LG, your post on "deviant behavior" says that 10% of all US males have had experiences with prostitutes."  I have no reason to doubt that, but that doesn't tell us much.  What I would like to know is the extent of hobbying among men who have sufficient disposable income to indulge occasionally or as a regular.
This is just guesswork, but here's what I'd estimate:
TER males range mostly between the ages of 35 and 70 with income between 100-300,000 per year.  More TER males make more than $300,000 than less than $100,000.  Among males as a whole who fit this category across the US, I'd guess the number who are occional or regulars in the hobby are much higher than 10%, probably up around 60%.  True?  So:

1.  What is the % of ALL US males who are occasional or regular hobbyists in the cohort between ages of 35-70 and with a minimum yearly income of $100,000?  

2. Would frequency of hobbying vary significantly between urban and rural populations?  I would think so, with much greater frequency in cities over 500,000 pop.  

3.  Is frequency greater among the Switzer-type population of politicians, businessmen, celebs and conservative clergymen who gravitate towards high end providers (over $500/hour)?  

Any figures or plausible guesstimates on these areas of male behavior?  Males I know who fit this category, including friends and neighbors, hardly make their dalliances public, but I'd guess the percentage, if we could know it, would be much higher than people think.

In other words, quality hobbying is far greater than anyone cares to admit, and responsible research is not encouraged by the citadels of respectability.  Is this a plausible supposition?  I'd think hobbyists and providers would also have plausible insights on these matters.

my guess is that you are vastly overestimating the income of TER users

nice humorous potshot at conservative clergymen as well . . . well played

...when you include Street walkers, escorts, sugar daddies etc.

Only 10% of the US population has income above $100K
so I think your estimates of the TER population's disposable income is overstated.

It's entirely possible I've overestimated -- it's clear I'm no statistician.  And the income category of $100,000-300,000 is very broad, maybe too much to be statistically meaningful.  But I would guess that there must be millions of US males between 35-70 in this income group, and just extrapolating from what I've subjectively observed as an urgan male somewhere in this category, I'd still guess the figure of occasional players might still be close to 50%.  You see?  I've already reduced my supposition from 60% to 50% under pressure from you experienced observers. I appreciate your observations.

Dear Polaris 2,

In the absence of more reliable statistics, I will stick to published research of prostitution studies worldwide. In the USA, appx 10-15% max. In western Europe, about the same. In Thailand, anything from 30-70%.

"Responsible research" is most certainly encouraged, but trying to find a representative sample is another matter. The last random samples in various countries indicated the figures above.

"Plausible insights" is not what we're going for in research,
The Love Goddess

sample was as non-representative as anyone could get at the time. Going to prostitutes is NOT a common thing among men, despite what this site may lead you to believe.

I do have the 411 on this one,
The Love Goddess

10% = 14 million people.  

Also, before internet age, I think that number was much lower (maybe less than 1 million?).  Internet has been around for only about 20 years, and it has been serving the hobby for maybe only about 10 years.

If that's the case, in 10 years, internet has created over 10 million hobbyists.  What will it be in the next 10 years?

Also, internet has attracted a new different population of people into the hobby.  Providers population improved tremendously as well.  This will have a compound effect going forward.

Actually only a little more than 5% of individuals earn more than 100k. The middle class of the US earns far less than most people think.

I think a lot of people lie even on a confidential poll.

My guess is the percentage is a lot higher than reported because people lie, even to themselves.

...since its inception 10+ years ago.  TER covers only major 25 metropolitan areas.  I'm sure there are men in Bumfuck, Nebraska who patronize their local providers yet have never heard of TER.

For every one TER member, you can conservatively estimate that at least 20 others have paid for sex.  This means that 25,000,000 men out of a potential 100,000,000 men of hobbying age (150,000,000 - 50,000,000 too young or old) have paid for sex, or at least 25%.  

Personally, I believe one-third to 40% have paid for sex at one time or another.  It doesn't matter how secret a survey is, you'll never get more than 10% of American men to admit to visiting a prostitute.

My hunch is that your last sentence is on target, BPS.  LG stated that "going to prostitutes is NOT common," but my query was prompted by skepticism over the low estimate of US males having visited a prostitute, and although I'm only guessing I think that estimate is way too low.  Also, I was prompted by the earlier post about "deviant behavior," and the widespread moralistic idea that hobbyists and providers manifest "deviant behavior."  My opinion is the hobby is a mainstream activity, and of itself there's nothing "deviant" about it.

Dear BigPapasan,

Unless you have an in with the administration at TER, you will never know exactly how many men have created accounts at TER. TER is also patronized by providers, madams and others in whose interest it is to create and maintain accounts.

