Delayed Childbearing and Voting Behavior
Tuesday, November 17, 2009
The correlation between religious and moral values and voting behavior did not operate a generation ago.
It is commonly understood, I think, that American women are having their first babies later in life than they used to. Over a long generation, the change has been even more striking than I would have guessed. According to the National Center of Health Statistics at the Centers for Disease Control and Prevention, the average age of mothers at first birth in 1970 was 21.4 and in 2006 was 25.0. Behind those numbers lie substantial changes in behavior in those born circa 1949 and those born circa 1981. The average American woman with children in 1970 became pregnant before she was old enough to vote (the law lowering the voting age to 18 was not effective nationally until 1972). The average American woman with children in 2006 became pregnant more than six years and six months after she became old enough to vote (at least, if she was born in 1954 or later).
What’s also fascinating is that there is a very high correlation between women’s age at first childbirth and voting behavior. In 2004, every state where the age of first childbirth was below the national average voted for George W. Bush. And all but three states where the age of first childbirth was above the national average voted for John Kerry. The exceptions were Colorado, Florida, and Virginia, all of which voted for Barack Obama in 2008.
Yet the correlation seems to be slightly lower for the 2008 election. Of the 28 states where the age of first childbirth was under the national average, 22 states with 174 electoral votes voted for John McCain and six states with 63 electoral votes for Obama. On the other hand, all 22 states (and the District of Columbia) where the age of first childbirth was at or above the national average voted for Barack Obama.
The average age of mothers at first birth in 1970 was 21.4 and in 2006 was 25.0.
The conclusion seems fairly obvious to me. Voting behavior in recent years, particularly in 2004 and somewhat less so in 2008, has been highly correlated with religion, or degree of religiosity. And early childbearing seems correlate highly with conservative or tradition-minded beliefs—or perhaps it might be more accurate to say that late childbearing, delayed well beyond the time that has been normal for many generations, highly correlates with liberal beliefs. For a similar theory, developed at considerably greater length, see blogger Steve Sailer’s interesting writing on affordable family formation.
To oversimply a bit, Sailer argues that high real-estate prices and heavy immigration in coastal states has led tradition-oriented people with large families to move elsewhere, with predictable political results. This roughly lines up with Bill Bishop’s argument in The Big Sort: Why the Clustering of Like-Minded America is Tearing Us Apart, that Americans have been clustering in culturally congenial communities, with liberals shunning conservative areas and conservatives shunning liberal areas.
In 2004 every state where the age of first childbirth was below the national average voted for George W. Bush.
This argument is undercut somewhat by the National Center of Health Statistics’s data on age at first childbirth in 1970. When you map the results, you find that almost all the states that had below average ages of first childbirth in 2006 did so in 1970, and vice versa. Only a few states with average or above average ages of first childbirth in 1970 had lower than average ages in 2006 (Alaska, North Dakota, Utah), and only a few states with below average ages of first childbirth in 1970 had average or higher than average ages in 2006 (District of Columbia, Florida, Maine, Michigan). The first group of states does seem to have become culturally more conservative in this period, while the second group of states seems to have become culturally more liberal.
But another factor seems to buttress Sailer and Bishop’s arguments. The gap between the states on age at first childbirth was much narrower in 1970 than in 2006. In 1970 the age of first childbirth varied between 20.2 (Arkansas) and 22.5 (Connecticut, Massachusetts, NewYork)—just 2.3 years. In 2006, the age of first childbirth varied between 22.6 (Mississippi) and 27.7 (Massachusetts)—5.1 years. The cultural differences between the states seem to have widened substantially over these 36 years.
The gap between the states on age at first childbirth was much narrower in 1970 than in 2006.
In addition, there has apparently been little correlation between age of first childbirth in 1970 and voting behavior in the close elections around that year, 1976 and 1968. In 1976 states where the age of first childbirth was below average cast 89 electoral votes for Gerald Ford and 173 electoral votes for Jimmy Carter. States where the age of first childbirth were average or above average cast 151 electoral votes for Ford and 124 electoral votes for Carter. In 1968 states where the age of first childbirth was below average cast 157 electoral votes for Richard Nixon, 60 electoral votes for Hubert Humphrey, and 46 electoral votes for George Wallace. States where the average of first childbirth was average or above average cast 144 electoral votes for Nixon and 131 electoral votes for Humphrey. Which is another way of saying that the correlation between religious and moral values and voting behavior which has become familiar with in this decade did not operate a long generation ago.
Michael Barone is a resident fellow at the American Enterprise Institute.
FURTHER READING: Barone recently discussed the nation's decline in foreign-born population in "An Immigration Tipping Point,” and described how Democrats need their base much more than Republicans in "Republicans and Democrats: A Tale of Two Bases."
Image by Darren Wamboldt/Bergman Group.