Probably the best way to honestly estimate whose voters are smarter is to look at two exit poll demographic measures of the national House race that are known to positively correlate with intelligence: income and education.
Not surprisingly, the Democrats did best among poorer voters, winning 68% of those making under $15,000 annually, while the GOP won 51% of those making over $100,000.
I would roughly estimate that the median GOP voter had a household income of $85,000 versus $73,000 for the typical Democratic voter. (Midterm elections, with their low turnouts, tend to draw a fairly elite group to the polls, as compared to Presidential election, which bring out more of a mass electorate. That's why voters' incomes are quite high for 2006. Or, maybe, respondents are just exaggerating their incomes …)
What about education, as expressed in years of schooling? The Democrats did best among the 3% who are high school dropouts (winning 64%) and the 18% claiming to have undertaken postgraduate studies (59%), while the Republicans were strongest among those who attended college without graduating (47%) and college graduates (49%).
I estimate that voters in House races who voted Democratic averaged (or at least claimed to average) 14.72 years of schooling (i.e., 2.72 years of college), while GOP voters averaged … 14.73 years.
In other words, they came out, for all intents and purposes, exactly the same on years of schooling.
That's the same result as in 2000, when the two parties' voters also were tied on education. In 2002, the GOP had a small advantage on years of schooling, and in 2004 the Democrats had a small superiority.
So, the two sets of voters are almost identical on education. This year, there just were more Democratic than Republican voters. And, bottom line, that's what counts.