atteridgeville: the first opposite one of the township’s public health clinics and the second opposite the police station. The third site was situated further west alongside another public health clinic, and the fourth site within the Phomolong section of the township. The four sites were selected because of their accessibility and potential to facilitate the recruitment of participants. Permission to use the sites was attained from the Atteridgeville ward coordinator.

A research team comprising one supervisor and six to eight interviewees (second- and third-year sociology students from the University of Pretoria) were placed at each of the four sites. Potential participants were approached by the interviewees and asked to participate in the questionnaire survey. Participants were eligible if they were South African, over 18 years of age and resided in Atteridgeville. The duration of residence in the township, HIV status, social status or class were not considered in the sampling process. Participants were asked to give informed verbal consent before participating in the questionnaire survey. They were briefed by the interviewees about the background and purpose of study and were notified that their responses would remain anonymous. Participants could terminate the interview at any time, as well as choose not to answer some of the survey questions. They were also given the option of responding in English, Sesotho or Sepedi, the most prevalent languages spoken in the area. Upon completion, each questionnaire was handed to the supervisor of the site to ensure that the sections were correctly completed. The supervisor then registered the questionnaires under an age and gender column in order to calculate the daily target ratios (i.e. the number of men and women within their respective age groups to be interviewed for that day).

Institutional review board approval was obtained for this study from the Faculty of Health Sciences Research Ethics Committee of the University of Pretoria, South Africa, and the Human Investigation Committee of Yale University School of Medicine, United States of America.

Measures

Demographic questions included age, gender, marital status and education level completed. AIDS knowledge was assessed by a 15-item scale in which participants were asked to ‘agree’, ‘disagree’ or respond ‘don’t know’. Items included ‘A person can get HIV by not using condoms during sexual intercourse’; ‘HIV can be transmitted from mother to baby, through breast feeding’. Each correct response was given a score of 1 and summed to provide a total knowledge score of 0 - 15. Experience of AIDS was assessed with a single item question that asked if participants knew of someone close, or of an acquaintance who has or who had died of HIV/AIDS.

Two parallel, 11-item AIDS stigma scales were adapted from Bauman’s Perceived Stigma of HIV/AIDS scale (Bauman, Camacho, Westbrook, & Forbes-Jones, 1997; Bauman, Silver, & Camacho, 2000; Westbrook & Bauman, 1996) and were worded differently to measure: (a) the participant’s personal views on stigma (personal stigma variable; example of an item: ‘If you have HIV you must have done something wrong to deserve it’), and (b) the participant’s perceptions of stigma within the community (perceived community stigma variable; example of an item: ‘Most people believe that if you have HIV you must have done something wrong to deserve it’). Participants were asked to ‘agree’ or ‘disagree’ with the statements (participants could also choose not to respond to the items). The scale ranged from 0 to 11, with 11 representing a more stigmatising attitude. The internal consistency values for personal and perceived community stigma were good (α=0.75 and 0.85 respectively).

In order to assess the extent to which respondents endorsed contentious political views, the 2005 questionnaire included a scale of six items. Two of these items were adapted from an Afrobarometer survey that sought to understand the relationship between public opinion and HIV/AIDS across eastern and southern Africa (Afrobarometer, 2004). The remainder of the items were based on and developed from statements made by key government officials from 2000 onwards. Participants were asked to ‘agree’ or ‘disagree’ with these statements (participants could also choose not to respond to the items). The scale was summed out of a total of 6, with 6 representing more support for contentious political views on AIDS. The internal consistency for this scale was acceptable (α=0.60).

