Monday, September 06, 2004

Breeding for Conservatism?

Jason Soon links to an interesting article in the Washington Post suggesting that religion-minded conservatives are outbreeding liberals by a mile. The kicker is the concluding passage excerpted below:

If Gore's America (and presumably John Kerry's) is reproducing at a slower pace than Bush's America, what does this imply for the future? Well, as the comedian Dick Cavett remarked, "If your parents never had children, chances are you won't either." When secular-minded Americans decide to have few if any children, they unwittingly give a strong evolutionary advantage to the other side of the culture divide. Sure, some children who grow up in fundamentalist families will become secularists, and vice versa. But most people, particularly if they have children, wind up with pretty much the same religious and political orientations as their parents. If "Metros" don't start having more children, America's future is "Retro."
This, to my mind, is clearly nonsensical based on the empirical evidence we have to hand. It's safe to say that in the West, the highly religious have been outbreeding their wealthier and more liberal counterparts in places like the US and Britain for at least a century now*, and arguably a century and a half in the British case, and yet the trend in both countries has been towards liberalism rather than away from it. Were the thesis outlined here correct, we ought to have seen a monotonic increase in social conservatism throughout the 20th century, and that this is not the case illustrates why it is always helpful with arguments like this one to state one's assumptions in an explicit mathematical form, as the devil is always in the details. Saying "most people, particularly if they have children, wind up with pretty much the same religious and political orientations as their parents" is a uselessly vague prop for the conclusion the article's author seeks to draw from it.

As an aside, one thing mentioned in the WaPo article struck me as both surprising and doubtful. In particular,
African Americans, according to the National Center for Health Statistics, now have a lower average fertility rate than whites, and they are no longer producing enough children to replace their population.
Where can such numbers be found? Looking here, the claim seems untrue, at least as of 1999, while this document, which has figures for 2002, also contradicts the claim made above, giving a general fertility rate (GFR) of 57.4 per 1,000 for non-Hispanic white women, 67.4 per 1,000 for non-Hispanic black women, 58.0 per 1,000 for Native American women, 64.1 per 1,000 for Asian & Pacific Islanders, and 94.4 for Hispanic** women. The closest that the latter document comes to supporting the dubious claim made above is in suggesting that black and white fertility rates are converging, as non-Hispanic white women recorded a decline in GFR of 8-9% over the 1990-2002 period, while non-Hispanic black women recorded a 24 % decline over the same period. I don't see why anyone should trust in this guy's reasoning when even his data seems of dubious provenance.

*Few people seem to remember for instance that the 1920s and 1930s also witnessed a low in Western fertility rates, and this was only reversed after World War 2. Birth control did not begin in the 1960s with the pill.

**"Reconquista" paranoiacs would do well to note that while GFR declined 14% for Mexican women and 21% for Puerto Rican women over the period in question, the one Hispanic subgroup for which it rose was for the Cubans (i.e. predominantly GOP-voting whites), and that by a hefty 12%.