NOTICE: The Editor-in-Chief has retracted the article after a post-publication review revealed that its original results and conclusions were incorrect and unsupported by the data. Although the author’s own re-analysis confirmed these analytical errors, they disagree with the decision to retract the paper.
A recent study published in the Journal of Population Economics sheds light on the unintended consequences of Sweden’s ban on the purchase of sex. The research indicates that the regulation, aimed at curbing prostitution, may have led to an unexpected increase in rape offenses.
The regulation of prostitution remains a highly debated issue globally. In the United States, prostitution is illegal except for in Nevada. European countries vary in their approach from decriminalization, like in Denmark, to outright bans on purchasing sex, such as the Nordic model adopted by Sweden, Norway, and Iceland.
Proponents of the Nordic model argue that it protects human rights and reduces human trafficking by penalizing the buyers of sex rather than the sex workers themselves. However, quantitative evidence on the effectiveness of these laws in reducing related crimes has been sparse and primarily qualitative.
“So far the economic literature studied the effect of decriminalizing prostitution on rape. Yet, little was known about criminalization of prostitution. To this extent, this study analyzes the effect of criminalization of the purchase of prostitution,” explained study author Riccardo Ciacci, an associate professor of economics at Comillas Pontifical University.
For his study, Ciacci utilized data from the Swedish National Council for Crime Prevention, spanning from 1997 to 2014. This period allowed for the analysis of crime rates before and after the introduction of Sweden’s sex purchase ban in 1999.
To tackle potential biases in the analysis (known as endogeneity issues), Ciacci employed several advanced statistical techniques. These included regression discontinuity, which takes advantage of the timing of the law’s implementation; instrumental variable techniques, using the availability of flights as a proxy for access to sex tourism; event study analysis; and fuzzy difference-in-differences estimation methods. These approaches helped isolate the impact of the sex purchase ban from other factors that might influence crime rates.
The primary outcome of the research suggests that instead of decreasing sex-related offenses, the ban coincided with a substantial increase in the number of reported rape offenses. Specifically, between 1997 and 2014, the study estimates a rise in rape incidents by approximately 44% to 62% following the implementation of the ban.
Ciacci did not find evidence that the increase in rape was due to a decrease in the supply of prostitution. Instead, the findings hint at a possible demand-driven effect, where the inability to legally purchase sex may have led some individuals to commit rape as an alternative.
These outcomes challenge the anticipated benefits of criminalizing the purchase of sex as a means to protect individuals from exploitation and reduce sex-related crimes. The study’s implication is that while the intention behind the law is to decrease demand for prostitution and thereby reduce harm, the actual effect might have been to displace behaviors into more harmful and non-consensual acts of sexual violence.
“There is causal evidence that criminalizing the purchase of prostitution leads to an increase in cases of rape,” Ciacci told PsyPost. “First, it might be debated that these results suggest that the purchase of sex should not be criminalized. This current of thought might be motivated on the basis that if purchasing of sex is not criminalized, there will be no increase in rapes.
“Second, it might be also debated that, to the extent that prostitution is paid rape, these results tell us that society might alter human behavior and thus, this policy needs to be accompanied by further measures targeting a potential boost in rape to prevent it. In other words, one might suspect that had this policy been accompanied by policies targeting rape as well, the results might have been different.”
Ciacci acknowledges several limitations within the study. The data used were aggregated at the regional level, which could mask more nuanced local variations in how the law’s effects were experienced. Additionally, the reliance on reported crime rates may not fully capture unreported incidents, a common challenge in studies of sensitive crimes like rape.
Looking forward, Ciacci expresses a desire to explore the connections between prostitution and rape more deeply, to better understand the dynamics at play and to inform more effective public policies. This would ideally involve more granular data, such as at the municipal level, and potentially exploring different methodological approaches to strengthen the validity of the findings.
“Unfortunately, these topics were long forgotten by the economic literature,” Ciacci said. “Now that we (academic economists) have the the empirical tools (causal inference methods) to explore them, it is paramount to do our best to study these issues as well.”
The study, “Banning the purchase of sex increases cases of rape: evidence from Sweden,” was published March 14, 2024.