Tuesday, May 14, 2013

Not the Virtuous Time You Think It Was

There's a lot of chatter going on in the world about the Mormon dialog around virtue, chastity, and virginity.  I've not really wanted to get into it too much because there are so many better articulated pieces about it.  While reading about it, there were a couple things that I wanted to explore more and a couple of things I wanted to emphasize in the discussion.

One of the common themes we hear in the Mormon church with resepect to sexual morality is that societal standards have deteriorated, and that we should return to those more virtuous days when sexual morality was the norm.

Let's be clear about something though.  Those virtuous days are a myth.  They never existed.

When I was in graduate school, I read a report from the Guttmacher Institute1 that claimed over 95% of Americans had sex prior to getting married.  What's more, that trend has been in place since before the 1950's.  I didn't believe the report was accurate, so I replicated the study using CDC data from the National Survey on Family Growth.  Much to my surprise, I found very similar results.

Premarital sex is the norm in the United States (and the world), and it has been for much longer than you think.

I recently listened to a podcast from Mormon Matters in which one of the participants described a article that reviewed births in the 19th century with a focus on births that happened less than 9 months after the wedding.  Unforutnately, they neither used specific numbers nor cited a reference that I can find, but I'm trying to verify the claim.

The evidence I have gathered so far indicate that 40% of women in the mid-19th century were giving birth less than 8 1/2 months after marriage2.  Certainly, some of these were preterm births, but if you add in miscarriages and women who were having sex with the good fortune of not getting pregnant before their marriage, I would imagine that percentage is still higher than 40%.

Another, less scientific, evidence I came across was in a book Tracing Your Family History.  The author advises people not to look for marriage records only prior to the birth of an ancestor, but also after.  He says:
Marriage records are often missed because people assume their ancestors practised the rule of 'no sex before marriage.' After many years as a full-time genealogist I have serious doubst as to whether many people have ever heeded that maxim...
So, anyway, I really doubt the past was any more virtuous than the present.  At least if by 'virtuous' you mean the distorted definition of the word that is in use by the Church (and rightfully, a definition that is currently receiving much criticism).  It would be more accurate to say that I really doubt the past was any more chaste than the present.  (I have no doubt that in many ways, the present is more virtuous than the past3).

If you're going to say that anything has changed between the past and the present, you are probably limited to saying that we talk about it more now than we have in the past.  There are good things about that, and there are bad things about that.  But when we have discussions about sexual morality, let's not fool ourselves into thinking that our ancestors were so pure.  They weren't.

1 Summary report and detailed report

MINTZ, Steven, and Susan Kellogg. Domestic Revolutions: A Social History of American Family Life. NY: Free Press, 1988. (summary here)

3 For instance, women who were raped in the past were stigmatized, considered damaged and impure.  Such a vile stigma is the antithesis of virtuous.  


  1. This post is full of mushy thinking masquerading as objective observations. There is a difference between "standards" and observed practices. For example, one foot is a standard of measurement and is comprised of 12 inches. A carpenter may hope to cut something according to the standard, but he may mess up and only cut it 11 inches. It was a mistake. But he doesn't change the standard and say, I now declare 11 inches to be one foot. The standard has not changed, but he has fallen short. There was a time in American society where the social standard was to abstain from pre-marital sex. The rate at which people successfully adhered to that standard has nothing to do with the standard. Standards do change over time. Sometimes for the better; sometimes for the worse. But if you seriously think that the standards are the same as they were 100 years ago, I've got a bridge I can sell you.

    Aside from your specious "standards not changing" argument, I would also go further to question your assertions about the rate at which people fail to meet those standards. Unless I'm wrong, which did happen once, didn't the National Survey on Family Growth start in the 70s? A little hard to detect a trend from before the 50s when the study started in the 70s. I realize that many of the women in the 70s were adults at the time, and thus may have provided data from the 50s, but that still wouldn't let you go back that far before the 50s to detect a "trend." Also, the very fact that you described it as a trend suggests that it was changing. Lastly, I've been around the block enough to question the assertions of places like the Guttmacher Institute. On a parting note, I won't take credit for this thought, as I saw it on a comment board from someone else. If you stick one fit in boiling water and stick another foot in a bucket full of ice, statistically, you're fine.

