Listening to an EconTalk podcast with David Brady of Stanford University, the phrase "all-important independents" was used on multiple occasions by host Russ Roberts' guest. I've heard both that American politics are becoming more partisan and that independents are coming to represent an increasing portion of the electorate (Brady emphasizes the latter, though President Obama's approval ratings by party provide evidence for the former--and he is continuing a trend that began as far back as the Johnson administration). Those two things would seem to be contradictory on their faces, though in Red State, Blue State, Andrew Gelman notes that while voters seem to be moving away from party labels and toward self-described political independence, the primary process helps ensure that politicians are generally to the right (if Republican) or the left (if Democrat) of their constituencies. Further, whether those in question represent a random sample of all adults, registered voters, or likely voters matters.
Despite opening with it, though, that's a digression from what I want to look at in this post. It sparked a tangentially related question of how partisan self-identification has changed over time among people grouped by varying levels of intelligence. The GSS wordsum distribution is ideal for this sort of categorization, as slicing score ranges into five groupings creates something akin to a normal distribution; Really Smarts (wordsum score of 9-10, comprising 13% of the population), Pretty Smarts (7-8, 26%), Normals (6, 22%), Pretty Dumbs (4-5, 27%), and Really Dumbs (0-3, 12%). In the following graphs, the blue lines represent those who self-identify as Democrats, the purple lines independents, and the red lines Republicans.
Society's upper echelon has been pretty evenly split three ways for the last several decades, and as of 2008 the partisan distribution among them was almost identical to what it had been in 1974. This might seem to counter the concern among some on the right that Republicans are losing ground among the most intelligent, but for three of the other four IQ groups, Republican self-identification has increased modestly over time. Relative to the bulk of the population, high-IQ types seem less enticed to move toward the Republican label than others are.
Among this solidly middle class portion of the population, the story is similar to that of the Really Smarts, albeit a little better for the GOP. The spike in the late-eighties looks like randomness to me, but I could be missing something. If I had to designate one of the five groups to which President Reagan would have appealed most to, I'd choose this one.
It is among Normals and Pretty Dumbs where the Democrats have lost the most ground. I suspect this is in large part due to the shift among southern whites towards the GOP and away from the Democratic party in the eighties.
It has been noted previously that when the perennial narrative describing how independents will determine the outcome of whatever election is at hand is given, the spotlighted fence-sitters tend to be duller than their more decisive compatriots are. That is apparent here.
Also apparent are the underpinnings of the PC trap white conservatives are able to snare SWPLs with--in aggregate, Republicans are, on average, more intelligent than Democrats are. The primary reason for this is the overwhelming support NAMs (especially blacks) give to the Democratic party. SWPLs could conceivably respond that among whites (the only group they're really concerned about how they measure up to) this is not the case. But of course they refrain from doing so, because making such a point would be racist. Appearing to be less sharp than white conservatives, while surely frustrating, is not as awful as being racist. Indeed, nothing is.
GSS variables used: WORDSUM(0-3)(4-5)(6)(7-8)(9-10), YEAR, PARTYID(0-1)(2-4)(5-6)