Data analysis

The data were coded and analysed using SPSS. To assess aim 1, independent t-tests were conducted comparing the 2004 and 2005 cohorts on personal and perceived community stigma, AIDS knowledge and experience of AIDS. To assess aim 2, point biserial correlations and Pearson correlations were used to assess the relationship between the demographic variables, AIDS knowledge, experience of AIDS, personal and perceived community stigma and the endorsement of contentious political views. We conducted correlations at the bivariate level to parallel the multivariate analyses used in this study. A hierarchical regression analysis was therefore used. First, we entered the demographic variables, AIDS knowledge and experience of AIDS into the model. Second, the endorsement of contentious political views variable was entered into the model to assess whether these views related to the personal and perceived community stigma, above and beyond the demographic factors, AIDS knowledge and experience of AIDS variables.

RESULTS

The mean age for the 2004 and 2005 cohorts was 36 years, with females representing 52% of the sample population; 23% of the participants were married and 24% of the participants in each sample possessed a tertiary education. In the 2005 sample, significantly more participants (86%) knew of someone living with HIV/AIDS, as compared with the participants in the 2004 sample (71%), (p<0.001). There was no difference between the levels of AIDS knowledge between the two cohorts (p=0.64). Results did reveal that personal stigma was significantly lower in the 2005 cohort than the 2004 cohort (Table 1); however, the 2005 cohort perceived the community to be more stigmatising than the previous year.

Table 1.

Comparison between 2004 (N=500) and 2005 (N=500) cohorts

Variable	Cohort	Mean	Standard deviation	t-score	Sig. (2-tailed)
Personal stigmaa	2004	2.73	2.38	-2.46	<0.01
	2005	2.37	2.26
Perceived community stigmaa	2004	6.68	3.30	2.74	<0.01
	2005	7.25	3.31
AIDS knowledgeb	2004	11.15	2.34	0.47	=0.64
	2005	11.22	2.09
Experience of AIDSc	2004	0.73	0.45	-5.57	<0.001
	2005	0.87	0.34

^aScore of 0 = minimum stigmatising attitude, 11 = maximum stigmatising attitude.

^bScore of 0 = minimum AIDS knowledge, 15 = maximum AIDS knowledge.

^cScore of 0 = no experience of AIDS, 1 = experience of AIDS.

Results for the endorsement of contentious political views are displayed in Table 2. The mean score was 2.33 (standard deviation=1.63) and 42% of the cohort endorsed at least 3 of the 6 statements. A bivariate analysis revealed that older participants (r=0.09, p<0.05) as well as participants without a tertiary education (r=0.11, p<0.05) were more likely to agree with contentious political views about AIDS (Table 3). Using a multivariate analysis it was found that a model including the variables age, gender, tertiary education, AIDS knowledge and experience of AIDS significantly accounted for the variation in the endorsement of contentious political views variable (p<0.05); however, these variables only accounted for a small percentage of the variation (2%).

Table 2.

Agreement with political statements for 2005 cohort (N=500)

Statement	Number	Per cent
1. President Mbeki has said that poverty rather than HIV is the primary cause of AIDS in South Africa. Do you agree/disagree with Mbeki’s view?	274	55%
2. Mbeki believes that poverty related issues such as unemployment and the lack of housing are more important than the problem of HIV. Do you agree/disagree?	226	45%
3. HIV-positive patients should be taken care of by family, friends or a partner, and not by the government.	211	42%
4. The government has said that the people are better served when money is spent on general health-care issues such as TB, malnutrition, malaria, cholera and opportunistic infections. It is better that the government does not spend too much money on AIDS medication.	180	36%
5. There are many other problems facing this country, people with HIV/AIDS should not receive too much attention from the government.	151	30%
6. President Mbeki said that he had not known personally of someone who died of HIV/AIDS. Mbeki is right to believe that AIDS is not affecting as many people as is made out to be.	108	22%

Table 3.

