The Demand for, and Impact of, Learning HIV Status

Publication Details

American Economic Review, December 2008, v. 98, iss. 5, pp. 1829-63. Available From: Available from:

Link to Source
Rebecca Thornton
Institutional affiliations
None specified
Grant-holding institution
None specified
Sub-Saharan Africa (includes East and West Africa)
Health Nutrition and Population
HIV and AIDS, Preventive Health and Health Behavior
HIV and AIDS, Preventive Health and Health Behavior
Equity Focus
None specified
Evaluation design
Randomised Control Trials (RCT)
Journal Article


This study evaluates the influence of monetary incentives on the demand for human immunodeficiency virus (HIV) testing and assesses whether learning HIV status can affect sexual behaviour. The debate over the use of HIV testing and counselling as a prevention strategy is partly fuelled by two assumptions. First, it is believed that significant budgets must be invested into mass campaigns to overcome psychological and social barriers keeping people from learning their HIV status. Second, it is assumed that learning HIV status encourages HIV-positive individuals to protect their partners and HIV-negative individuals to protect themselves. The present study investigates the validity of such assumptions in the context of an HIV testing experiment conducted in rural Malawi. Trained nurses were sent to 120 villages to offer free HIV tests and distribute incentive vouchers redeemable upon picking up test results and attending a 30-minute counselling session.
The author estimates the intervention’s impact using a randomised evaluation (RCT) design in which 3,185 individuals living in the study area were randomly offered an HIV test (2,812 accepted) and a voucher of a value randomly ranging from US$0 to US$3. Individuals could then collect test results at temporary test results (VCT) centres. The author uses a regression model in which attendance at a VCT centre is regressed on voucher reception, the amount of the incentive and distance to the VCT centre, controlling for age, gender, HIV status and geographic location. The author then uses a second regression model in which another outcome variable, condom purchase after receiving test results, is regressed on VCT attendance, HIV status and an interaction term between attendance and status, similar control variables being included. Participants were interviewed once at baseline (1998) and twice after the intervention (2001 and 2004).

Main findings

Overall, 69 per cent of the individuals who were tested visited VCT centres to collect their HIV results. The author finds that incentives had a significant effect on the demand for test results, with a VCT attendance rate about twice larger amongst those receiving positive-valued vouchers than amongst those receiving no incentive. The findings also show that for each additional dollar of incentive, the probability of attending a VCT centre increases by 9.1 percentage points. The author notes that distance to the centre also had a significant impact on demand, with the probability of seeking HIV results 6 percent higher amongst individuals living less than 1.5 km from centre than amongst those living farther away.
The results found over the influence of learning HIV status on condom demand are mixed. On one hand, the probability of purchasing condoms was twice higher amongst HIV-positive individuals who collected their results than amongst those who did not. On the other hand, HIV-negative individuals having knowledge of their HIV status were as likely to purchase condoms as those who chose not to collect their results. In all, the study suggests that although small financial incentives can significantly increase the demand for learning HIV status, such knowledge turns out to have limited impact on the protection behaviour of HIV-negative individuals. As a result, the author recommends that other types of prevention programme be investigated to identify reasonable alternatives to costly door-to-door testing interventions.

Scroll to Top