Exchange Sex and Network Structure Among Lgbtq Homeless Youth
Methods: Individual and social network data were collected from a sample of 225 homeless youth collected using Freeman’s Event Based Sampling. During one month of services, all clients receiving services at MFP were asked to participate in the survey (youth were ages 13-24); only 18 youth (6%) declined to participate. Social network data was collected in a face-to-face interview conducted by a trained interviewer. After youth finished nominating their egocentric networks, attributes of each nomination were then collected, including: first and last name, aliases, age, gender, race/ethnicity, tattoos, and whether the nominee was a client of the agency. A whole network dataset was created by linking participants in the sample. Participation in exchange sex (trading sex for money, food, drugs) was based on self-report. Network variables were created with UCINet and merged with self-interview for analysis with logistic regression models.
Results: Relative to heterosexual males, gay, bisexual, and heterosexual-identifying males who have sex with men were all more likely to report exchange sex (OR=15.5, OR=6.8, OR=7.4, p<0.05). Among females, bisexual and heterosexual-identifying females who have sex with women were more likely to report exchange sex (OR=5.8, OR=5.1, p<0.05). Whole network data allow one to examine indirect ties and direct ties. In this network, being directly connected to a peer who exchanges sex or having a two-step indirect tie to a peer who is engaged in exchange sex increased the odds of reporting exchange sex (OR=12.7, p<0.01). Youth who were nominated more by other youth in the network were more likely to have either a direct tie or a two-step indirect tie to a peer who exchanges sex (OR=3.8, p<0.001).
Discussion: Put colloquially, if their “friend” or their “friend’s friend” is involved in exchange sex, males are more likely to be involved in exchange sex. Males who are more “popular” in this network are more likely to have a “friend” or a “friend of a friend” who is having exchange sex. Whole network data expand peer-influence models and theories beyond dyadic social learning and homophilous selection. Such data allow one to understand how a small number of ties aggregate into larger, whole network structures, and how positions in whole networks, and clustering of direct and indirect relationships, impact the risk-taking and preventive behaviors of a population. Such data are critical in determining appropriate peer-based intervention strategies.