Bivariate relationships between demographic characteristics, stigma and contentious political views for the 2005 cohort

		(N=500)	Contentious political views a	Personal stigmab	Perceived community stigmab
Age (years)	18-25	125	2.06	1.94	7.71
	26-35	163	2.37	2.15	7.15
	36-50	132	2.41	2.42	7.19
	51 and over	80	2.54	3.41	6.85
Gender	Female	240	2.36	2.07	7.44
	Male	260	2.30	2.65	7.07
Education	No schooling	17	2.40	3.90	6.75
	Schooling	362	2.43	2.65	7.30
	Tertiary	118	2.01	1.31	7.18
Marital status	Not married	383	2.24	2.35	7.34
	Married	115	2.62	2.42	6.94
Experience of AIDS	No	61	2.62	2.76	7.08
	Yes	433	2.28	2.32	7.27

^a0 = minimum score, 6 = maximum score for endorsement of political views.

^bScore of 0 = minimum stigmatising attitude, 11 = maximum stigmatising attitude.

Results for the 2005 survey indicate that younger people (r=0.21, p<0.001) and females (r=0.13, p<0.01) were less likely to stigmatise. Participants with a higher AIDS knowledge score (r=0.40, p<0.001) and participants with a tertiary level education (r=0.26, p<0.01) were also less likely to stigmatise. No correlation was found between personal experience of AIDS and stigmatisation. In both surveys, a satisfactory statistical model could not account for the perceived community stigma score.

Participants who endorsed contentious political views about AIDS demonstrated more stigmatising attitudes (r=0.20, p<0.001). In order to investigate this relationship more closely, the variables age, gender, education, AIDS knowledge, experience of AIDS, and the endorsement of political views (independent variables) were entered hierarchically into a multiple regression analysis; this model accounted for 23% (p<0.001) of the variance for personal stigma. Controlling for the demographic variables, AIDS knowledge, and experience of AIDS, it was found that the endorsement of contentious political views variable contributed significantly to the personal stigma outcome (Beta=0.15, p<0.001).

DISCUSSION

In this study, an attempt was made to examine the relationship between political factors, AIDS-related stigma and knowledge in the context of a South African township community. Specifically, the rationale for this study was developed in the light of the decision by the South African president, Thabo Mbeki, to debate the scientific orthodoxy and treatment of HIV/AIDS. Mbeki’s views had real-time effects and consequences for AIDS policy-making in South Africa, which ultimately delayed the implementation of a national ARV rollout programme until 2004. As a result, both Mbeki and the South African government were criticised by commentators in the media and elsewhere for lacking the political will and leadership to put AIDS at the centre of the public agenda. Moreover, this disposition was seen to contribute to an already heightened climate of AIDS stigmatisation and misinformation in South Africa.

This study sought to examine the impact of the eventual implementation of an ARV rollout programme on the levels of AIDS-related stigma and knowledge in the township of Atteridgeville. In comparing the results for both the 2004 and 2005 cohorts, the level of personal stigma was shown to have significantly decreased, and that there was a striking increase in the number of participants who knew of someone living with AIDS. It would be difficult to attribute these positive changes solely to the institutional availability of ARVs, as these two results may have been influenced by documented increases in AIDS-related knowledge, the impact and visibility of AIDS awareness campaigns, and increases in HIV testing rates across South African communities (Stein, 2003). The results do show that the 2005 cohort perceived the community to be more stigmatising when compared to the 2004 cohort. In an attempt to explain this result, Visser et al. (2005) suggest the tendency of an individual to present the self in a more positive light than the average. Thus, participants may have perceived others in the community to be more stigmatising than themselves and so reflected this attitude in their survey responses.