Nor will you know how many of those men acted on meeting with escorts/massage providers/tantrikas, etc. Some people buy a membership out of curiosity, look into it and decide not to pay for sex.

Unless the sample is representative, it is useless to speculate and state that "you'll never get more than 10% of men to admit." Reliable statistics are based on probability samples, among other things. For this, the sample needs to be random. And we don't have that here at TER.

Thank you,
The Love Goddess

LG, I can well understand why you, engaged in formal research, have a disdain for anecdotal information and subjective opinion.  But I don't think the 10% figure is sustainable in light of the factors that several posts have mentioned.  Where's the beef behind that figure?  I've done public opinion research and used other research methodologies on various issues, and in some cases the "gut feelings" I got from people on the ground and who were directly involved were much more accurate and informative than what "objective" research turned up.  I'm not disuputing any of your points, but only stressing that we have to pay due attention to plausible subjective information (e. g. the opinions of informed participants) when objective research on the hobby is so limited and media and interest group commentary is so often biased.

My answer to the question would be Heck NO.

...TER members there are.  A poster on the L.A board recently mused about his 10+ years on TER and the fact his Member ID no. was under 400.

This piqued my curiosity because I didn't know my own member ID no. and couldn't find it in the Account Manager.  After a little searching, I located mine (around 64,000) and also found it is possible to determine many member's ID nos.  It is TER public information.  Using this method, I found a hobbyist who joined in May, 2009 had an ID no. close to 1,200,000.

Providers have separate TER ID nos. There are approximately 150,000 providers who have been reviewed.  The number of providers and madams who would set up accounts as hobbyists is negligible as they would readily be discovered by TER.

Maybe it's anecdotal, but at least 50% of the men I know have paid for sex.  I guess you could chalk it up to the decadent California lifestyle, but more likely, the 10% figure is just the tip of the hobbyist iceberg.

especially considering it's illegal. What I find most fascinating about the numbers that LG brings up is the similarity in percentages with Europe, where in most countries is either legal or not persecuted ( other than for traffickers and pedophiliacs). That speaks a lot of how little criminalizing it does to change sexual behaviors. Cultural, sociological factors being similar in these two samples adds to that interpretation.
My curiosity however is whether 10% is a number sufficient to explain how a specific behavior can persist through centuries and civilizations.

It's actually a randomized study done with computers. The responses were also scored by computers. It's not like a census taker who knocks on doors, nods and winks in a conspiratorial manner. Check out the attached, please. Go to page 24-25. Also, check out the GSS link at:

http://www.norc.uchicago.edu/GSS/GSS+Resources.htm

Although that would be funny,
The Love Goddess

Thanks LG for the links.  I checked them out, and they confirm that men have probably under-reported their contacts with prostitues, which is what some of us have been asserting anecdotally.  I read further in the references, and while I have the greatest respect for valid objective studies, what I read in this case, and in many other instances of research on various topics, is that data only confirms what common sense and experience have already ascertained experientially.  Now, this is assuredly not a criticism of you and your research.  The TER study you have already done of 100 providers is full of fresh insight and useful information.

between "underreporting;" my study which had nothing to do with a random or even remotely representative sample; hunches; and the 40-50% being speculated by some posters in this thread.

Assumptions drawn from anecdotal data or membership of TER have nothing to do with representative samples, random sampling and social surveys.

Once and for all - and I really mean this - read mrbonobo's post and reflect; 10% of the ENTIRE American population is about right.

TER members = snowblind to stats, it seems,
The Love Goddess

To understand the problem read the following explanation: "The Myth, the Math, the Sex" ( http://www.nytimes.com/2007/08/12/weekinreview/12kolata.html ) from the New York Times. It is clear enough there is no point in my echoing what it has to say.

For a much more detailed analysis of the problem see the Related Link.

There is a phenomenon that the military calls "mirror imaging" (maybe psychs have a different term for it).  I've never been in the military, but as a market strategist and long-range planner, I adopted their term to explain the human tendency to assume that others share our values, or think and act as we do.

From the military's standpoint, it's critical to understand that your enemy might not be fighting by the same set of rules as you do (kamikaze pilots or jihadists who want to die, not stay alive).  For marketers, it's important to remember that not everyone shares your taste or preferences, and it's also helpful in removing one's personal bias from a discussion or decision.

I think hobbyists tend to fall into the trap of mirror imaging as a form of rationalization that what they're doing must be OK, and they do this by projecting their behavior onto a large percentage of the population.  They seek to justify a behavior that we all know would be looked down upon by society (and that's illegal) by saying that most men do it, therefore it's OK.

