The Behavioral Response to Male Circumcision: Experimental Evidence from Kenya

This paper investigates the behavioral response to male circumcision (MC) for HIV/AIDS prevention. Recent evidence from randomized controlled trials (RCTs) indicates that medically performed MC significantly reduces men’s risk of acquiring HIV through sexual intercourse. However, HIV/AIDS incidence will not necessarily fall if the lowered risk of infection induces a higher propensity for risky sexual behavior. This paper emphasizes the role of an individual’s belief about MC in affecting his sexual behavior: an individual will only exhibit a behavioral response to MC because of his concern about the risk of HIV infection if he believes the effect of MC on this risk. The paper first establishes a behavioral model to study an individual’s tradeoff between pleasure from risky sex and risk of HIV infection, analyzing the effect of the safety-improving MC on this tradeoff. The paper then presents empirical evidence from a large-sample RCT of MC conducted in Kisumu, Kenya. In the RCT, MC increased sexual riskiness among the participants who were randomly assigned MC, whether they believed the protective effect of MC against HIV/AIDS. However, circumcised participants who believed the efficacy of MC (the circumcised believers) exhibited significantly less risky sexual behavior compared with those who did not believe this fact, so that the net effect of MC on the risky sexual behavior of circumcised believers was not significantly different from zero. Based on the empirical evidence, we discuss some policy implications for MC campaigns aiming for HIV/AIDS prevention. JEL classification: D81, I18