The Challenges of Estimating Prevalence in Human Trafficking Research

East Tower 

In the 25th anniversary year of the Palermo Protocol, there is an increasing demand for prevalence estimations of different exploitative practices that fall under the categories of slavery, forced labour, and human trafficking. This paper sets out the many challenges of estimating prevalence and how different empirical methods overcome the fundamental problem of unobservability inherent in hard to find, or elusive populations. It sets out a series of arguments for how prevalence estimation contributes to anti-trafficking, the risks and challenges associated with prevalence estimation, and the variety of methodological approaches that can be and have been applied for prevalence estimation. It draws on the extant quantitative research on human rights to ground its discussion of unobservability, political methodology on the identification of countable units for prevalence estimation, a review of extant studies and methodological options, and examples of live prevalence estimation projects in Malaysia, India, Niger, and the United States, including their use of complementary novel data streams and computational methods to provide greater context and understanding in which prevalence is situated. Throughout its discussion, the paper focuses on the relationship between samples and statistical inference, and how various methods can reduce uncertainty and improve estimation. The paper concludes with a stylised summary of the remaining lacunae and challenges, examples of best practice, and how prevalence estimation makes an imperfect and partial contribution to the broader efforts at combatting human trafficking.