Using City Attributes to Predict Human Trafficking Business Types

Abstract

Human traffickers employ a variety of business models with which to traffic victims and profit from their victimization. Awareness of human trafficking, particularly sex trafficking, and how to recognize it are key facilitators in reducing the problem and preventing further victimization. Determining predictors of sex trafficking business models will contribute to building awareness, and this analysis seeks to establish if sex trafficking business models and city characteristics are correlated. Logistic regression, multinomial logistic regression, and random forest models were trained on data collected from the court documents of federally prosecuted human trafficking cases in the State of California as well as city attributes from Census data – e.g. population density, median home value, education statistics, etc. – to determine their correlation. Results indicate that certain city characteristics are correlated with sex trafficking business models, which can be used to provide indications of what type of sex trafficking business models will be problematic in a given city based on these key characteristics.