My Affordable Family Formation theory isn't about who wins nationally, it's about how, given a particular national level of support, which states will be solid blue (Democrat), which ones purple (mixed), and which ones solid red (Republican).
Of course, George W. Bush ran in both 2000 and 2004, so maybe he was the reason my theory worked so well in both elections. Thus, 2008, with its quite different candidates, was a good test. Or maybe the Housing Bubble and its subsequent popping would have changed results dramatically.
Before getting to the results, let me review my AFF theory. It holds that what paints the electoral map red and blue is "affordable family formation" was validated once again. Taking a quick and dirty look at McCain's and Obama's shares in each state (plus DC) with 92% of the national precincts reporting, the same two demographic variables that drove the results in 2000 and 2004 showed startlingly high correlations once again.
My basic theory is that Democrats do best in states with metropolitan areas where land for homes is scarce because they are hedged in by oceans or Great Lakes; while Republicans do best in inland areas where homebuyers can look around for homes in a 360 degree radius around job sites. I call this the Dirt Gap: Republicans are found more in areas with more dirt and less water.
This means that homes in inland areas tend to be cheaper because the supply of land within a certain commuting time is greater. In turn, cheaper homes mean that non-Hispanic whites tend to marry earlier and have more children, which means they attract family oriented people and their cultures tend to be more family-oriented, making Republican family values appeals more appealing there. In contrast, "Living by the Water," which is #51 on the Stuff White People Like website, correlates with Stuff White People Like political views. (You can read about Affordable Family Formation in detail with graphs here.)
Take a look at the Average Years Married between ages 18 and 44 among non-Hispanic white women in the 2000 Census. That's a statistic I invented to be the marital analog of the well-known total fertility rate measure (which estimates from the latest available year's birth behavior how many children a woman will have in her lifetime). Likewise, Average Years Married estimates how many years out of the 27 between 18 through 44 will a woman be married. The Average Years Married for non-Hispanic white women does a remarkably good job of predicting McCain's (or Obama's) share of the total vote across all races in the states.
Thus, McCain carried 19 of the top 20 states on Average Years Married among non-Hispanic whites, while Obama carried 18 of the 19 lowest states. The correlation coefficient was r=0.88, on a scale where social scientists usually call r=0.2 "low correlation," r=0.4 "moderate correlation," and r=0.6 "high correlation." So, in the social sciences, r=0.88 would have to be something like "extremely high correlation." This is, however, down from the astonishing 0.91 level seen in 2004, but, keep in mind, the demographic data I'm using is now 8.5 years old. (It was collected on April 1, 2000 for the last Census.)
Looking at the 2002 Total Fertility Rate among non-Hispanic Whites, Obama carried the bottom 15 states, while McCain carried 14 of the top 15. The correlation coefficient was r=0.82. The demographic data is now 6 years old. (In 2004, when the demographic data was fresher, it was 0.86.)
Keep in mind that this is based on incomplete 2008 voting results with 8% of the precincts and who knows how many of the mail-in ballots missing, so the correlations will likely change.
By the way, this explains much of the Sarah Palin Hysteria: with her five children, she elicits the SWPL whites' secret dread that they are being outbred by the non-SWPL whites.