Independent Woman

This exchange between Steve Sailer and Brad DeLong led me to some data on the median age at first marriage in the United States which I used as the basis for my chart:

The story we see here is that in 1890 men waited until the relatively old age of 26.1 years before getting married -- perhaps representing the time at which they could acquire some land and support a family, and they married substantially younger women -- 22 years old on average. Over time, growing prosperity led the first marriage age for men to drop steadily and substantially to 22.8 in both the 1950 and 1960 snapshots. At the same time, women's age at first marriage declined steadily and slightly to 20.3, giving us the "traditional" (i.e., postwar) family.

At this point, the trends reverse and average age at first marriage rises steadily. By 1980, women are getting married at the 1890 age again, though men are still getting married younger than their 1890s counterparts. By 1990, however, 1890 marriage ages are back in style for men, and women are getting married later than their 1890s counterparts. By 2000, at 26.8 men are getting married slightly later than they did back in the way, but women at 25.1 years old are now getting married way than either midcentury or late 19th century women. From 2000-2003, ages continue to creep upwards for both genders, and after that I have no data.

Comments

Re: the age dip in the 1940's:

Wasn't there a bit of a marriage boom as couples married prior to deployments in WWII (perhaps marrying younger and sooner than they would have without the war), or I have I seen too many movies?

Posted by TheOaf | July 11, 2007 10:58 AM

More importantly, the Pope has implicitly declared that James Dobson, who famously declared that Fred Thompson is not a Christian, is also not a Christian since he isn't Catholic.

I'm waiting for round 2.

Posted by FS | July 11, 2007 11:06 AM

it's important - as historical demographers and family historians can expound upon at great length - to understand the context of the numbers matthew has here.

for example, until social security, in many families it was the responsibility of the children, and particularly the youngest child, to live at home and care for the aging parents until they died, whenever that was.

for example, in a period of much tighter social constraints on inter-marriage across ethnic and religious lines, availability of a suitable partner for marriage may have been quite limited.

for example, pace TheOaf, war and depression affected marriage patterns.

in short, there's much more to it than just an issue of increasing prosperity....

Posted by howard | July 11, 2007 11:09 AM

Step back and the graph takes an obvious dip right around the post-WWII period, when economic prosperity was widespread.

Seems to me that this correlates with the new book "Staying Afloat": if age of first marriage has an inverse correlation with economic opportunity, in that age of marriage goes down as economic opportunity goes up, then what we are seeing since WWII is marrying age going up as economic opps. go down.

Posted by nikkos | July 11, 2007 11:14 AM

my "economic opportunity" right out of college was great - a programmer in 1994? c'mon.

but after i graduated, i spent a couple of years enjoying my (relative) prosperity, and freedom: playing in bands, buying records, trying out a few jobs, hanging out with all my college friends who had also graduated into lucrative careers, etc..

but, i did get married at 26, in 96 - right on schedule.

Posted by cleek | July 11, 2007 11:20 AM

in other words...Brooks is (partially anyway) correct.

which to those of us who knew the stats...was obvious...despite the anecdotal remarks from people about getting married in their 30's in the 60's, 70's and 80's.

the 90's and the present are later than those decades...and considerably so with women.
as an unmarried Manhattan male at 32...yeah, the culture has changed.

Posted by Nathan | July 11, 2007 11:27 AM

In the last sentence of the posting, "ages continue to creep upwards for both genders" should be "for both sexes." "Gender" is a grammatical term. Employing it to classify real, live, flesh and blood people is a species of prudery.

Posted by renito | July 11, 2007 11:28 AM

I think this all has something to do with the gay marriage thing. What that is I don't know but I am having the words spinster aunts, confirmed bachelors, and images from It's a Wonderful Life bubbling through my mind.

Posted by Felipe | July 11, 2007 11:41 AM

In the dictionary that's sitting on my desk, the first definition of "gender" is indeed the grammatical one. The second definition is "the fact or condition of beign a male or a female human being ..."

Posted by pedants suck | July 11, 2007 11:42 AM

There's a confounding factor that may be significant in these numbers; I imagine that fifty years ago, it would have been extremely unconventional for an unmarried couple to be living together. The point at which the couple moved in together and started planning a family was right after marriage. For the past few decades that traditional pattern has been breaking down, with increasing numbers of people living in common-law marriages and other "less conventional" (though not any more) relationships. Would it be surprising if the age at which people start getting hooked up with the longer term in mind is about the same, but they just don't get a marriage license?

Posted by RSA | July 11, 2007 11:43 AM

RSA - "For the past few decades that traditional pattern has been breaking down, with increasing numbers of people living in common-law marriages and other "less conventional" (though not any more) relationships."

You also have to consider that the break-up of marriages by something other than death used to be much less common. People may not have gotten married earlier, but maybe there were far fewer non-married people in their 30s and 40s back then because divorce wasn't as common. The idea of people getting married later may be based more on the combination of the never-marrieds with the used-to-be-marrieds into a single social group.

Mike

Posted by MBunge | July 11, 2007 12:13 PM

"From 2000-2003, ages continue to creep upwards for both genders, and after that I have no data."
If my personal experience is a useful guide, I'd estimate that males' ages rose about four years in the 2003-07 period. A pure guess here, but perhaps females' ages rose about the same amount.

Posted by old guy | July 11, 2007 12:37 PM

I'm surprised at this data. In 1850 the average life-span was 35 years. By 1900, life-span had increased somewhat to around 41 years. I would have expected people to marry much younger, relatively speaking -- this is like waiting 'til your twilight years to start a family.

Oddly enough, as life-span increased, people started marrying earlier. I don't get it.

