Wednesday, March 15, 2006

One Baby, Two Baby, Red Baby, Blue Baby

Via Kevin Drum, I see that USA Today of all places has published another one of those silly claims about "Red states" out-breeding "Blue states" by one Phillip Longman. Here is the gist of it:

It's a pattern found throughout the world, and it augers a far more conservative future -— one in which patriarchy and other traditional values make a comeback, if only by default. Childlessness and small families are increasingly the norm today among progressive secularists. As a consequence, an increasing share of all children born into the world are descended from a share of the population whose conservative values have led them to raise large families. [...]

Today, fertility correlates strongly with a wide range of political, cultural and religious attitudes. In the USA, for example, 47% of people who attend church weekly say their ideal family size is three or more children. By contrast, 27% of those who seldom attend church want that many kids.

In Utah, where more than two-thirds of residents are members of The Church of Jesus Christ of Latter-day Saints, 92 children are born each year for every 1,000 women, the highest fertility rate in the nation. By contrast Vermont -— the first to embrace gay unions -— has the nation's lowest rate, producing 51 children per 1,000 women. [...]

This dynamic helps explain the gradual drift of American culture toward religious fundamentalism and social conservatism. Among states that voted for President Bush in 2004, the average fertility rate is more than 11% higher than the rate of states for Sen. John Kerry.
This is one of those bizarre claims I've seen floating about a lot lately, mostly by right-wingers who use it as some kind of triumphalist taunt and/or threat. You just wait! We're going to out-number you! I guess they figure that if they can't implement their culturally repressive schemes due to a lack of popularity, they'll just wait until enough like-minded dopes have been born into the world.

Unfortunately for them, Longman's thesis is wrong. In fact it's so bad, it's hard just to list all of the ways in which it fails. There is the fact that children don't automatically adopt the politics of their parents. There is the fact that political movements are historically contingent. There is the fact that most of our net population gain is due to immigration, and immigrants don't tend to vote Republican. And then there is the fact that Longman doesn't even bother to find any causal link between political belief and birth rate; at best he simply finds a correlation. Does it occur to him that having children makes people more conservative and more religious? If so, he doesn't bother saying.

Longman also doesn't bother to break-down fertility by demographic group. If he had, he would notice that white people in general (most of whom vote Republican) are out-bred slightly by blacks, and a whole lot by hispanics, both of whom tend to vote Democrat. On top of that, it also known that poorer people have higher fertility than rich people, and uneducated people have higher fertility than educated people, etc. So that raises the question: is the high fertility rate of Texas, for example, caused by Bush-loving oil-rich conservatives, or by poor hispanics who vote overwhelmingly Democratic? Almost certainly the latter. Indeed, when you look at the different states, those with high Hispanic populations (including true blue California) tend to have the high fertility rates. The New England states, with relatively few racial minorities and high standards of living, have the lowest rates. Utah, with its large Mormon population, is an outlier. The whole red-state/blue-state thing appears entirely incidental. Is there any correlation with political belief? Possibly, but while Longman says there's a strong correlation, he doesn't back it up with any actual evidence, much less quantify it. If fertility does correlate with political belief, I'll bet my future first-born that it pales in comparison to differences in ethnic fertility.

But what really makes the argument stupid is the underlying assumption that the relatively small differences fertility rates among American subpopulations, whether by ethnic group or by political belief, will a) remain constant, and b) add up to noticeable difference within our lifetimes, or even our grandchildren's lifetimes. They won't.

Below I made a chart for what happens over the next 250 years when you take Utah's high birth rate (21.2 per thousand in 2003) and population (2.2 million in 2000) and compare it going forward with the birth rate (15.2) and population (33.8 million) of California. I use the same data that Longman gets his fertility rates from, but I used the birth rate because it's easier than using the general fertility rate.

If it looks like Utah isn't about to out-breed the rest of us and steal all of the electoral votes, it's because they're not. In spite of Utah's abnormal fertility rate, the fact is that everywhere in the US has a low fertility rate, and given this, it takes a long time for noticeable differences to add up. The Mormons have been breeding like crazy for a good while now, and Utah still only has a measly 2.2 million people. Yet Longman expects us to believe that the rise in fundamentalist politics over the last 20 years was caused by differences in breeding? Nuts.

So when will the populations of Utah and California converge? In 464 years. And they will converge at, get this, 36 billion people. If that sounds unreasonably high, it's because this model ignores mortality, migration, and most importantly, changes in fertility rate.

About that changing fertility rate. Longman labors under the assumption that fertility rates we see today will remain as they are into the indefinite future. Nonsense. Fertility rates have dramatically fallen for everyone over the last four decades. The white general fertility rate has fallen from 113.2 in 1960 to 65.1 in 1999. Among African-Americans, it'’s fallen even more sharply over the same time period: from 153.5 to 70.1. Whereas the black fertility rate was once 36% higher than the white rate, it'’s now only 7% higher. The same thing will happen to Hispanics. How about those fecund Mormons? Their rates have plummeted too. Total fertility went from 4.3 in 1960 down to 2.64 in the late 90s. It is expected to continue to fall. Indeed, almost every demographic group'’s fertility does the same thing everywhere: As its members become more and more economically prosperous, their fertility rates decline down to just below replacement levels. Please note that the Mormons have a religious imperative to have as many children as they can afford, and yet their fertility rate still shot to the ground. This is a demographic law that even God can't change.

So Utah and California aren't going to reach 36 billion people, and it's safe to say that Utah will never catch up to California's population, at least not by means of natural birth rate. The fact is, America's population boom is over. Most of our future growth will come from immigrants. If you want to find a population boom, try Africa, with a total fertility rate of 5.4. That makes the Mormons, with their puny 2.64, look downright childless. Fortunately, Africa's fertility rate is falling too, and as their economy improves, they'll eventually drop down to replacement levels like those of us in the developed world. But the fact remains, the world'’s main demographic trend is not an increasing number of babies being born to fundamentalist crazies who refuse birth control, it'’s an exploding population of Africans and Asians. America, by comparison, isn'’t growing at all.

So Longman can drop the nonsense about the country being overrun by religious nuts. Maybe he can go back to the standard conservative demogoguery about the impending take-over by the brown hordes. Of course that's a stupid argument too, for all the same reasons, but I like it better because at least it causes the 'wingers some serious heartburn.

Update: I thought I'd make a more direct evaluation based on national birth rate data rather than simply comparing Utah vs. California as per the graph above. Longman writes the following:

Among states that voted for President Bush in 2004, the average fertility rate is more than 11% higher than the rate of states for Sen. John Kerry.

Sounds like a big difference, right? Wrong.

Below I charted what would happen if you take a given birth rate and compare it to one that's 11% higher over time. Again, I'm using birth rate rather than fertility rate because it's easier, not because it makes any real difference. (The data are taken from here, which is Longman's source.) I'm assuming that the country starts off containing exactly 150 million blue staters and 150 million red staters. The blue state birth rate is set to the national average of 14.1 per thousand, whereas the red state birth rate is 11% higher at 15.65.

You can see that after 50 years, very little has changed. The overall poplulation has grown to 619 million, but the percentage of red staters has only increased from 50% to 51.9%. Whoopity doo! How long would it take to for the red state population to increase to 60%? About 270 years, give or take.

So again, even if we ignore the fact that the difference probably has nothing to do with politics, if we ignore immigration, if we ignore changing fertility rates, and if we believe against all evidence that people have the same politics as their ancestors did several generations ago, the numbers still don't support Longman's thesis. The tiny differences in fertility between red states and blue states would require a very long time to add up to anything meaningful at all. It is just that bad.