As Hillary and Obama went at it, the details of their respective healthcare plans remained obscure, as Pat Buchanan humorously pointed out. The Democratic contests underscored the wisdom of Singapore founder Lee Kwan Yew:
In multiracial societies, you don't vote in accordance with your economic interests and social interests, you vote in accordance with race and religion.States where exit polling data were gathered for caucuses is not included among the polled state numbers because the number of total voters is unknown. They're included in the non-polled numbers instead. When estimating voter totals for the two Democratic candidates, I looked to demographically and politically similar states for guidance but this time party totals are what I'm after, so the exit polls aren't as much help.
Because the Republican nominating process essentially ended on March 4, exit polling data for voters on the GOP side only extends up to that date. The exit polls cover 74.5% of the US' total population. To adjust for the quarter of the population not represented for whites, I took the national white percentage and divided it by the total white percentage (adjusted for population size of course) of the population not covered in the exit polling and multiplied the result by the white percentage of total votes from the population covered in the exit polling. The same process was then repeated for the other racial groups. I then adjusted proportionally to get the percentages of all five groups to equal 100% (I've never liked the "Percentages may not add up to 100 due to rounding").
The white percentage in states not covered in the exit polling represents 80% of the non-polled states' total population, compared to just 62% in the states polled. Recall most of the states that held caucuses are sparsely populated white flyover states with the glaring exceptions of DC and Hawaii (and to a lesser extent Nevada and Colorado). But because even though whites made up less than two-thirds of the population in polled states they still comprised 89% of Republican voters, the adjustment effect only nudges the GOP's white total up a couple of percentage points. It correspondingly drops the Hispanic percentage by a little over one percent (Hispanics made up 17% of the population in polled states compared to only 8.3% in non-polled states).
The same method was applied on the Democratic side. The effect of the adjustments are even more marginal here, as only 11% of the population was not covered in exit polling due to caucuses instead of primaries being held. This ups the white percentage by 1.5 points by trimming the black percentage by a point and the Hispanic percentage by half a point.
Here are the percentage totals by gender and by race with the aforementioned adjustments included (with pre-adjusted totals in parentheses) for states where exit polling was conducted and individual votes were tallied (click on graphics for a better image):
In Freedonomics, John Lott shows that female suffrage pushed US politics to the left starting in the early 1900s. A century later, that, in addition to the Fifteenth Amendment, continues to push the country leftward.
I wonder if Bill Clinton is still exulting over the minoritization of whites in the US. You might say the chickens have come home to roost. He should be more careful what he wishes for in the future.
In what year will that become the case on the Democratic side? Doing some quick estimating, my guess is sometime around 2030, but that's projecting a long way out. At that point the rubicon will have been crossed. As the Democratic party continues to represent West Virginians less and less, will there be an exodus to the Republican side, essentially creating two political parties--the white party and the non-white party--in the US? That is already the case to some extent, of course.
A few other interesting tidbits:
- Oklahoma and Utah are the only two states where the gender trend was bucked. In both cases, women comprised a majority of Republican voters. Pulling that off requires being so conservative as to move the female distribution far enough to the right that a big chunk of the left side of the distribution is still in Republican territory--enough to overcome the smaller male distribution that is inevitably to the right of female distribution in all states, irrespective of political environment. In no state did men make up a majority on the Democratic side.
- In Utah, Mitt Romney received 255,218 votes. The combined total for all other candidates from both parties comes to 146,705. This was one of the two only instances where one candidate garnered more votes than all the other candidates from both sides combined while the Republican race was still competitive. The other occurence was Washington DC, where Obama dominated with 85,534 to all the other candidates' pooled total of 32,984. Tribalism anyone?