The Challenge

We found that 90% of young people in Botswana were unaware that 40-year-old men had the highest HIV prevalence at around 45% - many believed them to be the safe option. For girls in Botswana, HIV prevalence triples between ages 10 to 24 while male prevalence remains constant. This divergence is partly caused by girls dating older partners they believe to be safer. This is indicative of a widespread problem: over 60% of girls in East and Southern Africa lack comprehensive HIV knowledge. The dominant approach to address these problems has been abstinence education. While optimal in theory, many young people do not abstain and, in the absence of alternative safe sex education, when they do become sexually active they engage in risky, unprotected sex.

The Solution

Choices busts the myth that older partners are safer by revealing that older partners are riskier. This is done via a 90-minute class delivered to junior secondary school students (ages 13-16). The program encourages and empowers young people to make healthy relationship choices, through a series of fun games and interactive activities. If they must date, they are encouraged to date their age-mates, who have lower rates of HIV and with whom they can negotiate for safer sex, rather than older partners. The resulting behavior change is simple, but catalytic. Crucially, the intervention is delivered by young people, which we found to be highly effective: the messenger matters!

Scaling Choices

Since 2014, we have continued to test and optimize the Choices program, scaling delivery in Botswana to half the country’s regions and piloting in other countries with similar contexts. Currently Choices is delivered in schools by near-peer facilitators, deployed by Youth Impact directly or by the National Youth Service Program. An additional phone-based delivery model targeting high-risk schools is being piloted. As well as scaling nationally and internationally, we scale impact through iterative program testing to optimize the program and make it more effective. We do this by conducting rapid, rigorous, randomized A/B tests on a termly basis to link program design with learning.