Thursday, July 02, 2009

Young women do get around, albeit once every few years

I previously attempted to refute the presumption thick in some areas of the Steveosphere that today's young women are nymphos. The GSS does not lend any credence to blogger Whiskey's assertion that it is not unusual for educated, urban white women to have more than 50 different sexual partners.

He has subsequently questioned the utility of the survey on the grounds that even for questions with 600 responses, for any given year of age there might only be 5 or 10 participants. The criticism is inane. Unless the issue at hand is how 53 year-olds feel about something, that doesn't matter. If it did, polling organizations that gauge things like President Obama's approval rating or the level of support for same-sex marriage would be out of business unless they could devise an economically viable way to 10,000 people for every poll they conducted. Instead, they conduct surveys with 3% margins of error in either direction using samples a tenth that size or smaller. If statistical reliability was desired not just for age in years, but age in months, the required sample size would grow by another order of magnitude.

But Whiskey's is the place for that to be addressed. The purpose here is to show that there is no massive shifting in female sexual behavior currently taking place in the US. The following graph shows the average number of male partners women aged 25-30 report having had since turning 18, beginning in 1989, the first year respondents were queried on it:

The trendline is pretty flat, perhaps rising marginally over the last couple of decades. The anomalously high mean in 2000 is the result of one woman reporting 122 partners. Removing her from the calculation yields a mean of 5.15, nearly identical to the 2002 result. For the last 15 years, young women of middling salacity have, with the exception of 2002, consistently reported having had three partners. The mode (most frequently occuring number of partners) in every year is 1. That is, if you ask a woman in her late-twenties how many people she's had sex with, the reply you are most likely to receive is that she's had just one.

The following table shows the percentage who say they have had ten more more partners as well as the total sample size for women aged 25-30 who were asked about their sexual behaviors:


The prevalence of 10+ women appears to have edged up a bit over the last couple of decades, with about one in seven having had partners in the double-digits by the time they reach the age of 30. Even sleeping with ten different people over ten years hardly qualifies as slutty behavior--it amounts to a different person each year. Not all women are made for long-term relationships, after all.


OneSTDV said...

Can you split this data up by race and/or socioeconomic status? Really curious how that would look.

Also, how dishonest can we assume these women are? Surely their answers would be affected by being asked the question face-to-face and being asked by a male.

silly girl said...

Yeah, this is not Lake Woebegone where all the women are above average. Every year there are more obese women in these numbers, seems that would impede finding partners. Also, I read the post carefully and didn't find the word "single" so I assume these numbers include married women the % of which generally would increase with age. In Texas 10% of women 18-20 are married. These numbers seem pretty believable.

Jokah Macpherson said...

Silly girl, it seems counterintuitive, but being obese somehow HELPS females in finding partners, at least as far as quantitiy is concerned. I don't know why this is but suspect it may result from a loss of "bargaining power."

This may be part of the answer to the reason for the slight upward trend in the 10+ partner contingent.

OneSTDV said...

"I don't know why this is but suspect it may result from a loss of "bargaining power.""

Black men love fat women.

TGGP said...

I agree with you on the sample sizes per year of age. I got into a similar argument with him elsewhere:

I do think Whiskey is getting at something important. I think the shift in gender relations is one of the most significant things that happened in the 20th century, and I'm surprised how little attention it has gotten from folks like, say, Mencius Moldbug who focus so much on race. Perhaps because America has such an unusual amount of racial diversity the issue stands out, but there are two genders everywhere and a similar change seems to be happening around the world. Justin Wolfers & Betsey Stevenson seem to have done the most analysis regarding this change.

The Undiscovered Jew said...


Check this out:

Sarah Palin Resigns

Hey UJ:

OT, but did you check out the preliminary reports out of Arizona regarding births for 2008? I know we've discussed this elsewhere, but I can't remember if the reports breaking down the numbers for 2008 had come out yet at that point.

In case not:


Latinos: 45,731
Whites: 42,216


Latinos: 42,639
Whites: 41,925

Seems like Latino fertility is falling faster than white fertility (3,000 fewer Latino births vs. 300 fewer white births) in Arizona.

Posted by: Mark | July 03, 2009 at 08:10 PM

Mark, I've actually been following the 2009 numbers. I didn't know the demographic breakdown for 2008 births had been released already.

Arizona births (no ethnic breakdown yet) in 2009 are down by 7.4% so far compared to 2008:

Posted by: The Undiscovered Jew | July 03, 2009 at 08:15 PM

Also see this for national 2008 birth data:

Posted by: The Undiscovered Jew | July 03, 2009 at 08:19 PM

Jack said...

I think this is actually pretty significant increase in female promiscuity. The % who have been with at least ten guys has gone up about 50%, from 10% or so to 15% or so. This means that there is an increasing subset of promiscuous women, and these women are likely to be the single ones in their late 20' dating pool as a 31 year old guy.

I would like to see the data for just white, educated women to see how they fare.

Audacious Epigone said...


