HIV is a huge health issue around the world and especially in sub-Saharan Africa. Many American NGOs have promoted abstinence programs as a way to prevent the spread of HIV/AIDS. However, most evidence finds that this approach has been ineffective.
An NBER working paper by Dupas (2009) adds more support that abstinence programs do not work. In the paper, the author used a
“…randomized fi eld experiment involving 328 primary schools to compare the effects of providing abstinence-only versus detailed HIV risk information on teenage sexual behavior. Half of the schools, randomly selected, received teacher training on the national HIV/AIDS curriculum, which focuses on abstinence until marriage, but does not discuss risk reduction strategies (such as condom use or selection of safer partners). In 71 schools, randomly selected after stratifying by teacher training status, an information campaign provided teenagers with information on the prevalence of HIV disaggregated by age and gender group (the relative risks information campaign).”
The authors finds that the abstinence program had no effect on pregnancy rates. However, the “risk reduction” educational program decreased the probability a girl had started childbearing within a year by 28%. The decreased pregnancy rates were not, however, due to less frequent sexual activity. Instead, teenage girls switched their sexual partners from older partners to teenage boys in their age cohort.
This leads to the finding that teenage girls are having the same amount of sex, teenage boys are having more sex, but pregnancies are decreasing. Why is this?
The author explains that when teenage girls have sex with teenage boys, “…teenage girls report higher rates of condom use, presumably in order to avoid pregnancy with resource-constrained teenagers.” It is also possible that teenage girls can more easily convince boys of their same age to wear a condom whereas it may be more difficult to convince older men to use a condom.
Thus, we see that these “relative risk” educational programs do not decrease sexual activity on the extensive margin (teenage girls are having the same amount of sex), but do decrease risky behaviors on the intensive margin (more condom use when teenage girls have sex with people of their same age).
- Pascaline Dupas (2009) “Do Teenagers Respond to HIV Risk Information? Evidence from a Field Experiment in Kenya,” NBER WP #14707.