    1. You may have a valid case that the language is not clear enough. If you look back at the second paragraph, I think you are applying the content of this post to "societal standards" instead of "those more virtuous days when sexual morality was the norm." The rest of the post goes on to describe how this norm (as in, normative behavior) wasn't the reality of the past.

      To borrow your carpenter example, it would be more like half of the carepenters in the country cut wood to 11 inches instead of a foot, but still went about their business like it was a full foot and everyone just ignored deviation. And then when they dropped the pretense and admitted that they were only cutting it to 11 inches, everyone declared that carpentry practice had started going down the tubes even though nothing had actually changed in the practice.

      If you want to argue that we were better off in the days when people were having premarital sex and not talking about it, you're welcome to do so. You'll just be very hard pressed to convince me.

      Again, the post isn't addressing societal standards. It's addressing the actual behavior.

      Regarding the National Survey on Family Growth, yes, I believe you are correct on it's start date. It also surveyed people as young as 15 and as old as 44. Obtaining data on sexual behavior as early as the 1950's is certainly plausible.

      "Trend" is an interesting word. Collquially, it means change, but it's mathematical representation is usually a slope. A trend of no change means a slope of zero. Alternatively it can mean "the trend of high sexual activity." Both uses are statistically appropriate.

      You may say what you wish of the Guttmacher Institute, but professionally speaking, I've rarely found their statistical methods to be lacking. Furthermore, I already stated that upon replication of their study I found similar results.

      The quip, "If you stick one fit in boiling water and stick another foot in a bucket full of ice, statistically, you're fine." is comical, but fundamentally misunderstands the nature and purpose of statistical analysis, which is to make descriptions, inferences, and decisions about populations, not individuals.

    2. One question, then, and answer honestly.

      Is it possible for someone to come up with a sound statistical analysis that is contrary to your position in this post?

  2. I entered my question somewhat off the cuff. I did a quick google search. I know nothing of statistics, but this looked interesting:

    "For all ages combined for each of these periods, the proportion of adolescent women who reported having had premarital sexual intercourse increased steadily (from 28.6% in 1970 to 51.5% in 1988 (Table 1)). For each 5-year period from 1970 to 1985, the amount of increase declined (i.e., during 1970-1975, 7.8 percentage points; during 1976-1980, 5.6; and during 1981-1985, 2.1). However, from 1985 through 1988, the proportion increased 7.4 points, or approximately one third of the increase in premarital sexual experience among adolescent women for the entire period 1970-1988. This trend persisted even after adjustment for the influence of changing age composition by comparing age-adjusted proportions.

    For each year of age during 1970-1988, the proportion of adolescent women who reported having had premarital sexual intercourse increased at least 55% (Table 1). The largest relative increase occurred among those 15 years of age (from 4.6% in 1970 to 25.6% in 1988). The cumulative absolute effect of these changes was greatest among women 18 and 19 years of age."


    No doubt you will scrutinize on a level that you refuse to apply to the Guttmacher Institute.

    1. My reply to your earlier comment was going to be that it isn't a matter of statistical analysis, but of data.

      The study you have cited is interesting, and notable. It differs from the Guttmacher Institute in population. Your CDC report only looks at the surveyed 15 - 19 year olds. The Guttmacher institute did not limit its age groups. Also, the CDC report seems to be reporting binary trends, whereas Guttmacher used cox proportional hazard modeling. Neither is more appropriate than the other, but they will shed light on different aspects of the data.

      Looking at the two studies in tandem suggests that the rate of premarital sex is unchanged, but the age at which it begins has shifted down. There are probably a few factors that contribute to this. One may be that we talk about sex more. One may be that of opportunity.

      One of the things I looked at when I did my reanalysis of the Guttmacher report was whether having a parent at home made a difference in when a person started having sex. I found that people who had a parent at home tended to begin having sex later than those who did not have a parent at home. This effect was seen up until age 18, but withered away by 21 (the college years).

      That change in age is a topic interesting of discussion and investigation. But it doesn't convince me that past generations were more chaste. They were just quieter about it, and maybe a little later.

    2. Well, you got me with "cox proportional hazard modeling." Damnit!

      Just to recap, though: It looks like you've conceded that (1) standards regarding premarital sex have changed, and (2) women are having premarital sex at a younger age. Consequently, I think the Church is spot on in telling the youth that standards have changed and admonishing them to wait until marriage. Maybe that's just silly old me.