Posted by raff | July 11, 2007 12:49 PM

raff:

The biggest driver of increased life expectancy is lower infant/child mortality. People who made it to age 21 in 1850 didn't neccessarily live significantly shorter lives than people doing so today. Even the Bible waxes poetically about man's allotted "3 score years and 10."

nikkos:

That's daffy. The average age of marriage has been steadily going up since 1960 - through recessions and booms.

Posted by sd | July 11, 2007 12:59 PM

Another factor that might be affecting the pattern is education. In 1940, only 24.5 percent of adults had completed high school. The rate rose rapidly after that point--34 percent in 1950, 41 percent in 1960, 55 percent in 1970, and 69 percent in 1980. If people are staying in school longer, they start their careers later, and wait until later to marry.

Posted by Bob | July 11, 2007 1:08 PM

RSA...extensive cohabitation appears to primarily occur among people who don't get married at all (i.e. lower economic strata)...so that probably doesn't change the average marriage age that much.

Posted by Nathan | July 11, 2007 1:12 PM

I'd be interested in seeing the ethnic, geographic and economic distribution of the median ages of first marriage across that timespan. My intuition suggests to me that the net data for the entire United States is not useful, as its a large country that encompasses a lot of ethnic and cultural groups.

Posted by Patrick | July 11, 2007 1:49 PM

Steve Sailer is like a turd in a swimming pool-- if he's around, I ain't even dippin' a toe in. Ditch the bigot, Matt, you're a big boy now.

Posted by Le Duc Violet | July 11, 2007 2:23 PM

I don't think these numbers take into account the group of "never married". If people in their thirties aren't marrying--and a significant percentage of them aren't--then the statistics above won't fully reflect changing patterns unless and until they do marry.

Posted by Thomas | July 11, 2007 2:42 PM

Response to Raff:
Average life expectancy is just that. Infant mortality rates were much much higher in the 1800s. So there were a lot of children dying who were pulling down the average. If you made it into adolescence, you could probably expect to live until your late 50s, early 60s.

Mortality in the cities was always higher in the country for adults and children. Greater exposure to untreated water, and the poor sanitary conditions exacerbated these problems as more people moved off the farm into urban environments.

Looking at the age of death for most of my ancestors (I've got data on 450+)-- and for their immediate families -- I find that the my Puritan ancestors (and immediate family) were dying at an average age of 76 years for men, and 79 years for women. Mind you this is for adults. The average age of death must have been much much lower, since approx 15 percent of children seemed to die before age 5. The mortality rates were probably higher, because not all the infant deaths got into the records. But in rural New England in the late 1600s, if you got past the first 5 years, you were likely to live to a ripe old age.

On the other hand, by the 1840s, when many of my ancestors started moving into medium-sized towns and cities, I see women dying in childbirth much more frequently, dropping the average adult age of death down into the mid-40s for women, and the mid 50s for men (I've got a fewer number of ancestors and immediate family to work with in those more recent generations, so I'm a little less sure of the statistical validity of the data). And infant mortality went way up, including more older children dying before adulthood. I can only attribute this to poor sanitation (and possibly the temperance movement, which unwittingly encouraged people to drink contaminated water rather than bacteria-free beer ;-)

So aveage age of mortality is just an average. It doesn't really describe your life expectancy after you attain a certain age.

--Beo

Posted by beowulf888 | July 11, 2007 4:37 PM

Beo,
My own family tree is rather different. Apart from an ancestor who died fighting in the Civil war, and his widow who followed him just a year later, and a badly abused grear-grandmother dying in childbirth with her 10th pregnancy, most of my ancestors made it to what would be a respectable age even today. However, the childhood death rates were awful: almost 1/4 of the kids in my family tree pre-1900 did not make it to 20. I should mention that most of my ancestors lived in small cities (e.g., Kalamazoo) or in rural villages so they probably escaped the worst of big city pollution and crowd plagues.

Posted by JonF | July 11, 2007 7:08 PM

The rule was "you can't marry my daughter until you have a farm (or other evidence of economic independence and success) of your own,"

It's the modern, independent twenty-something woman that is the truly new sociological phenomenon...

And the result has been a dramatic increase in the percentage of children born to un-wed mothers over the last 40 years. And not teen mothers either, but children born to modern, independent twenty-something women with access to birth control. Today 37% of all babies are born out of wedlock, the highest percentage in the industrialized world, while America's teen birth rate is it's lowest on record. In 1960, before the pill and abortion, 5% of all births were out of wedlock.

As a parent with two small children, I know how hard raising children is. It's not something I would recommend that anyone try to do alone if they can help it. It takes a lot of money, time, and effort to raise kids and thus the "old rule" still carries much wisdom. Prudent parents are waiting longer to have kids. Any trend which decreases the commitment, money, time, or effort spent on raising children is one that augurs poorly for the quality of our culture.

I don't know what the answer is, I'm certainly not for returning women to chattel, but I don't think we can ignore the consequences of the new sociological phenomenon of Girls Gone Wild.

Posted by Just Karl | July 11, 2007 8:33 PM

Re: 37% of all babies are born out of wedlock

Aren't the percentages higher in Scandianavia where many couples do not wed until their second child is born? Of course we may be confusing two things here: single mothers and unwed mothers. The two are not synonymous. In Europe, and I suspect generally among the unwed middle class mothers in the US, the baby's father is living with the mother in a long term relationship which they may intend to formalize with marriage someday. It's generally only among the poor (where there is a lack of "marriageable", i.e., income producing, men) and among the rich (where money and servants abound) that true single motherhood is found.

Posted by JonF | July 11, 2007 9:56 PM

http://ise.uvic.ca/Library/SLT/society/marriage.html

Posted by misc | July 11, 2007 10:07 PM