University of Oxford
2 files

Prostitution Legislation_and_Sex Trafficking_joined dataset

Version 2 2024-01-01, 15:12
Version 1 2024-01-01, 15:06
posted on 2024-01-01, 15:12 authored by Amy ForzaAmy Forza

This study was conducted to explore the effects prostitution legislation has on sex trafficking rates. This issue holds paramount importance in the fields of legal studies and human rights. By leveraging advanced machine learning techniques to analyze data from the Counter-Trafficking Data Collaborative (CTDC), encompassing 180 countries, this study aims to uncover the relationship between various prostitution legislation types and sex trafficking occurrences. The exploration begins with extensive cleaning, merging, and filtering of the CTDC dataset, integrating it with prostitution legislation data from the World Population Review. This process ensures a harmonized dataset that accurately reflects the global landscape of sex trafficking in relation to legislative frameworks. The machine learning model initially concentrated on prostitution legislation as a key variable but evolved to include a broader range of factors like registration year, population, growth rate, gender, and citizenship. This expansion was crucial in developing a more accurate and holistic model.

This study offered a nuanced exploration of the impact of prostitution legislation on sex trafficking, employing sophisticated data analysis and machine learning models to parse through extensive data. The advanced RandomForestClassifier was key in the research, achieving an 87% accuracy rate for predicting instances of sex trafficking and demonstrating the need to incorporate diverse predictive features. Notably, the analysis emphasized the importance of the legislative feature in accurately predicting sex trafficking, despite the inclusion of other variables to improve overall model precision. These findings underscore the significance of a multifaceted approach, considering factors like demographics and socio-economic indicators, to gain a comprehensive understanding of sex trafficking trends.

Complementing the machine learning insights, a logistic regression model scrutinized the specific effects of different legislative approaches on sex trafficking. The analysis revealed that legislative frameworks such as legalization, abolitionism, decriminalization, and neo-abolitionism have a considerable influence on reducing sex trafficking rates, suggesting their potential as effective legal strategies. Alternantively, prohibition legislation is found to corrrelate with significantly higher sex trafficking rates. These results serve as a critical resource for policymakers and advocates engaged in the development of informed, evidence-based approaches to address the global challenge of sex trafficking.


Usage metrics

    University of Oxford



    Ref. manager