In 1999 Sweden introduced an innovative asymmetric legislation against prostitution that criminalizes sex buyers but not sex sellers. The Swedish example was followed by Norway and Iceland in 2009, which gave it the catchy nickname “the Nordic Model”. This model has been increasingly seen as a palatable third way between bilateral criminalization, and legalization with regulation. The first is often criticized on the ground that sex workers are typically victims rather than perpetrators, and should therefore not be criminalized; the second because it may have the effect of increasing demand and erode social norms against paid sex. Although some recent policy reports express concerns about the overall effects of the policy (e.g. Amnesty International 2016; Vanwesenbeeck 2017), the Nordic Model proved very popular with politicians, perhaps because it seems to reduce the number prostitutes on the streets: it has been adopted by Canada in 2014, Finland and Northern Ireland in 2015, France in 2016, Ireland in 2017, Israel in 2018, and many other countries are currently discussing its possible introduction. A lively debate is taking place on the overall effects of this legislation, mostly between enthusiastic politicians and skeptical social workers and NGOs.

The debate shares similarities with the standard arguments put forward for or against alcohol prohibition or drug liberalization. The criminalization of an activity is most likely shrinking the corresponding market, because it increases the cost of participation, but also functions as a signal of what a society deems acceptable or not, and coordinates behavior to potentially change social norms. At the same time, however, it pushes the remaining market in the darkness, where criminal activity and violence are likely to increase, and health education and controls to prevent unsafe sex and related diseases become hard. The policy reports informing this debate, however, are typically based on scattered and tentative suggestive evidence, as identifying the causal effects of these policies on the prostitution market is difficult. Moreover, recent economic studies, suggest that intervening on a market (be it for drugs or prostitution) may have complex spillover effects on society that go well beyond the targeted market.

To better inform this debate, and connected policy choices, this paper provides more rigorous evidence on the effects of the 1999 Swedish prostitution law, in particular on violence and related health outcomes in Sweden.

The empirical analysis shows an increase in violence as a consequence of the criminalization of buyers, but mostly ascribable to domestic violence rather than violence towards sex workers. We propose a model that rationalizes our conclusions on the source of the increase in violence and to simulate the effects of other policies.