Agreement with Mbeki’s views and the government’s position on AIDS was prevalent across the 2005 cohort. More than half of the participants agreed with the statement that poverty was the primary cause of AIDS, and just less than half believed that poverty-related issues were more important than the problem of HIV. The high level of agreement with these two statements is of interest, particularly in a geopolitical context where the link between poverty and AIDS has been so ardently debated. As discussed earlier, Mbeki has in the past considered poverty-related issues to play a prominent role in the progression of the AIDS epidemic, while mainstream scientific research has suggested a more complex relationship between socio-economic status and HIV-prevalence (Fenton, 2004; Gie, Schaaf & Barnes 1993; Halperin & Allen, 2001; Shelton, Cassell & Adetunji, 2005). Although poverty is not a direct cause of AIDS (Makgoba, 2000; Sewankambo et al., 2000), studies have shown a connection between poor social conditions, high-risk behaviour and susceptibility to HIV-infection (Kalichman et al., 2006; Karim et al., 1995; Shisana & Simbayi, 2002). For example, it is known that those with limited access to financial assets and resources often resort to risky sexual behaviour in return for jobs, accommodation, money, transport, food and other survival necessities. Additionally, poor educational experiences and illiteracy prevent people from gaining an understanding of the means by which HIV is transmitted, and limited access to health facilities in economically disadvantaged areas can compound the problem (Evian, 1993; Mitton, 2000).

In this regard, it is not surprising that South Africans who are confined to this socio-economic reality tend to view the problem of poverty as interrelated with the problem of AIDS. In their study Kalichman et al. (2006) show that survey respondents who experienced a range of complex social problems did not necessarily perceive the problem of AIDS to be unique, even though these respondents reported a higher level of risk behaviour for HIV infection. Across South Africa, a significant number of township dwellers considered the social problems of crime, violence, unemployment, and a lack of education to be equal to, or even more pressing than, the problem of AIDS (Kalichman & Simbayi, 2003b; Kalichman et al., 2005; Whiteside, et al., 2002). Whiteside et al. (2002), in a discussion of their findings, suggest the usefulness of ‘rational prioritisation’ as an explanatory concept for these perceptions. In such a scenario, the downgrading of AIDS (as an important issue) can be explained as a rational response by the poor who must confront an array of urgent social problems on a daily basis. Subsequently, the poor would expect immediate government attention and resources to be devoted to these social problems rather than specifically to the problem of AIDS. To this extent, the process of rational prioritisation could partially account for the levels of support for the government’s position on AIDS. As the results of this study demonstrate, almost one-third of the 2005 cohort believed that the government should not devote too much attention to AIDS, more than two-fifths believed that the care of HIV-positive patients should not be shouldered by the government, and just less than two-fifths believed that the government should not spend too much money on AIDS.

To what extent could the support for the government’s position on AIDS be the result of factors other than rational prioritisation? The lack of educational resources and opportunities in township communities affected by poverty and other related issues could be a significant factor, given that the results of this study showed a significant negative correlation between knowledge and education on the one hand, and AIDS-related stigma, on the other. It is therefore necessary to ask whether the endorsement of contentious political views was likely to be influenced by a participant’s level of education and knowledge about AIDS. Results indicate that the level of education completed and AIDS knowledge accounted for only a slight variation in the contentious political views variable. Thus while these variables have some effect they do not substantially explain whether one will endorse contentious political statements about AIDS. Alternatively, the role of ‘political affiliation’ can be loosely acknowledged here, although it was not examined in the current study. This factor may be of importance, given Mbeki’s leadership of the African National Congress (ANC), the current ruling party, which gained more than two-thirds of the vote (from primarily black Africans) across the country and more than 90% of the vote in Atteridgeville (Independent Electoral Commission, 2004). It is possible that a substantial number of participants in the 2005 survey agreed with a number of the political views or statements about AIDS because of their political affiliation with Mbeki and the government. Previous research has shown that of the four population groups, black Africans generally tend to assess the government’s response to AIDS more favourably. For example, investigators of two Human Sciences Research Council (HSRC) studies found that black Africans were more optimistic that the South African government was allocating sufficient resources to AIDS, that political leaders were publicly recognising the importance of AIDS, and that there was a political commitment to controlling the AIDS epidemic (Shinsana & Simbayi, 2002; Shinsana et al., 2005).

