Does Legalizing Sex Work Increase Human Trafficking?

One of the most influential social science papers of the 21st century argued that when countries legalize prostitution between consenting adults, it causes more people to be coerced into sex work.  

The study, published in 2013 in the journal World Development, has been used to stop legalization initiatives around the world and to justify harsh new laws that turn customers of voluntary sex work into criminals, often in the name of stopping human trafficking.

Unfortunately, the authors of the study used a flawed economic model and abysmal data to reach their conclusion. When crucial information was missing, they guessed and filled it in. Then, when the analysis didn't yield what seemed to be the authors' desired finding, they threw out the data. There is no evidence that legalizing prostitution increases human trafficking.

Despite its obvious flaws, the paper has been widely influential, cited not only in the press but by advocates and lawmakers writing policy. The Canadian government referenced the paper when crafting a 2014 law criminalizing the purchase of sexual services, and it influenced a similar law passed in France. An open letter signed by 800 feminist activists pointed to the study as evidence that legalization had failed to reduce "the harms that surround prostitution."  

The Nevada Independent cited the paper as one of "[n]umerous studies…show[ing] that prostitution and sex trafficking are inextricably linked." It has also been referenced in policy debates all over the United States, as various localities have debated decriminalizing sex work. 

So, how do you demonstrate that allowing consenting adults to exchange money for sex causes more people to be driven into sexual slavery?

They classified countries based on 4,950 accounts of human trafficking from 1996 to 2003, tabulated in a dataset put together by the United Nations Office on Drugs and Crime (UNODC). 

The UNODC report was compiled from official government reports, news and opinion articles, and materials produced by activist groups. These sources aren't consistently trustworthy, and yet the study authors weighted them equally and didn't bother taking into account the number of reported victims in each incident. Almost half the accounts were missing crucial data, and the U.N. only included English-language sources.

Another problem with the data is that it included human trafficking unrelated to sex, like people forced to clean houses and prepare food. The dataset is also limited by only counting people trafficked across international borders, ignoring domestic exploitation. 

The authors conceded that given all of these problems, their data "needs to be interpreted cautiously." But they plowed ahead anyway, asserting that their index was still "meaningful."

After tabulating the human trafficking incidents by destination country, the authors looked at whether or not sex work was legal in each place. A problem is that most countries allow some types of sex work but not others, and the laws often vary in different parts of the same country. Some places changed laws significantly during the study period. Enforcement also varies widely, from non-existent to very strict. The binary classification into legal or illegal that the authors used misses more information than it reveals. 

And yet, the dataset they compiled showed no statistically significant link between legalized sex work and trafficking—until the authors eliminated 34 countries from their analysis.

What was their rationale for deleting data? The authors claimed that some of the countries were so poor that their citizens wouldn't have enough money to pay for sex work. That is unlikely to be true, and in any case, wasn't a good reason to exclude data, since the study already controlled for per-capita income. 

If the authors really believed this, they shouldn't have omitted the fact that they suppressed 23 percent of their data. The abstract should have read, "Legalization increases trafficking in rich countries," and claimed that only 116 countries were included in the analysis, not 150.

After boiling the list down to 116 countries, they reran the analysis, but there was another problem. The authors had made so many adjustments and "imputed"—that's a fancy word for guessing—so much of the missing data that the results were statistically unreliable. But they ran with them anyway.

It turned out that the study's strongest finding was that human trafficking destinations happen to be countries with democratic governments, not where sex work is legal. Why didn't they make that the banner claim of their study? Can you imagine any journalists or policymakers citing such a finding to argue that we need more dictatorships?

So the authors ran with the sixth strongest effect they found, suggesting, falsely, that legalizing prostitution caused more human trafficking. That's the finding that would sell.

But logically, we expect legalizing sex work to reduce human trafficking. 

Criminalization discourages voluntary but not coerced sex work, causing trafficking to increase to fill the vacuum left by departing voluntary workers. 

When sex work is a crime, formerly legal providers face the trauma and stigma of jail, along with the cost of fines and bribes. They also lose the ability to complain to authorities about rape, robbery, and other abuse in connection with their work.

The situation is entirely different for traffickers. They were violating the law all along, so they face no additional costs, and the trauma is borne by their enslaved workers.

Therefore, we expect criminalization to mean more coerced and less voluntary sex work.

There's no good reason for the government to interfere with competent adults choosing to exchange money for sexual services. But even people who think the government exists to force their moral choices on unwilling others should not support the criminalization of sex work.

If the goal is to reduce trafficking, put more resources into enforcing anti-coercion laws. If the goal is to discourage paid sex work, decriminalization, which means removing criminal penalties but allowing fines, sin taxes, and other penalties, accomplishes the goal with far less undeserved suffering and official corruption. It also allows sex workers to access police protection against rape, robbery, and assault in connection with their work.

Sex work criminalization leads to trafficking for the same reason alcohol prohibition in the U.S. created bootleggers. They had the same policy goal as temperance activists, who opposed drinking on moral grounds. That's the origin of the term in economics, "bootleggers and Baptists." Criminalization brings together people who dislike an activity and criminal providers who want to discourage competition.

It's unfortunate that such a poorly executed study with a conclusion that defies economic common sense received so much attention from advocates and policymakers. Its perverse finding has likely only led to an increase in human trafficking while making willing, adult sex workers and their customers considerably worse off.

• Motion Graphics: Adani Samat
• Video Editor: Cody Huff
• Audio Mix: Ian Keyser
• Graphic Designer: Nathalie Walker