The Invasion of Privacy: Third-Party Effects in Developing-Country Survey Research

Jenny Trinitapoli, Pennsylvania State University
Alexander Weinreb, University of Texas at Austin

The two-actor model—a single interviewer, a single respondent, questions flowing one way and information the other—provides the basic blueprint for survey interviewing across the globe. The privacy it promises is thought to be the best way to get respondents to admit to some of the non-normative behaviors and attitudes that animate social research. So much for blueprints: in practice, a “third party”—typically a family member or close friend—is often co-present during interviews. Not including spouses and children, third-parties have been noted in more than 20% of DHS interviews in Brazil, Jordan, Morocco, and Congo-Brazzaville and more than 35% of DHS interviews in Mozambique and Guatemala. Yet, to the best of our knowledge, no systematic analysis of these trends in developing countries, or of their effects on data quality, exist in the scholarly literature. Filling this gap empirically is our primary goal in this paper.