Lots of other special interest or minority groups do it too, not just men who pay for sex.  The gay community, for example, would have you believe that nearly everyone is gay, wants to be gay, has thought about being gay, or has a gay experience of some sort.  They do this to make being gay more mainstream.  Maybe many people do, but not in the numbers that are tossed about when an advocate is trying to make a point or advance an agenda.

This is the reason you do quantitative studies in the first place- to remove biases, to base conclusions on data that can be reproduced by others, and not anecdotal experience.  And perhaps most importantly, to identify results that are representative and can objectively be expanded to the larger population as a whole.

People who are quick to dismiss large quantitative studies based on the imperfections of any study, somehow never see the problem with relying on anecdotal experience or their own personal experiences.  I've never understood how a sample of one is supposed to yield better results than a sample of 500 or 1,000, but this seems to be the basis of their argument.

10-15%? Wow! I'm shocked, I would have thought much, much higher. But it's your business, you're more likely to know than me, I guess, though I seem to recall having seen a couple of times that 50-66% of men had paid for sex **at some point** in their lives.

Does paying for a handy at a strip club, for example, count? Or are we talking just regulars?

When you analyze it, you bleed all the interest from it.

Here we go, straight from admin:

The over 1,281,714 registered user names represent anyone that has ever created a user name, including agencies and providers.

Thank you,
The Love Goddess

maybe you feel better thinking everyone or half of the population cheats, not true. Not even close when it comes to p4p. that would mean half or your friends are payin for it.

Ok... I haven't read the second article yet, but there is a major flaw in reasoning apparent in the NY Times article. I'm not sure whether this error can be attributed to the reporting or to the mathematician quoted; I really, really hope it's the former.

Anyway, in one paragraph, it says:
"One survey, recently reported by the federal government, concluded that men had a median of seven female sex partners. Women had a median of four male sex partners."

In the next paragraph, it says:
"But there is just one problem, mathematicians say. It is logically impossible for heterosexual men to have more partners on average than heterosexual women. Those survey results cannot be correct."

This is not true; it is ENTIRELY possible for men to have more partners "on average" depending on what kind of "average" is being given. Basic high school statistics teaches the difference between  mean, median and mode, which can all be considered different kinds of "average." If you have a bunch of numbers, you get "mean" by adding them all up and dividing by how many numbers are in the set. You get "median" by arranging them in order from smallest to largest and picking the number one in the middle. You get "mode" by choosing the number that appears most frequently.

If everybody was being honest, I agree that it is mathematically impossible for the MEAN number of partners to be different for heterosexual men and women. However, it is ENTIRELY possible for the MEDIAN number to be different, which was the "average" that they referenced.

Let's do an example.

Imagine you have 10 men and 10 women, who answer surveys with 100% honesty.

Initially, all are neatly paired off into monogamous couples. Each man has only one partner, and each woman likewise has only one partner. In this case, both the mean AND median number of partners equals 1, for both groups.

Now imagine the same group of people, behaving the same way. Except, suddenly, ONE woman decides to have sex with all the other men as well, not just her initial partner. The number of partners that would be reported would then be:
Men:   1  2  2  2  2  2  2  2  2  2
Women: 10 1  1  1  1  1  1  1  1  1

The MEAN number of partners (for both groups) is now 19 / 10 = 1.9. However, the MEDIAN number of partners is now 2 for men and 1 for women. Isn't that interesting? If you're reporting the MEAN, then men and women both average 1.9 partners. But if you're reporting the MEDIAN, men suddenly average TWICE as many partners as women! And yet no math has been violated.

The "High School Prom Theorem" provided is only relevant if you're concerned with the mean number of partners reported. (In this example, G = B = C = 19.) I'm going to assume that a UC Berkeley math professor would know this, and that it's the writer's mistake for referencing statistics that quote medians.

I'm not saying anything about whether I think people lie on surveys or not; I'm merely pointing out that the "evidence" cited in this article does nothing to support this claim. You would have to compare means in order to gain any insight on this question.

~Luci



-- Modified on 12/5/2009 2:21:58 AM

Sampling works well when the distribution of the characteristic being sampled is "normal" or Gaussian. It is usually assumed that the characteristic has a Gaussian distribution in the population. When this is true the mean, median, and mode are all the same.

The second paper cited makes the point that the distribution of the characteristic "number of sexual partners" is far from Gaussian in the population. Your example of a closed population of ten men and ten women (which I like) illustrates this (as well as the intended point that people shouldn't confuse mean and median). Try obtaining the correct population mean from a sample of any three men or three women in your example. The sample mean and the population mean simply won't agree.