A few years ago a friend of mine participated in a program that was designed to use sports to educate children in Africa about the danger of unprotected sex. His plan was to spend several months travelling with a group of students around the Gambia bringing this program to the schools. When he arrived in the capital city Banjul, however, the local authorities made it clear that while they welcomed the project, the only message they would be permitted to leave with the students was they should say no to sex before marriage. No discussion on condom use would be tolerated.
The Gambia is not the only Sub-Saharan African country to take this approach. Kenya, which has an HIV prevalence rate of 3% for girls between the ages of 15 and 19, 9% for women between the ages of 20 and 24 and 13% for women between 25 and 29, has an HIV education program that teaches moral values, refusal skills, and abstinence until marriage. There is no mention of condoms in the curriculum that instead teaches that the best strategy to avoid HIV is to just say “no”.
Despite this call for abstinence 21% of grade eight girls and 48% of grade eight boys report that they have had sex.
One of the reasons why HIV is so prevalent among girls in Kenya--four times higher in fact than among boys of the same age--is that girls are having sexual relationships with much older men. There is evidence to support that as well, among girls who have become pregnant within a year of grade eight, 49% report the father is more than five years older and 16% report that he is more than 10 years older.
Here is the (economic) issue. The girls know that becoming pregnant is one of the risks of being sexually active. If they become pregnant with a boy their own age they have little hope in terms of financial support. With higher incomes, older men are much more willing to give money even without pregnancy and are much more successful at negotiating unprotected sex. So, the girls have a preference for unprotected sex, essentially in exchange for money, with “sugar daddies."
Of course, while the girls can observe the income distribution of their potential sexual partners, they can’t observe the distribution of HIV infections and may not be making optimal decisions about sexual partners if they make their decision based on income alone.
In a novel experiment, published last month, a group of researchers introduced an education program into a randomly selected subset of schools in Kenya that gave students one simple piece of information: they informed them the prevalence of HIV by age and gender in the closest nearby city. Presenters visited schools and spoke for only 40 minutes about the evidence. In that time they showed an educational video on sugar daddies and spoke of the risks of cross-generational sex. In accordance with the national guidelines they did not raise the issue of condom use, but did answer questions that were raised by students on that topic.
In the year following the program, the pregnancy rate of girls who received the cross-generational information was 28% lower than the control group. The decline in pregnancies was almost exclusively among girls who were pregnant with men who were more than five years older than themselves – that rate fell an incredible 61.7%. Finally, the girls in the treatment group were 36% more likely to report having used a condom in their last act of intercourse relative to the control group.
Pregnancy rates is an imperfect measure of risky sexual behavior. We all know there are equally risky substitutes to vaginal intercourse, but the decline in unprotected vaginal sex with older men suggests that girls changed their behavior in response to the new information on the distribution of HIV infections. We don’t know if this simple-to-implement program reduced the incidence of HIV, but in a forty minute visit to just 71 schools the researchers managed to avert the birth of 30 children born to mothers who are just fifteen years old.
That seems to me a policy worth pursuing.
Dupas, Pascaline (2011) "Do Teenagers Respond to HIV Risk Information? Evidence from a Field Experiment in Kenya" American Economic Journal: Applied Economics 3: 1–3