In this study, we found that there was a significant positive correlation between the endorsement of contentious political AIDS views and AIDS-related stigma. Thus, participants who demonstrated a higher level of agreement with contentious political AIDS views were more likely to demonstrate a higher level of stigma than those participants who disagreed. At first, this result seemingly adds weight to the claims made by commentators in the media and elsewhere that the unorthodox position of Mbeki and the government may have potentially exacerbated an already heightened climate of stigmatising attitudes and perceptions about AIDS.

While the above result indicates a tentative relationship between the government’s position and AIDS-related stigma, a word of caution must be expressed, particularly with respect to the limitations of this study. To begin with, it would be difficult to discern a direction in the relationship between contentious political AIDS views and personal stigma. In other words, this study did not develop a causal model to test whether the endorsement of these views led directly to the presentation of stigmatising attitudes and perceptions about AIDS. More generally, it would also be difficult to conclude that the political views of Mbeki and the government were directly related to an increase or decrease in AIDS stigma levels across South Africa.

The attempt to isolate a causal relationship would need to account for a range of other potential factors not included in this study. For example, this study did not examine the psychological or social processes by which individuals or communities are likely to be influenced by political issues and rhetoric. It is important to note that the actions of government officials and the events occurring on a political stage are often mediated through a range of social institutions, practices and personal belief systems before the direct impact thereof can be quantitatively measured. It would also be useful to consider the experiences of the economically disadvantaged in South Africa, particularly regarding the translation of these experiences into potential support for the government’s AIDS policies.

The findings of this study and the studies cited above indicate some degree of consistency between the political perceptions people have about AIDS and the government’s response to the epidemic. In this respect, the views of Mbeki and the government, the socio-political context from which these views emanate, and the notable support for such views should be taken into account when examining the political context of AIDS-related issues. Certainly, such an investigation would have to penetrate further than the kinds of descriptions that have been offered in the media and elsewhere of the government’s position on AIDS as ‘irrational’ (‘Sadly, Mbeki liability’, 2002), ‘warped’ (Mbali, 2002b), ‘arrogant’ (Trengove-Jones, 2000) and even ‘evil’ (‘Huge protests threatened’, 2001).

Finally, the results of this study were limited to only one township community over a 1-year period. It would not be possible to generalise these findings across the broader South African context. In addition the time frame of 1 year may not be an adequate indication of the changes in levels of AIDS-related stigma and knowledge. Nevertheless, the 1-year period from 2004 to 2005 was an important one in the South African story of AIDS, and it is hoped that the findings of this study will help to shed light on what is clearly a complex relationship between political factors, AIDS knowledge and AIDS-related stigma.

Biographies

Brian Forsyth is Professor of Pediatrics at Yale University and a Deputy Director of Yale’s Center for Interdisciplinary Research on AIDS. Since 2002, Dr Forsyth has collaborated with investigators at the University of Pretoria in conducting the Serithi Project, a multidisciplinary research project focused on addressing the effects of HIV/AIDS on women and children.

Alain Vandormael is an associate fellow of the Department of Sociology at the University of Pretoria. Since 2004 he has been involved with the Serithi project and Yale’s Center for Interdisciplinary Research on AIDS (CIRA). His research focuses on the political response of the government and civil society to the AIDS crisis in South Africa.

Trace Kershaw is an Assistant Professor at Yale University School of Public Health. His research focuses on sexual and reproductive health of young couples.

Janis Grobbelaar is Professor and Head of the Department of Sociology, University of Pretoria. She has done wide-ranging fieldwork on poverty, development, socio-political transformation, local government, human rights and race and ethnicity in the eastern, western and northern Cape, in the north west, the Free State, Limpopo, Mpumalanga, Gauteng and in Natal. She was seconded to the South Africa Truth and Reconciliation Commission for 2 years and acted as Deputy Provincial Electoral Officer for the Pretoria sub-province on secondment during the 1994 general elections held in Pretoria. She has produced numerous research reports, published a monograph and a number of journal and book articles dealing with South African studies.