How Common Is Sex Trafficking, Really?

How common is sex trafficking in the western world? There’s a lot of concern about it, particularly in recent Onlyfans discourse. In some circles, sex trafficking is implicitly viewed as an epidemic, something we should make strong moves against.

It’s hard to get data on how common it is. People might not consider what happened to them ‘sex trafficking’, maybe it happens disproportionately to lower classes who are harder to survey, maybe people are extremely private about it, or ashamed, and would never check the ‘yes’ box on a survey.

So I figured I’d try a survey where I ask people about “who they know” – I’d ask a bunch of people (mostly from the US or western Europe) if they knew anybody who was sex trafficked, and also ask them if they knew a bunch of people with other traits, so I could see how social knowledge of people’s networks for different traits compared to each other.

For example, sex trafficking victims probably keep it quiet, so I also included questions about knowing rape victims or pedophiles – both things people also have a lot of incentive to keep quiet.

I also asked about other things – stuff like “do you know someone who’s a doctor” and “do you know anyone with lung cancer”, things that both have really concrete data so we can check to see how real rates compare with people’s knowledge of them.

I also asked about stuff around class to make sure I wasn’t only getting data from people exposed to upper-class communities where nobody has ever been sex trafficked. I asked about trump support, lottery winning (disproportionately played by lower classes), homelessness, and imprisonment.

I also asked about tribe stuff to see if people weren’t contacting insular communities. I asked about knowing active orthodox jews, practicing Muslims, fundamentalist Christians, trans people,

I also asked if people knew sex workers, both online and in person, as sex work communities presumably have high overlap with the types of people who might be trafficked.

I also asked about some classic rare things; winning the lottery and never learning to read before the age of 18 to get a sense of the ‘lower bound’ of things that should almost never happen.

In the survey, I asked people to tell me the ‘closest’ circle of those they knew; do you know someone who e.g. voted for Trump, personally? Are they someone in your wider network? An acquaintance? A close friend or family member? Is it you, yourself? The closer it was to home, the higher the score, and I took the average score of each trait to see how sex trafficking ranked.

I got around 3300 responses. Each question got an average score between 0-4 points for “how common” it was in people’s social networks. An average score of 4 would mean it happened to literally everybody who answered (e.g., breathed air), and an average score of 0 would mean nobody had ever even heard of someone in their networks it happened to (e.g., has been to Mars).
Here’s the total chart, from most to least common:

Before continuing, make a prediction – where do you think community knowledge of sex trafficking ranked among other things? How often would you expect people report knowing someone in their community had been a victim of sex trafficking?

Average rating – the question people answered – actual population rate estimate at the time of survey, if available

Do you know someone:

  • 1.92 – who’s switched their major political party affiliation after the age of 18
  • 1.82 – who’s tested positive for Covid-19? 2.8% at time of survey, Oct 2020
  • 1.73 – who’s been raped? 9.6%
  • 1.68 – who voted for Trump? 22.5%
  • 1.53 – who’s currently a medical doctor? 0.3%
  • 1.35 – who’s polyamorous? 45%
  • 1.24 – who’s transgender? 0.5%
  • 1.22 – who is currently a practicing Muslim? 0.3-1%
  • 1.17 – who is currently a fundamentalist, evangelical Christian? 0.3-24%
  • 1.12 – who’s committed suicide? 0.07%
  • 1.07 – who experienced a significant and permanent psychotic break? 0.1-1%
  • 0.94 – who is openly and explicitly racist; for example, clearly believes in the inherent superiority of one race over others? ???%
  • 0.90 – who’s been diagnosed with lung cancer? 0.3%
  • 0.89 – who’s been incarcerated for over 1 year? 0.26%
  • 0.83 – who’s an online or no-contact sex worker (e.g., stripper, camgirl), or has done this in the last five years? ???%
  • 0.80 – who’s been homeless for 3+ months? 1-2%
  • 0.74 – who was homeschooled for the entirety of their k-12 years? ~1%
  • 0.58 – public figures who are famous enough that news outlets write about their lives? ???%
  • 0.51 – who’s an escort or a prostitute, or who has done this in the last five years? ~0.3%
  • 0.50 – who is currently an orthodox Jew? 0.2%
  • 0.46 – who’s been murdered? 0.03%
  • 0.42 – who’s died from Covid-19? 0.09% at time of survey, Oct 2020
  • 0.33 – who won over 10k in the lottery? 0.1-0.5…??%
  • 0.32 – who you’re very sure is a pedophile? ~0.3%
  • 0.30 – who never learned to read as a child (though they might have learned to read after the age of 18)? (???%)
  • 0.10 – who’s been a victim of sex trafficking? This does not include consensual sex work. 0.02-0.04%

Sex trafficking came in dead last with a huge gap – well behind even not learning to read before the age of 18, which surprised even me. I also was surprised to find that the ‘actual’ numbers of sex trafficking in the US, as reported by the study from the top google result, reported roughly comparable numbers to the social visibility on my survey – one of the lowest total prevalence of any of the things I asked that I could estimate numbers for. This in itself might not even be true; sex trafficking numbers have historically been conflated with voluntary sex work.

While this evidence isn’t perfect – there’s a lot of small things that could nudge the results in different directions and I’d like to see it replicated by someone else – it’s still a piece of evidence in favor of “sex trafficking in the US is actually really, really rare.” Your chance of being a sex trafficking victim in the US each year is somewhere in between your chances of dying from bee stings and electrocution/radiation (but, to be fair, seven times more likely than dying from being struck by lightning. Seven!). Classifying this as an “epidemic” seems deeply dishonest.

“People wouldn’t report being sex trafficked” doesn’t seem like a very strong argument; it seems like social visibility of rape is pretty high, and that’s probably subject to the same pressures – social shame and fear. “You’re not asking the right people” also seems weird – I’m asking full social networks, networks which seem to have no issue picking up stuff like homelessness or 1+ years imprisonment.

The survey results are affected by ‘how public people tend to be’ about it; for example lung cancer, which hits the same percentage of the population as being a medical doctor, is ranked far lower in social visibility (0.9 vs 1.53). The suicide and homicide social-visibility numbers are depressed a bit because nobody could answer “happened to me”, but are still elevated because everybody talks about it when somebody dies. Interestingly, rape is more visible than “voted for Trump”; it’s not like I have a mega-liberal followerbase (~20% are Trump supporters and a lot more are libertarians), so I’m not sure exactly what to conclude from this.

(also; I’m using the word ‘visible’ here, but these numbers don’t differentiate between visible and prevalent; more prevalent, invisible things (e.g., prostitution) might get a similar social-visibility score to very visible, uncommon things (e.g., public figures))

Some more deets: though not all of my responders were from the US, I decided to use “rates in the US” as a baseline, assuming that the percentage of things (e.g., lung cancer, doctors, winning the lottery), were not hugely different in the US vs other western countries such as Canada or France.

I got respondents mostly from twitter and reddit.

I gave detailed instructions about stuff like “only report for people you’ve known in the past 5 years”, and stuff like who qualified as an acquaintance. I included a hidden test for people to verify they’d read the instructions, and I doubled the score weight for people who passed the ‘I read the instructions’ test.

I’m planning on going through the data a bit more generally to find other interesting things (I also asked everybody their a/s/l, so gotta see if there’s any juicy correlates around there), and will probably make another blog post eventually.

Velit sapiente autem ut consequatur ea assumenda voluptatem hic dolo