The Oregonian’s “fact check” on “average age” myth: They were aware of the falsehood for almost three years

Last month, while distributing a error-ridden “fact sheet” on commercial sexual exploitation of youth (CSEY/CSEC) to a roomful of audience members, Multnomah County Collaboration Specialist for CSEC Joslyn Baker said that the line that stated (inaccurately) “the average age of entry into commercial sex industry in the U.S. is 12 years old” might need to be changed because there is a “pushback from local media.” I did not know what she was referring to at the time, but surely enough, The Oregonian finally decided to fact-check the claim in its PolitiFact Oregon column last week.

While I was glad to see that The Oregonian now officially acknowledges that there is no basis for this oft-repeated yet demonstrably false claim, its investigation falls short of what I, just a local activist, wrote almost three years ago; in fact, everything PolitiFact Oregon writer Janie Har examined and wrote about in her column was already in my three-year old blog post, including the 2001 University of Pennsylvania study, and Shared Hope International’s 2009 report that shows a pie chart that does not match the claim made in the main text and does not include a citation. Har contacted Shared Hope for further clarification and received no response, which is what I had already done in July 2010.

Furthermore, Har makes this issue about the truthfulness of Multnomah County Commissioner Diane McKeel’s column published in August 2010 as well as her more recent comment to Oregonian columnist Elizabeth Hovde this past January, but somehow fails to point out that Elizabeth Hovde claimed in her Oregonian column that “the average age of entry into prostitution is 13” the month before Commissioner McKeel wrote it. More than likely, The Oregonian misled Commissioner McKeel in the first place, before it realized the mistake and decided to blame it on her.

I first read the claim about the “average age” in Hovde’s column, and was immediately suspicious. There are certainly young people 13 or even younger who are sexually exploited for commercial gain, but they are definitely outliers. If the average age was actually 13, there would have to be many more 8 or 10 year olds being forced into prostitution that is realistically possible to counter-balance all other people who are entering in their late teens or adulthood. So I started looking up the source, and it was very easy to find out that the figure was not based in any actual evidence. My investigation led to my writing the aforementioned blog post which explained how the claim was false, in more detail than Har was able to articulate three years later.

I also contacted Elizabeth Hovde on July 3, 2010 to share my findings.

From: Emi Koyama
To: Elizabeth Hovde
Date: July 3, 2010
Subject: Average age of entry

Ms. Hovde,

I’m contacting you to correct the error in your Oregonian column about sex trafficking.

You cite USDOJ as the source to state that “the average age of entry into prostitution is 13.” This is incorrect.

DOJ has not conducted any such study, but cites a report from researchers at University of Pennsylvania titled “The Commercial Sexual Exploitation of Children In the U. S., Canada and Mexico” (attached for your viewing). This report is NOT based on representative sampling of people who work in (or have worked in) prostitution, or even of youth who are/have been in prostitution, as the report itself states.

The “average age” is also based on the minors who were studied, which means that no adults in prostitution, including those who started working in prostitution after the age of 18, were included. This is like studying the average age of death for those who died as minors: the average age of death would be probably something around 13, but it has no bearing whatsoever on the average life expectancy of the general population. Similarly, the “average age of entry” among minors has no bearing on the actual average age of entry for all people who are or have been in prostitution.

Further, because the study only surveys minors, those who entered prostitution early have much greater chance of being studied than those who started at 16 or 17. That is, someone who started at 13 has five years to be studied by the researchers (because that person can be 13, 14, 15, 16, or 17 at the time of the study), while those who started at 17 only has one year. As a result, the number of people who entered at 13 are inflated by the factor of five compared to the number of people who entered at 17.

But even without knowing this, common sense should tell you that the average age of entry cannot possibly be 13. Let’s consider two possibilities: 1) the distribution of the age of entry is normally distributed (bell curve), or 2) it isn’t normally distributed. If the age of entry is normally distributed, that would mean that there are equal number of 8 year olds entering prostitution as there are 18 year olds–which you know isn’t true (if it were, we’d see much more media coverage about 8 year olds being prostituted). If the distribution isn’t normal, it would likely mean that there are many times more 11-13 year olds entering prostitution compared to 16 and up (to compensate for the fact there are very few pre-teens entering prostitution)–which once again cannot be true.

The only conclusion that is consistent with logics and common sense is that the average age of entry isn’t 13, but is closer to 18 or older. That doesn’t diminish the fact that some 13 and 14 year olds are being recruited into prostitution, and we should do something about it. But we need to keep our conversations based on reality and reason, rather than falsehood and panic.

(Cc: to Dr. Stephanie Wahab, Regional Research Institute at Portland State University. If you need help deciphering the UPenn study, she might be able to help you better than I can.)

Emi Koyama

Here’s her reply:

From: Elizabeth Hovde
To: Emi Koyama
Date: July 5, 2010
Subject: Re: Average age of entry

I appreciate the information, Emi. I was using two sets of information. The DOJ Web site lists the Pennsylvania info. The DOJ also commissioned a study with Shared Hope INternational (The National Report on Domestic Minor Sex Trafficking), which found a 12-14 age of entry in field surveys in 10 cities (I think … need to revisit to be sure).

