So remember like a month ago when i asked people to help me come up with virginity quest films for a paper I was working on? I’m still working on that paper. Yesterday, while working on this with my project partner, Miranda, we decided to start tracking real life statistics on virginity. If you follow me on twitter (@chrismaverick) then you might have seen me tweet out my frustration about it:
You'd think on the internet, it'd be easy to find a source of avg age of virginity loss over time. You'd think. It's the internet. Nope!
— Mav (@chrismaverick) October 26, 2014
It’s funny. I was like “I have a sex question. But hey, I have the internet. That’s perfect. Sex is what the internet was invented for, right?” I mean… that and cat pictures. You’d think it would be trivial to find this stuff. Well, you’d be wrong. It turns out it’s ridiculously hard to get the data I was looking for online. You see random pseudo statistics that basically say “kids are having sex younger and younger these day. The age of virginity loss has decreased by 5 years! Teen pregnancy is on the rise! We need prayer! We’re all going to hell” and such. Or you get articles that say stuff like “Teen sex is actually on the decline. According to ‘a study’ the age of virginity loss is up 3 years!” But no one ever actually provides raw data or a citation or anything useful like that. Not random blogs, not government websites like the CDC. Not academic journals. Not the Kinsey Institute, which you know, pretty much EXACTLY ONLY EXISTS TO DO THIS SORT OF THING. Several of them post sources for a survey JUST for the year in which they are writing, but no one uses historic data. In fact, even news outlets make the error that I have complained about while grading dozens of student papers: according to a study… WHAT FUCKING STUDY?!?!? Isn’t it like so obvious that your argument immediately seems stronger if you write the name of the study? That’s science, dammit!!!!
Anyway… after hours of web surfing, and some help from others who saw my tweet/FB post, I eventually found a study by the Guttmacher Institute. This was just what I was looking for (here’s a PDF to the full paper, because I’m a better person than everyone else)! One of the most interesting things about it, is that this was published THIS YEAR. Basically, the only reason I was able to find this out at all, is because last year, someone had the same question as me, and they ended up doing all the tabulating. Well, good, at least I don’t have to do it.
So their research methodology is interesting. They tabulated from bunch of different sources… they were able to get cohabitation and marriage data from the census. That makes sense, but their primary sexual data they gathered from the National Survey on Family Growth (NSFG), a periodic CDC study of that is actually devoted to discovering trends in childbirth and contraception use amongst childbearing age Americans. They actually even had more information than I needed. For instance, they track the deviation in menarche amongst pubescent women over time, which didn’t change as much as I actually thought it would, but that’s a different topic (but you know, in case you’re wondering why I thought there’d be a shift… Bovine Growth Hormone) and when birth control pills are first used. Anyway, here’s my version of the chart (click it to make it bigger):
So a few really interesting things of note. First off, to read the chart, the years represent BIRTH years, not what was happening in that year. So, that means in 1960, the average age a female lost her virginity was 18.5 and the average age a woman had a baby was probably around 21 or 22. The numbers listed for 1960 (18 and 25.5) actually occur when women born that year reach that age (so 1978 and 1985, respectively). Second, there’s no data for men before 1960, because the survey covers reproductive age adults (15 to 44) and it didn’t occur to anyone to survey men about this until 2002 (they started with women in 1982), which says something interesting about gender studies right there. You’ll notice several of the lines also drop off after the 1980s. This is because the ages for that event have risen past the point where it can be meaningfully plotted. That is to say, that the median age of marriage for men born after 1985 appears to be at least 30, and therefore hasn’t happened yet.
Second, a few interesting points come out of the chart. The age of virginity loss hasn’t really changed that much. Assuming honesty in the poll, (more on that in a bit), the age of virginity loss for women has only varied between 19 and 17 and currently rest right in the middle of that gap. Men has varied between 17 and 18. Marriage and childbirth ages for women are both relatively on the incline. Marriage for men is slightly on the incline, but childbirth has actually remained relatively constant. This basically means, despite what you hear from conservative abstinence groups, teen pregnancy is way down. One of the really neat things you get to see is the converging of lines. Girls born in the mid-70s and after, apparently don’t see saving their virginity as anything more important than boys do. Also around the same time women basically start getting married because they are pregnant, or so I would guess from the data (the same thing appears to happen for men born in the late 70s, early 80s).
I was discussing the data with other people who responded to my quickie tweet and there is a question about what’s going on in 1960. It would appear that 17-year-old boys are commonly boinking 18-year-old girls (and probably being considered sluts for it). But I actually expect that’s not the case. I expect it is more likely that a larger number of sexually active males are doing it with a smaller number of sexually active females of the same age or slightly lower. Keith Irwin broke this down in a comment on my Facebook:
For example, let’s consider that we have a population with four boys and four girls. Hypothetically, the boys are willing to be sexually active starting at age 15 and so is one of the girls, but the other girls are waiting until 18. Each year, they pair up into different couples (fairly long-term relationships). The first year, one boy and one girl lose their virginity at 15. The second year, one boy loses his virginity at 16, but no girls lose theirs. The third year, one boy loses his virginity at 17, but no girls lose theirs. The four year, one boy and three girls lose their virginity at 18. In that scenario, everyone lost their virginity to someone the same age as them, but the average age for boys was 16.5 and the average age for girls was 17. That doesn’t even require that the girl who starts earlier be particularly promiscuous or have a bad reputation. Given our attitudes towards sex and virginity in our culture, it would not be at all surprising to me to see scenarios like that play out in practice and that would likely cause the age for girls to skew higher.
Finally, and probably of most relevance for my project which is about loss of virginity as a marker of coming of age, is the fact that even at the earliest point in the study (those women born in the 40s), the average age of virginity loss was still three years before the average age of marriage. That means that at no point in history of anyone likely to be sexually active today was “saving yourself for marriage” really a thing. It’s always been a cultural invention that is claimed but apparently not lived up to, at least for most people.
Of course, there is still a question of data validity. The Guttmacher Institute claims that “the NSFG is arguably the best source of data on U.S. women’s and men’s sexual, contraceptive, and childbearing behaviors” and I am inclined to agree. But, as Laura Stewart pointed out on my Facebook post:
You should note that consistently these studies reveal a pattern that’s numerically impossible. People lie based on differential gender pressures. This can be somewhat explained by a small number of women having a very large number of partners and then not getting recorded in random samples. But when these patterns show up in every study, that explanation starts to seem fishy. I suspect similar fishyness with Keith’s interpretation of this similar data.
I’m not saying we can’t get ANY information from the study you found, but we should really really take it with a grain of salt.
So I do wonder, how much can we trust this data. But, this is a problem with any self-reported survey data. You can’t force honesty, especially when it comes to sex. That said, I’m inclined to believe that their sample size (which ranges from 7600 at the earliest studies to 23,000) probably does a decent job of normalizing out any fallacious data, at least enough to read an actual trend. In particular, since cultural attitudes, particularly at the earlier times of the study should have forced any lies to imply that (at least for women) loss of virginity and date of marriage were coincident with each other and the data shows this not to be the case it is probably relatively trustworthy. I would doubt that the age of marriage, childbirth or cohabitation is incorrect (and the Guttmacher report confirms that those are all inline with US Census data). If anything, I’d say the biggest lie we could possibly assume would be that the age of first sexual experience would be lower for any point on the timeline (particularly the older ones) which would mean that if the data is lying, virginity might be even more steady around the 17-18 year old line.
So anyway, I’m interested in people’s thoughts. Both about the data and the methodology. What do people think?