To get more than "white", "black", and "other", we have to look from 2000 forward. Sample sizes are small for non-whites. The white mean is 6, the same as the black mean and higher than the Asian or Hispanic means, but again, the sample sizes for the latter two are too small to put much stake in (they're both in the 30s).

Income is tough to measure, because married women tend to have higher family incomes but often lower individual incomes.

Re: honesty of respondents, how does the presumption that promiscuity is rapidly on the rise spread? Previously, Whiskey argued that women understated the number of partners they had, but also pointed to a NYT (if memory serves) article about a couple of profiled women freely talking to an unknown reporter for a national newspaper about their free-wheeling sex lives, implying that it's everywhere. I'd think that an anonymous, detailed survey about a host of personal information would solicit the most honest self-reported data available, but that is only my speculative assumption.

Silly girl,

It includes all women that age, not just those who are single. If 75% of women were married and devoted to their husbands, the salacious behavior of the unwed minority wouldn't be much of a story.

Also, since women are getting married at a later age than in decades past, presumably this should cause some partner inflation.


I am not hostile to Whiskey's treatment of the (broad) topic and am only taking issue with instances where he makes claims that are refuted by available data.

Agnostic does a lot with gender, generally finding that re: interaction between the sexes, the changes have been overblown, and he backs his stuff up empirically.


I have. Once again, encouraging news!

Anonymous said...

Obesity may not be an issue for women when it comes to finding sex partners, but it's certainly a major issue for men. Obese men get less sex than anyone else.

When I saw that part about the woman reporting 122 partners I thought that maybe she was a hooker, but then figured that knowing the precise number didn't make sense. Could it be some sort of data entry error?


Billare said...

OT: To better complete my knowledge of your positions, TUJ, did you go by the nomen "JRR" on the "WA/us" forum? I ask, because I am curious and because much of my thinking on certain ideas has already been advanced there, and you might have some insights to share that I need only just Google for myself.

The Undiscovered Jew said...

To better complete my knowledge of your positions, TUJ, did you go by the nomen "JRR" on the "WA/us" forum?

No. I think Jewish Race Realist said that he is an Israeli doctor.

The Undiscovered Jew said...

Also, so you know, I'm half Jewish on my dad's side and half non-Jewish white on my mom's.

silly girl said...

" it seems counterintuitive, but being obese somehow HELPS females in finding partners, at least as far as quantitiy is concerned. I don't know why this is but suspect it may result from a loss of "bargaining power."

Jokah, what I infer from what you are saying is that it keeps them find a permanent partner, therefore more temporary partners. Is that your meaning?

silly girl said...


Jokah, what I infer from what you are saying is that obesity keeps them from finding a permanent partner, therefore they have more temporary partners. Is that your meaning?

Jokah Macpherson said...

Silly girl, you are correct, that is what I am saying. The main point I was making (the weightiest part of my comment) is that weight gain in women seems to contribute to the number of sex partners rather than having the reverse effect. As for the reason this is the case, I am only speculating.

David said...

I grew up in the wrong era!

Audacious Epigone said...


For white women with 16+ years of education (essentially a bachelor's or better) aged 25-40 (widened to get large enough sample sizes), mean number of partners since 18 by year:

1989: 6.36
1990: 4.21
1991: 5.09
1993: 4.91
1994: 5.44
1996: 6.54
1998: 4.36
2000: 4.71
2002: 5.08
2004: 5.10
2006: 6.45
2008: 4.20

A little higher than the average for all women aged 25-30, but not by much, and the age range encompasses another decade, giving them more time to accumulate partners.

Black Sea said...

This may seem a minor semantic distinction, but to me, anyway, saying that something is "not unusual" doesn't really mean that it constitutes the statistical norm, or anything close to it. For one thing, if I remember correctly, Whiskey was talking about a subset of a subset of women, and he was focusing on those who have adopted a certain attitude toward sex and relationships that isn't easily reducible to a statistical measure.

Let me try an example. I'm in my forties. It's not unusual for me to run across people around my age who used cocaine or other fairly serious drugs when they were younger, although I'd suspect that, from a statistical point of view, they're a distinct minority. What I mean by "it's not unusual" is that it's not unusual ENOUGH to be shocking or even particularly surprising. Similarly, Obama's oblique admission of cocaine use, or the rumors of George W's cocaine use didn't raise that many eyebrows, even though it's probably not the norm for people of their educational background and age groups.

Anonymous said...

Obese women prob have more partners b/c they have less self-control, but their fatness doesn't decrease the partner count since there's always another fat guy who'll have sex with them. There are even more fat guys than fat girls.

Audacious Epigone said...

Black Sea,

Yes, semantics can be played, but it seems fair to presume that, unless otherwise specified, "not unusual" refers to at least 1 in 10 cases.

Spanning 2000-2008, among white women aged 25-35 with 16 years or more of education (bachelor's+), 63%--nearly two-thirds--are married. The Sex and the City nympho is definitely the exception, not the rule.

Black Sea said...

I do agree that women who have 50+ sexual partners in their lifetime are statistical outliers, and I that degree of promiscuity is certainly unusual.

