The information dynamics behind the sexual networks of prostitution

 

PETTER HOLME

UMEA UNIVERSITY

 

Prostitution is an unusual socioeconomic phenomenon. The high prices of escorts (average prices of about $ 250 / h has been reported) are still an open question among economists. Since it is so expensive, customers should logically be keen on gathering information before an encounter. However, as prostitution is a stigmatized phenomenon, banished from the main mass-medial channels of communication, this information is often shared laterally over social networks, rather than the more hierarchical structure of traditional marketing and consumer advice. Our starting point is a large dataset of sex-buyers discussing their experiences with escorts from which we can derive a temporal network of claimed contacts. We characterize the dynamic and topological properties of this system, focusing on feedback mechanisms from ratings on the discussion forum to the commercial success of the escorts, and vice versa, and also detect long-term temporal correlations. Can such mechanisms and correlations give alternate hypotheses about the high price levels? Furthermore, we discuss spatial and demographic scaling patterns, including a somewhat surprising sublinear scaling of overall activity with city size. It has been argued that escorts and their clients use protection too frequently for this prostitution to be a factor in the spreading of sexually transmitted infections (STIs). Still, can the observed contact patterns tell us anything about STI epidemics?