I will contact appropriate sources about how to phrase the information better. You’re right that my column should make reference to the fact that the average age of entry for a YOUTH is 13, rather than a blanket statement. I will relook at my column and list a correction as needed. I haven’t visited the column since the edit. Does that phrasing make sense to you?
Thanks for the note and the concern.

Onward,

Elizabeth

I was not aware of the Shared Hope report, so I downloaded a copy and started analyzing it.

From: Emi Koyama
To: Elizabeth Hovde
Date: July 5, 2010
Subject: Re: Average age of entry

Hello – thank you for the reply.

I’ve looked up the report (“The National Report on Domestic Minor Sex Trafficking from Shared Hope International”), but I can’t locate any original research by SHI. On page 30 of the SHI report, there are two statements: “The average age that a pimp recruits a girl into prostitution is 12 to 14” and “Research has shown that the average age of entry into prostitution and pornography is 12 to 14 years old in the United States.” They are sourced to SHI’s training material titled “Prostituted Children in the United States: Identifying and Responding to America’s Trafficked Youth.” So I searched for and found this material on SHI’s website and looked into it, but it cited the DOJ as the source for this “statistics.” In other words, SHI does not seem to have an independent source other than the same DOJ fact sheet (which misrepresents the UPenn study) for the “12 to 14 year old” figure.

There is also a chart and table titled “Average Age of Entry into Prostitution,” with the breakdown of victims’ ages (11 thru 17). I find this chart puzzling for a couple of reasons. First, it doesn’t cite any source for the data, so I don’t know which study it is supposed to represent. The sample size is n=103, while the UPenn study included 63 boys and 107 girls. Second, if you actually calculate the numbers shown in the chart, the “average” os actually almost 15 (14.89), which contradicts the claim that the average is between 12 and 14. The median age (which is probably a better indicator of general tendency) is 16.

That said, it’s still not accurate to state that the average age of entry for a youth is 13. It is accurate to say that the average age at which UPenn research participants reported to have entered is 13, but there are many problems with generalizing this figure. The biggest problem, as I’ve explained in the first email, is how the research method artificially inflates the number of participants who entered into prostitution earlier.

Let’s say, for the sake of argument, that an equal number of youth enter prostitution at age 13 and at age 17. Members of the first group might be interviewed when they are 13, 14, 15, 16, or 17, while the latter can only be interviewed when they are 17, because those who become older aren’t included in the study. In any given year when the research takes place, we could encounter 13, 14, 15, 16, and 17 year olds who each entered when they were 13, while we will only interview 17 year olds form the latter group, because those who entered at 17 and are now 18, 19, 20, or 21 are not part of the study. When the study is compiled, you would find that there were five participants aged 13-17 who all say that they entered at 13, while there was only one person who reports having entered at 17. It is still wrong to conclude that youth is five times as likely to enter into prostitution at 13 compared to entering at 17, since we started from the scenario that the number of 13 year olds and 17 years olds entering prostitution is equal.

Thus, since we are only studying youth 17 and under, we can’t simply add numbers (age at which the participant has entered into prostitution) and divide by the number of participants to calculate the “average age of entry”; we need to compensate for how the study excludes those who entered into prostitution toward the end of the cutting-off point (18), and make adjustments. Using the data from the chart/table on page 30 of the SHI report, I calculated the adjusted “average age of entry,” which turns out to be almost 16 (15.96). This number (age 16) still isn’t entirely satisfactory, since the research subjects aren’t randomly selected and there are other ways errors can happen (for example, some participants were interviewed in groups with their peers, which is known to distort the data). My calculation is also rudimentary and at best an approximation, since I don’t have access to the complete data. But I suspect that it is much closer to reality than the 12-13 figure.

And details like this matter. Social policies we must enact to prevent and stop sexual exploitation of minors would differ greatly if the average age of entry is 13 or 16. I feel that many people use the lower figure for shock value, to arouse strong emotional reaction toward the issue, but the distortion of reality is irresponsible. We need to understand reality as they are and craft rational and sensible responses to the problem, rather than indulging ourselves in panicked frenzy.

Emi Koyama

I did not hear back from Elizabeth Hovde again.

In conclusion:

• Yeay! The Oregonian acknowledges that the claim is baseless! (But why is it rated “half-truth” if there is no basis for it? And why did they not mention any other study that contradict 12-14 claim?)
• I have a feeling that Janie Har read my blog post. How can she not, if she actually did any research? The fact that she mentions the same Shared Hope report and points out the same problems strengthens my suspicion. If she did read my blog, why did she not speak with me or give me credit?
• Janie Har writes as if the problem comes from Commissioner McKeel, and Oregonian columnist Elizabeth Hovde simply wrote down McKeel’s comment. But the truth is that Hovde herself perpetuated the false claim before McKeel did.
• Elizabeth Hovde has been aware of the problems Har points out about the University of Pennsylvania figure as well as the Shared Hope report, because I pointed out the exact same problems almost three years ago.
• Elizabeth Hovde and The Oregonian had the opportunity to stop perpetuating the myth for almost three years, and yet failed to do so as recently as this January. While Janie Har’s column is to be commended, The Oregonian and Hovde need to take responsibility for their part in the falsehood, rather than simply blaming McKeel for it.