National leadership on AIDS: does the presence of civil society organisations result in better government responses?
This paper investigates the (potential) relationship between the quality of national HIV/AIDS responses (specifically: HAART and PMTCT coverage as indicators of treatment and prevention responses respectively) and the presence of Civil Society Organisations (CSOs) active on HIV/AIDS. Owing to limitations in data availability, cross-country regression analyses are restricted to Sub-Saharan Africa (N=42). A number of indicators of CSO presence are emnployed, principally the number of organisations listing HIV/AIDS as a focus area in a 2004 UN directory of African NGOs and the number of CSOs responding to, as well as the number of employees reported by CSOs responding to a 2009 UNAIDS civil society survey. Models are constructed that control for population size, national income, international AIDS assistance, burden of disease and other factors expected to influence the outcomes. Results are ambiguous, with some models indicating a positive relationship between the presence of CSOs and HAART and PMTCT coverage and others no, or even a negative relationship. These results therefore do not support the conclusion that the mere presence of CSOs result in better government responses. However, despite time-ordering, the models may suffer from an insurmountable endogeneity problem in that it is equally plausible that CSOs become active in response to poor government responses, and may therefore be associated with relatively poorer rather than better outcomes.