Texting and sexual health: experimental evidence from an information intervention in Kenya

While text-messaging is an efficacious method of disseminating health information in developing contexts, we know less about how users adapt their behavior based on that information. Does it matter how the information is conveyed? This paper presents findings from a randomized field experiment that evaluates the impact of a Short Message Service (SMS) sexual health counseling service on individuals' knowledge and behavior in an urban informal settlement of Nairobi, Kenya. Subjects were randomly assigned to one of three treatment conditions which tested different mechanisms through which technology-enabled information provision could work to alter sexual behavior: (1) information gap reduction, (2) personalization and (3) social cues. The evidence suggests that personalizing the information and providing signals about how other people in the community are behaving can dramatically minimize sexual health risk, compared to simply reducing the information gap. Additionally, individuals receiving generic, non-personalized health information were more likely to engage in risky behavior compared to their counterparts.