Last month, after hearing nearly 14 hours of testimony, the D.C. council voted to table a controversial bill that would have decriminalized sex work in the nation’s capital. If they were going strictly by the economics, council members might have decided differently: When sex work is decriminalized, violence against women declines.
This is not just a local debate. The issue is becoming more prominent nationwide, with several Democratic presidential candidates considering it as part of a broader rethinking of the criminal justice system.
Among skeptics and opponents, two concerns predominate. The first is that prostitution erodes the moral fabric of society. The second is that prostitution is inherently violent and decriminalization would worsen the exploitation of women.
Economic evidence — and theory — has something to say about this second objection. Some background: The so-called Nordic model seeks to simultaneously reduce the imprisonment and exploitation of sex workers by addressing demand. In effect, that means prosecuting managers and customers, but not the sex workers themselves.
It sounds like a well-balanced solution to a complex social problem: Punish the exploiters, not the exploited. Combined with outreach, it promises to lift desperate woman out of a dangerous way of life.
The problem is that this approach lacks strong support from sex workers, who say that it complicates their screening procedures for customers and drives sex work underground. This jibes with economic theory, which suggests that the costs of any limitation on the market fall on those with the fewest options, whether buyers or sellers.
New research confirms just how powerful this effect can be. Economists studied Craigslist, which from 2002 to 2010 gradually introduced an “erotic services” section that allowed sex workers to advertise directly and anonymously on the internet.
The staggered rollout allowed the economists to measure the impact on each market as the service expanded. As expected, the market for sex workers expanded rapidly. More important, according to the 2019 paper, the expansion of Craigslist into a market “led to a 10% to 17% reduction in female homicides.” To be clear, that figure is not homicides among sex workers — which are difficult to measure in real time — but homicides among all women in the area.
This result so astounded the economists that they performed some tests to validate it. It passed them all. Moreover, effects have been demonstrated in other studies. Decriminalization in even parts of a city is associated with double-digit declines in sexual assault. A 2014 study of an inadvertent decriminalization of indoor sex work in Rhode Island from 2003 to 2009 found it resulted in a 30% drop in rapes. This isn’t mere correlation: Both the Rhode Island and Craigslist studies use several methods designed to identify causation.
Sex work experts do report that the introduction of Craigslist and other such online services made such work far safer. And economists say their data is unambiguous. Still, the question remains: Are such enormous effects plausible?
Economic theory offers a possible answer: Reducing the restrictions on sex work expands the set of available economic opportunities, especially to those at the bottom of the economic scale. This expansion shifts the overall balance of economic power toward poor women — giving them more leverage in other economic relationships that expose them to violence. That, in turn, alters the behavior of the men (and women) they work and live with.
Economists refer to these as general equilibrium effects, and they can cause small changes to have a large economic impact. In this case, reducing the danger associated with sex work could make it easier for women to avoid unsafe situations in other areas of work — thus multiplying the original effect.
This is only a theory. Meanwhile, as the data overwhelmingly suggest, full decriminalization makes sex work less dangerous. Public officials who want to reduce violence against women should embrace it.
Karl W. Smith is a former assistant professor of economics at the University of North Carolina’s school of government and founder of the blog Modeled Behavior.