•  
  •  
 

Abstract

Sex trafficking, a pervasive problem in many parts of the world, has become increasingly salient to policymakers and the general public. Activists, politicians, and scholars continue to engage in debates about how best to curb it. This Article discusses one especially contentious dimension of these debates: does banning prostitution reduce sex trafficking? Or is legalizing prostitution the optimal approach? Or is there a third, better way? Proceeding both theoretically and empirically, this Article seeks to cast light on the relationship between different types of prostitution laws and the prevalence of sex trafficking and human trafficking. It attempts to make three contributions to the literature. First, it builds on existing theories of the link between the demand for purchased sex and the supply of sex-trafficking victims to create a simple ordinal measure of prostitution laws. This measure, which the Article dubs the Prostitution Law Index (PLI), captures not only whether prostitution overall is legal or illegal, but whether buying sex is legal or illegal and whether selling sex is legal or illegal, which better reflects the actual cross-country variation in prostitution laws. The PLI takes into account scale, substitution, and replacement effects in the market for prostitution, where scale refers to increases in the prevalence of trafficking that are caused by growth of the overall market for prostitution; substitution to decreases in trafficking caused by current consumers who purchase sex with trafficking victims and, based on the risk of criminal sanction, shift to instead purchasing sex with individuals who voluntarily sell sex, thereby crowding out trafficking victims; and replacement to decreases in trafficking caused by new voluntary sellers of sex who, incentivized by changes in prostitution laws, enter the market and crowd out trafficking victims. The PLI ranks prostitution laws across countries on a four-point scale (from 1 to 4), based on their expected effectiveness (from least to most effective) in reducing sex trafficking. Second, the study uses a recent dataset provided by the European Union to map the statistical relationship between PLI scores and prevalence of sex trafficking, based on the Article’s theory of scale, substitution, and replacement effects. The analysis suggests that there generally is an inverse relationship between a country’s PLI score and the prevalence of trafficking in that country. Greater legislative efforts to reduce scale and to increase substitution and replacement appear, on average, to be associated with lower levels of sex trafficking. Third, the Article presents a basic Difference-in-Differences analysis—on the basis of extremely limited data and thus with an unusually large number of caveats— of Norway’s 2009 prostitution law reform. Tentative results indicate that the Norwegian reform, which made it legal to sell but illegal to buy sex, may potentially have helped reduce the prevalence of trafficking there.