      One final note, because I really am done with this insanity, your post doesn't say anything about promiscuity. I suppose in your world those unchaste folks in the 1800s were having premarital sex with as many different partners as people do today. Also, since there obviously wasn't any sexting going on then, are there any studies on the prevalence of sex-telegramming?

    3. (1) I'm not so sure. There are too many variables to consider if the attitudes toward premarital sex have changed. Just off the top of my head I can list birth control, abortion, women's suffrage, the industrial revolution, population concentration in urban areas, and expansion of public education. You're asking me to draw comparisons about whether people 100 years ago would have taken advantage of similar opportunities for premarital sex if they lived in the same conditions that exist now. What I can say is that historically, the data suggest that people have engaged in premarital sex at the same rate as they do now. Any other comparisons would be like trying to determine if George Brett was a better hitter than Ted Williams--too much has changed in the game to make an accurate comparison, but without question, they were both great.

      (2) That is quite possible. But I'm still not sure if that speaks to an opportunity discrepancy or if it speaks to a true difference in attitudes.

      I would posit that promiscuity would be somewhat comparable in urban areas. Maybe less so in rural areas. But I think that gets back to the opportunity discrepancy. I don't hesitate to speculate that if cell phones existed in the 1800s, sexting would be as popular then as it is now.

  3. I just wanted to comment on the "babies born 8 1/2 month after marriage comment". Your statistic of 8 1/2 months is more accurate than 9 months for a honeymoon baby. The beginning of 9 months begins with a woman's ovulation cycle, therefore conception takes place at the 2 week pregnant mark. From conception to birth is more like 8 1/2 months than it is 9 months. We learned this the hard way with our first daughter. We tried to plan a 9 month pregnancy with the last day of graduate school in mind. We were surprised to find out we were due two weeks before my last day of school. We thought we were safe. Our doctor explained the 8 1/2 month rule to us. Low and behold, she was actually born a week earlier than her due date, which means she was born three weeks early from our "safe date!" This is why doctors say to plan two weeks early and two weeks late from a due date, because each woman and each baby is different. Despite this, a honeymoon baby is actually born 8 1/2 months after a wedding date rather than 9 months after a wedding date. This of coarse depends on whether or not the woman in question is fertile the night of her wedding (14 days after the beginning of her ovulation cycle). Your statistic of 8 1/2 months is incredibly accurate for a honeymoon baby however, your assumption regarding that statistic is flawed.

  4. From webmed: You are considered pregnant before week 2! Conception is week 2.


    1. There are a couple of ways you can look at the question you bring up. The following link is the 2010 CDC report on birth rates. On page 4, using the singleton data, you can add up a rough estimate of what the < 38 weeks (probably a generous estimate of the 8 1/2 month rule). Over the six years listed, it works out to an average of 36.7% of births occur at less than 38 weeks (I assumed half of the 37-38 week deliveries were less than 38 weeks).

      That is pretty close to 40% which would seem to support what you're saying, except this is the percentage of pregnancies, regardless of when they were conceived. For the 1800 estimate I cited to be consistent, it would need to be that 40% of women who conceived on or near their wedding day delivered at less than 8 1/2 months...but the statistic says that 40% of women delivered in less than 8 1/2 months.

      If you take two weeks off of the medical description of full term (39 weeks), then the expected percentage of deliveries drops to 23%.

      The other way you can look at it is that if a woman chooses a random day to have sex, she has about a 25% chance of getting pregnant (she's fertile for about 7 days of a 28 day cycle). So, we might expect 25% of women to get pregnant on their wedding day. 40% of these women would be expected to deliver earlier than 8 1/2 months, or a total of 10% of all women.

      If you take out a week for her period, her probability of getting pregnant increases to 33%, in which case you'd expect 13.2% of all women do deliver earlier than 8 1/2 months.

  5. My brother worked phones for Ancestry for a while. So many angry callers demanding 'correction' of data showing that their ancestor was born less than 9 months after the marriage. So many of these people willing to accept the line that 'later pregnancies are always 9 months, but first babies are a little bit special'. I don't claim to know what proportion of their accepting that is ignorance vs. denial vs. giving up on the phone monkey, but I suspect all three play a role.