Matt Yglesias has posted some charts showing that
1) Childlessness among women is becoming more common
2) The variation of this state by education is disappearing
Here’s the chart which illustrates the second phenomenon:
I think the reason this may be occurring is a dilution of the sample bias of women who have higher education in relation to the general ppoulation. In other words, as more women attain advanced degrees the pool of those women become less atypical vis-a-vis the general population
To gauge the shift in education and peculiarity I only needed a few variables in the General Social Survey. I limited SEX to women, YEAR to 1992-1994 and 2006-2008, DEGREE allowed me to break down educational attainment, and finally GOD was a variable which probed them on a culturally indicative variable.
First you can see women as a whole have become more well educated. This is a well known dynamic. The absolute change in the proportion of women who have advanced degrees is small, only a few percent, but in the GSS the proportion increase is around 50%. This includes masters and doctorates into one category.
The sample sizes for GOD across the periods of interest are small, but look at the enormous increase in the proportion who have no doubts in the existence of God. There was no change in this result in the general population across this time period.
UPDATE: For the second chart I forgot to note that that’s only women with advanced degrees.
Razib Khan’s degrees are in biochemistry and biology. He has blogged about genetics since 2002, previously worked in software development, is an Unz Foundation Junior Fellow and lives in the western US. He loves habaneros.
June 27th, 2010 at 10:31 am
Hm, it’s puzzling that the % of atheists/agnostics have declined, seeing that all the other surveys say that it has increased over the same time period.
June 27th, 2010 at 10:33 am
In the Beliefs About God, that is across all women, of all educational attainment?
So if the overall numbers haven’t changed, fewer men have absolute confidence that God exists?
June 27th, 2010 at 10:59 am
sorry, that’s ONLY women with advanced degrees. i’ll update.
June 27th, 2010 at 11:03 am
[...] This post was mentioned on Twitter by Ron Simon and Kelly Smith, Jeff C. Coleman. Jeff C. Coleman said: Why educated women are having more children | Gene Expression http://bit.ly/cNmtkR [...]
June 27th, 2010 at 11:56 am
InquilineKea, it’s both puzzling and disturbing how agnosticism/atheism has declined if these statistics are correct. I wonder if this correlates with the shift away from optimism toward cynicism that has been plaguing us lately.
June 27th, 2010 at 12:09 pm
re: agnostic/atheism
1) sample size for that is TINY
2) this is only for women who have more than a bachelor’s degree
IOW, don’t put too much stock in that segment. overall there is been a broader moderate shift toward atheism/agnosticism.
June 27th, 2010 at 12:13 pm
I don’t see how the dilution effect explains more than about one-third of the drop in childlessness among women with advanced degrees. Obviously, if a some women “move” from from the BA category to the MA/PhD category (it’s not the same women, of course), you’d expect average characteristics among MA/PhD to become more like the BA average. But to bring childlessness down from one-third above the BA average (’92-’94)to exactly the same as the BA average (in ‘o6-’08) would require, theoretically, an infinite dilution, not the one-third dilution you’re suggesting.
In other words, the characteristics of highly educated women have likely changed. It’s not an artefact of our view of the data. Why? Family-friendly policies in the higher professions? The passing of first generation of (family-conflicted) feminists? Stabilization of the sexual revolution?
June 27th, 2010 at 12:35 pm
also, women who had put off having children in the 90s finally started to have in the 2000s.
June 28th, 2010 at 1:20 am
Regarding atheist/agnostic fraction in the graduate sample: if the “new” women with graduate degrees are drawn from a segment of the population that is more religious than the “old” women with graduate degrees then the data is consistent with both a higher fraction of athiests/agnostics overall and a lower fraction with a graduate degree.
For instance, in ’94 about 8% of women with grad degrees are athiest/agnostic. If you assume that the additional women with grad degrees come from a population that is 5% athiest/agnostic and that the new women account for 1/3rd of the total (which is consistent with a 50% increase) then you get 0.08*66% + 0.05*33% = 0.07, which to my eye looks about right in the chart above. If the general population increased from 4% to 5% athiest/agnostic between ’94 and ’08 then it would be an overal increase in that category and still look like a decrease amongst advanced degree holders.
June 28th, 2010 at 4:55 pm
I’d expect that this childless rate is at peak or will peak soon given the reversal in fertility trends.
It occurs to me that your second graph makes sense if low fertility is due to a specific point on human development/wealth. Since education is roughly correlated with income, one could see this as a proxy for the general trend of the last thirty years for the US and some European countries. The lower socioeconomic groups are now passing into the levels of economic wealth associated with low fertility, while those with more education are passing out of the trough and are concomitantly having more children.