As so often happens the discussion in the comments section to a recent post turned to the issue of delayed marriage & motherhood*. In this specific case we were discussing how a woman’s circle of friends and acquaintances is impacted by her own decisions in life, and how this nonrandom group tends to provide an inaccurate sense of what is normal for women. The basic phenomenon happens for all of us. Our choices in life tend to have the effect of surrounding us with others who have made very similar choices. This is only natural, but it can leave one with a skewed perception of the larger world. An unmarried woman focused on her education or career in her mid 20s is very likely to be surrounded by women of the same demographic, and very much unaware of how many women of her cohort are already starting their own families. The media can reinforce this sense by reporting statistics out of context on how many women are delaying marriage and childbirth.
Sometimes simple statistics can give us a very inaccurate impression of what is going on around us. The most common statistics around motherhood are the mean or median age women are having their first child, as well as the percentage increase in the number of children born to older mothers. As I’ve pointed out in the past, the breathless headlines about women having children ever later in life obscure the larger picture. Births to women in their late 30s are up, but births to women 40 and over remain extremely rare:
What statistics often fail to tell us is how the shape of the curve changed. The following charts are from the OECD publication SF2.3: Mean age of mothers at first childbirth (H/T Lavazza)**. The OECD retains the copyright to the charts. When we hear about the age of childbirth being delayed, I think it is normal to unconsciously presume that this involves a shift in the curve of childbirth and not a reshaping of the curve. For some countries this is certainly true (live births per 1,000 women):
But for other countries the change is much more about a loss of births to women at a younger age with little or no increase in births to older women:
One thing which is common throughout all of the charts is that the most recent curve shows a peak in births no later than the early thirties (Ireland appears to be the max at 33).
The other problem with statistics on late life childbirth is it doesn’t tell us how many women waited too long and didn’t become mothers at all. I found a Pew Research Center report titled Childlessness Up Among All Women; Down Among Women with Advanced Degrees with some data from the US Census on this. One thing I found interesting is the definition of “childless”:
This report uses the standard measure of childlessness at the end of childbearing years, which is the share of women ages 40-44 who have not borne any children.
So much for 40 being the new 20!
Not surprisingly, the overall percent of women who are childless has increased over the last 30+ years:
Among all women ages 40-44, the proportion that has never given birth, 18% in 2008, has grown by 80% since 1976, when it was 10%.
But as I mentioned at the beginning of the post, the group one is part of can be very different than the average. Here is what it looks like broken down by level of education (following charts were made by me using their analysis/presentation of US Census Data):
Interestingly women with masters degrees and higher seem to have made adjustments towards having children recently. This makes me suspect that they learned from a percentage of their predecessors who inadvertently waited out the reproductive clock.
Even though out of wedlock births are much more common now than in the past, married women are still much more likely to have children than their unmarried counterparts, especially for women with some level of college or higher education:
The reluctance of college educated women in the US (and the UK) to have children out of wedlock could make for some interesting choices for the group of women currently in their 20s who are delaying marriage. This puts additional pressure on their need to find a beta provider in their 30s.
They also broke this out by race and education. Here is the breakdown for white women:
*I hesitated to do yet another post on this topic because discussions on this often turn personal. However, the data was simply too tempting. Please avoid bringing up another reader’s choices in the discussion on this unless they specifically bring them up first in response to this post. Even in cases where the reader brings up their own choices first, please avoid making this personal as much as possible.
** Update Aug 13 2013: The updated version of SF2.3 is located here (the old link automatically redirects though), and the xls version of the publication is available here.
