Steve Sailer, touching on the tendency for presidential elections to bring out marginal voters who don’t participate in mid-term elections:

The big difference between 2012 and 2014 is that Presidential elections, being big whoop-tee-doos, bring out the dumb and disorganized, the Julias, so Democratic rhetoric in 2012 was pitched at the lowest common denominator. Midterms bring out fewer but better voters, civic-minded citizens who are more likely to be annoyed than energized by the stupidity and nastiness of the Democrats’ “war on women” rhetoric.

That seems a plausible working assumption, though trying to quantify the electoral differences between mid-term and presidential election cycles has revealed it to be less obvious than I would have assumed it would be. Taking averages from the 2006 and 2010 mid-terms and averages from the 2008 and 2012 presidential elections, here are the percentages of voters by income:

Mid-term under $30k — 18%
Presidential under $30k — 19%
Mid-term over $100k — 25%
Presidential over $100k — 27%

By education:

Mid-term no college — 22%
Presidential no college — 24%
Mid-term college grad — 48%
Presidential college grad — 46%

And by marital status*:

Mid-term unmarried — 32%
Presidential unmarried — 37%
Mid-term married — 68%
Presidential married — 63%

The presumed tendency is only very modestly perceptible in the educational and marital categories and is non-existent when it comes to income.

This does not, however, necessarily reflect information or engagement levels of the electorates. It seems quite possible that across all income and educational levels, those who vote in presidential elections but forgo mid-terms tend to be less politically informed and engaged than those who consistently vote in all election cycles are.

* The 2010 mid-term exit polls didn’t query participants by marital status. Instead they were asked whether or not they identified as gay or lesbian. The Cathedral hadn’t yet decisively wrapped up its victory in the battle over same-sex marriage in 2010, and not public surveys are exempt from conscription when necessity dictates! Total war requires total commitment. Consequently, the mid-term figures are comprised exclusively of data from 2006.