Eligible whites, blacks more likely to vote than Hispanics, Asians

Eligible whites, blacks more likely to vote than Hispanics, Asians

The Pew Research Center recently put out a report entitled “Dissecting the 2008 Electorate: Most Diverse in US History”. This translates as the white percentage of total voters being the lowest it ever has been in a Presidential election. Recognizing the slow but steady relative reduction in the size of the white vote is nothing novel. It has been occuring for several decades, to the enormous benefit of the Democratic party.

What I found interesting was the level of voter participation relative to eligibility, by race. The following graph shows actual voters as a percentage of eligible voters, by race, on a relative scale for the population as a whole. That is, if the graph included a line for the entire electorate, it would be a perfectly horizontal with a y-value of 100. If in a given election cycle whites represent 80% of actual voters and 75% of eligible voters, they have a civic score of 107 ((80/75)*100). Whites are represented by the white line, blacks by the black line, Hispanics by the brown line, and Asians by the yellow line:

With the exception of the steady increase in civic-mindedness among blacks over the last couple of decades, the relative values are pretty steady. Hispanic and Asian electoral participation has consistently remained below that of blacks and whites for as long as reliable demographic data has been kept. This helps in understanding why, at over 15% of the US population, Hispanics make up less than half that proportion of actual voters (with age structure and illegal residency status completing the picture) and also why anything that facilitates casting a ballot will tend to be supported by Democrats and opposed by (non-politically suicidal) Republicans.

A few other noteworthy facts gleaned from the report:

- The percentage of eligible white males who voted in 2008 dropped to 64.2%, down from 65.9% in 2004. That 1.7 point decrease was more than three times larger than the decrease among white females, from 68.4% in 2004 to 67.9% in 2008. White men, the GOP’s stalwarts, weren’t inspired by what was offered them last November.

- Eligible black women were the most likely group to vote in 2008, and also the most inclined to vote for Obama.

- The male:female participation rate ratio by race follows the common ‘hierarchical’ pattern found in so many other measures of social variables. Male participation among eligible voters as a percentage of female participation among eligible voters, by race, averaged for the 2004 and 2008 elections:

Asians — 95.1%
Whites — 93.5%
Hispanics — 91.6%
Blacks — 88.2%

Fascinating how this ‘ranking’ manifests itself in so many ways.

Red region, blue region, smart region, dull region

One of the major themes discussed by Andrew Gelman in Red State, Blue State, Rich State, Poor State is how affluence tracks more strongly with the tendency to vote Republican in poorer states than it does in wealthy states. Since blue states are on the whole wealthier than red states are, this translates into a larger income gap between Republicans and Democrats in red states than in blue states. As affluence correlates with intelligence, the IQ gap presumably follows the same pattern.

In states like Massachusetts, liberal whiterpeople who vote Democratic and whose friends all vote Democratic think of Republicans as simpletons of a lower social class than themselves. And in the Bay State, that’s understandable. Gelman points out that this is a problem when those localized assumptions are projected across the entire country, because in other regions of the country, they simply don’t hold. In the South, the monied classes are overwhelmingly Republican.

There are no data tracking IQ and voting patterns at the state level. The GSS, however, does break respondents up into the nine broad geographic regions used by the US Census.

First, I should point out that during the Bush years, in aggregate Republicans are smarter than Democrats are (with IQ scores converted from Wordsum results of 100.5 and 98.6, respectively–independents are lower than either, at 95.3). This holds in every region of the country except for New England, where Democrats average 4.9 IQ points higher than Republicans do*.

But in the interracial status games whites play against one another, minorities are often invisible. Considering the same period of time, white Democrats are slightly more intelligent than white Republicans are (102.1 to 101.7, with independents at 97.6). A noisome Republican can use this to force a whiterperson into a politically incorrect corner where he has to choose between conceding that Republicans are smarter than Democrats are or admitting that the Democratic average is dragged down by its one-third non-white contingent.

Among whites alone, Republicans and Democrats differ in red and blue states. Following is a table showing Obama’s advantage among white voters and the IQ advantage (converted from Wordsum scores with the assumption the white mean is equivalent to an IQ of 100 with a standard deviation of 15) of self-identified white Democrats by region**. Only responses from 2000-2008 are included to ensure that what we’re looking at is contemporary:

Region	Obama vote+ (%)	White Dem IQ+
New England	14.3	7.4
Pacific	10.0	1.1
Middle Atlantic	1.0	(0.3)
East North Central	0.2	0.2
West North Central	(6.4)	1.2
Mountain	(14.8)	(0.9)
South Atlantic	(24.0)	(0.6)
East South Central	(47.0)	(1.4)
West South Central	(49.3)	(2.6)

By region, white Republicans in red states are a little more intelligent than white Democrats are. In blue states, it is the reverse, with white Democrats being smarter than white Republicans. The correlation between Obama’s level of white support and the average white Democrat’s IQ advantage over the average white Republican’s is .72 (p=.03) at the regional level. This provides some explanation for why people in the Northeast may see the GOP as the party of rustic dummies while those in the South see it as the party of merit and prosperity.

Bundling states into nine broad geographic regions isn’t optimal, because red and blue states are lumped together in varying degrees, especially the East and West North Central areas. That such a pattern is nonetheless apparent suggests that at the state level it is even stronger. This isn’t surprising. After all, it’s implied in Gelman’s book, as IQ and earning power are correlated at the group level.

Data are available here.

GSS variables used: WORDSUM, YEAR(2000-2008), RACECEN1(1), PARTYID(0-2)(3)(4-6), REGION

* The Democratic IQ (dis)advantage by region for all races:

Region	Dem IQ+
New England	4.9
West North Central	(0.8)
Mountain	(1.3)
East North Central	(1.7)
Pacific	(1.8)
East South Central	(2.1)
South Atlantic	(3.1)
West South Central	(4.1)
Middle Atlantic	(4.2)

The only region that jumps out in surprise is the Middle Atlantic, which includes New York, New Jersey, and Pennsylvania. My guess is that many people involved in the financial industry from New York and New Jersey are Republicans, buoying the average. In the other direction, those three states are all heavily unionized, translating into lots of working class Democrats.

** To obtain white votes by region and compare them, I used exit polling data by state to obtain white percentages of each state’s total vote and multiplied them together. I then added up the total white votes for each candidate for all the states in a particular region to arrive at that region’s vote distribution for the 2008 Presidential election.

Advertisement