Whiskey said...

I don't think it's unusual, given self reporting as I noted here, note the number of women asked who responded openly about high number count partners. "I stopped counting after 50" was revealing. So too this and this.

There are sound reasons to think that a partner count of fifty or so by the mid thirties is not unusual. Is it the mode, or even the mean? No. But anonymous urban living, easy access to sex, and clubs and bars make living the life of Samantha for those who want it fairly easy.

I'll have to look at the data myself, I don't think you can trust it until you've loaded it into MySQL and done some queries to get a sense of say, how many White Women age 30 in 1998 were interviewed. If it was say, 6, then the sample would IMHO be non-random.

I'm still unable to get any sense of methodology and sampling throughout the years. If say, the interviewers just randomly called people on the phones years 1972-1980, and then switched to in-person interviews, that's a problem.

There's also a problem if the interviews were on say, college campuses with student volunteers asking their peers. A LOT of Social Surveys are junk like this, snapshots of college kids telling their peers what they think they want to hear. There's also geographic bias -- you're likely to get much higher numbers in NYC, Boston, LA, DC, and Chicago than say, Salt Lake City or Boise, where populations of young single people are too small to get away with anonymous behavior.

What's that bump in 2006 for White Women Bachelor's degree and higher? Things like that make me suspicious and want to dig into the data. Are you using MySQL, or the Web Tool? The Web tool is evil -- go with MySQL and Excel/Open Office. You can ALWAYS dump queries into text files, and then import into either spreadsheet for analysis. If you're a R wizard it can hook straight into MySQL, no problem.

I'll have to look at this data myself. There IS a map I think I may have seen on Spengler's site at the A-Times, with bubble maps by country for partners and then age of first sexual experience. China was low, with about 3 lifetime, 18 years of age, and Iceland incredibly high, 25 IIRC, with age 15 or so.

Has anyone seen this? It had the US but I don't recall anything but our relative position (not good).

TGGP said...

Whiskey, your evidence consists of anecdotes. That does not suffice to show something is not unusual.

It has already been explained to you why it doesn't matter there are very small sample sizes for specific ages-by-year in certain survey years. You can either answer the objection to show that it does in fact matter or move on to a different tack. But how can you expect to be taken seriously when you keep using logic your audience has told you it finds spurious?

How is the web tool "evil"? It's very quick to copy tables from there and paste it into an editor, all without leaving your browser. If you want to know how many people of a certain age were surveyed in a certain year, just run an analysis of AGE and YEAR with the web tool.

The GSS is regarded as the best single-source of sociological data, sometimes called "the gold standard" in the field. Without any evidence that it is the case you assume that a survey designed to be representative is making boneheaded errors. Some googling turned up this, where you can see the comparison of their sample population for different age/demographic groups vs the actual population shares. If you use the web tool and ask for statistics, it will show you the unweighted N along with the weighted N. I just ran AGE by YEAR and the weighted sizes are quite close to the unweighted ones.

Audacious Epigone said...


TGGP says it better than I am able to. The unweighted responses are virtually always close to the weighted responses, and correspond well to the general population (in terms of age, race, geographic distribution, etc).

Also, why the need for MySQL? The online analysis tools provide just about everything a user could want, and transferring data into Excel or SPSS to run additional analyses is not difficult (although it is a bit tedious).

googaw said...

These numbers do not go back nearly far enough to catch the radical change in Western culture. The last sexually conservative decade was the 1930s. (The 1950s were only sexually conservative during the first half; the second half was so radical I don't count the 1950s as a whole as conservative).

I expect if we could collect the numbers we would see a dramatic rise in promiscuity in the 1940s, a slight rise overall in the 1950s, two more dramatic rises in the 1960s and 1970s, and largely on a high plateau since then.

I am also skeptical of these kinds of studies based on self-reporting polls rather than solid evidence such as DNA. (We will someday be able to do some very good "who's the father" studies for past generations).

TGGP said...

googaw, you can use STD rates.

Rollory said...

"It has already been explained to you why it doesn't matter there are very small sample sizes for specific ages-by-year in certain survey years."

If by this you mean the statement in the OP, it's simply wrong. The issue at hand is how marriageable women are behaving (20-30, 15-35 at the outside), _not_ how 50-something women (or women in general) are behaving. The information needed here is fundamentally different from what is needed for a political poll, because all women of all ages are _not_ equal.

The older segments of the population represent behavior inherited from previous decades, not the current norm for the marriageable population, so mixing them in throws all the numbers off. And a survey sample of the marriageable age group that is particularly small is statistically not very meaningful. Whiskey's point stands.

Audacious Epigone said...


No, it does not stand. He claims the GSS is valueless because the sample size for married white men aged 53 in a single year is 8 or whatever. That does not make the database statistically unreliable, unless we are trying to determine what married white men aged 53 thought about abortion in 2002.

For the post at hand, the sample sizes for the age group in question is around 100 per year. The eight 53 yo married white guys from 2002 do not play into this at all. The sample sizes are ~100, not ~8.