Governance and AIDS Study

Name of Organization	Institute for Democracy in South Africa (IDASA)
Geographical Scope	Type of Initiatives: Country Regions: Africa Country: South Africa
Purpose	This study illustrates the impact of HIV/AIDS on electoral processes in South Africa and provides the first available evidence of the influence of the pandemic on the democratic process. It is part of an ongoing Africa-wide study by the Governance and AIDS Programme of the Institute for Democracy in South Africa (IDASA). The main message from this research is that the legitimacy and effectiveness of the South African democracy and its electoral processes risk being undermined by the HIV/AIDS epidemic. It does not equate elections with democracy nor does it reduce HIV/AIDS to a vote. Instead it provides empirical evidence of the effect of the biggest challenge facing Africa today and how it may shape the dynamics of our politics. It demonstrates that HIV/AIDS is not just a health crisis, but a pandemic that has implications for political and social processes. The analysis and results presented in the book show that HIV/AIDS may undermine the democratic project in South Africa and Africa by destabilising electoral systems; reducing political party support bases and the ability to compete; decreasing the participation in public policy processes of citizens infected and affected by the pandemic; and potentially undermining the capacity of Electoral Management Bodies (EMB)s to conduct elections effectively.
Area of Governance	Electoral Systems Governance and MDG
Website of initiative	http://www.idasa.org.za/programme_details.asp?RID=23
Publications	http://www.idasa.org.za/Outputs.asp?PID=23&OTID=2 Research Report: The Impact of HIV/AIDS on Electoral Processes in SA, [2004 November 26]. 'HIV/AIDS and Democratic Governance in South Africa: Illustrating the Impact on Electoral Processes'.
Funding sources	Rockefeller Brothers Fund
Source of Data	Using a combination of own and existing data
Type of Data Collection	Administrative Data Focus groups In depth interviews Panel of experts Secondary sources
specifications of type of data collection	Four key methodological approaches were used. Literature review: we consulted national and international titles from academic, activist and official discourses on governance and the epidemic. We analysed a number of key documents including the legal and constitutional frameworks governing the electoral process and central epidemiological reports, and we gathered specific information from the South African parliament, and various central election manifestoes from the parliamentary parties as well as from policy declarations by the IEC. Semi-structured interviews: In order to probe the policies and practices of some of the political parties we conducted interviews with their HIV/AIDS spokespersons. Statistical analysis: Our analysis of quantitative data (that is, information in the form of numbers) always starts with mere descriptions of how frequencies vary across different provinces, different groups in the population or whatever the case may be, but we then take one step further by exploring to what extent there are 'pattern' in the data by doing correlation analyses. The main purpose of these is to establish whether there is a statistical connection between factors A and B, for example whether it is true that a person who is very busy with providing home-based care is less likely to be politically active in civil society. In one instance we also explore the basis for a causal argument through a slightly more advanced regression analysis. Focus Groups: Through six focus group discussions with were privileged to get into direct contact with some 60 persons who were either People Living with HIV/AIDS (PLWHAs) and/or care givers who shared their personal and immediate experiences with us of political and social activity in general and of the April elections in particular. These six discussions were all held in three rural and three urban centres in KwaZulu-Natal. In addition to these four methodologies we opened our research process to critical scrutiny through a stakeholder dialogue in early April 2004 that brought together electoral officials, political party representatives, experts from the UNDP and UNAIDS, international and local research agencies, donor agencies and media representatives. Representatives of key government ministries were also invited.
Measurement Methods / Tools Generated or Used	In the statistical analyses, when using the Afrobarometer surveys for analysing the effects of the epidemic on people's opinions and on whether they are engaged in activities in civil society, we have to work with questions that are proxies for HIV and/or AIDS since direct questions on the HIV status of the respondent or of those in his/her care would be highly unreliable. Some of the questions we work with therefore capture also other illnesses and conditions than those caused by AIDS. A problem with our statistical analyses of voter participation on the basis of electoral data from the IEC is that we can only work at the level of the nine provinces. That is, in testing whether, for instance, it is the case that registration was lower in provinces with higher estimates of HIV prevalence we were only able to work with nine cases (N=9). Statistical correlation analysis should ideally be done on a larger number of cases than that, since the results become somewhat less reliable and more difficult to interpret with such few cases. This does not mean that the analyses generate invalid results, but that they need to be interpreted with some caution. The same applies for the correlation analyses we do on the information from the voters' roll. These limitations in the data we have had to work with imply that the statistical analyses we present are rather blunt methodological instruments for testing the link between the epidemic and non-participation in the election. This will be reflected in how we formulate our conclusions. On the basis of correlation analysis we can say whether there are any grounds for making the general argument that the epidemic has something to do with the level of non-participation on the level of the provinces. If such correlations can be established there are good reasons for exploring the issue in further depth also through other methodologies. One such attempt is made in this report in our commissioning of six focus group interviews with people who are infected by HIV or in other ways directly affected by the consequences of AIDS. The methodology used in focus group discussions can, with the help of special computer software, be analysed with a high degree of systematic rigour that can generate results with some degree of representativity if participants were chosen on correct grounds. We use the material mainly to explore further some of the results that were indicated by our statistical analysis. That is, given what we learned from analysing electoral and epidemiological data through statistical means, we wished to put questions directly to the people most concerned to see if they could shed further light on the possible links between the epidemic and questions relating to political participation more generally and in the recent election more specifically. In brief, we use the material from the focus groups with an explorative ambition, with an urge to learn more from the life stories and personal reflections of people who otherwise are seldom listened to.
List of Indicators
Main Outcomes (Products)
Main Users	Civil society Donor agencies International agencies Policy makers Researchers

How to

Areas of Governance

Country Initiatives

Regional Initiatives

Global Initiatives

UNDP Support