| This dissertation explores the effects of social interactions on HIV-related behaviors and attitudes towards domestic violence in rural Malawi. Data from the Malawi Diffusion and Ideational Change Project, a household survey developed for the purpose of exploring the role of social interactions in demographic attitudes and behavior, were used to conduct the empirical analysis. I use a variety of econometric techniques included fixed effects logit models, lagged variables and instrumental variables to deal with the well-known identification and endogeneity issues that may arise when estimating the impacts of social interactions.;The first study in this dissertation finds that social networks influence spousal communication about AIDS, infidelity and condom use, though not always in the desired directions. In some cases, social networks enhance the probability of safe behavior and in other cases social networks hinder the adoption of safe behavior. The second study in this dissertation finds that social interactions, measured by participation in local events, empower women in patrilineal societies to be less accepting of domestic violence. In addition, for men in patrilineal societies, high levels of social interaction at local events appear to be conducive towards negative attitudes towards domestic violence.;The results of this dissertation have several implications for HIV prevention policies and violence reduction policies. First, individual based policies and programs, such as public service announcements, may have important spillover effects as social interactions appear to be influencing HIV-related behavior and attitudes towards domestic violence. Furthermore, community information sessions, local outreach workers and other programs that take advantage of existing social networks may also be vital channels to changing HIV-related behaviors and attitudes towards domestic violence. |
