E.G.

As Egarwaen points out in comments, there are numerous problems with putting together a reasonable voter test.  We obviously would not want to test people on questions for which there is no consensus on the answer.  But politics is such that almost every question which is not extremely basic is contentious.  As a result, it’s difficult to come up with questions whose answers are both (1) non-controversial and (2) not immediately obvious to everyone; yet it seems that the questions on any useful voter test would have to be both.

I don’t have any firmly-held beliefs about what specific questions should go on a voter test.  But my tentative suggestion is that the test should include questions about extremely basic facts about recent political history.  These should be questions which anyone who has been paying reasonably close attention to politics for the last year or two can answer easily. 

The purpose of such questions would not be to weed out absolutely every unqualified voter.  A test which did so would be undesirable, since a test that rigorous would probably also weed out some qualified voters.  Being able to answer the questions on the test should be necessary, but not sufficient, for being a qualified voter.

Rather, the purpose should be to weed out potential voters who (a) have not even bothered paying attention to recent politics, or (b) who have been deeply misled by mendacious pundits or partisan hacks.  I believe that there are many such voters, and that their numbers are increasing each cycle.