It turns out that more knowledgeable voters benefit from the use of heuristics more than less knowledgeable voters do.
Science News has a report on recent studies by Richard Lau and David Redlawsk on the use of heuristics by voters. It affirms what political scientists have been saying for a couple of decades, which is that voters are more likely to rely on heuristics when voting than to actually pay close enough attention to the candidates to make a fully informed decision.
But, following along the lines of research by Gerd Gigerenzer (and in contrast to the arguments of Daniel Kahneman and Amos Tversky), Lau and Redlawsk find that heuristics, based on emotional reactions to candidates, can lead to better voting (defined as choosing the candidate closest to one's views) than fully informed voting does. The reason essentially is that too much information inevitably leads to points of disagreement between the candidate and the voter, confusing the voter (something I'm certainly experienceing this year). This is nice to see, since I've long thought Gigerenzer got the better of K&T.
But, interestingly, the positive effect is mostly confined to better informed voters. The less informed voters had, predictably, a harder time choosing the best candidate, but less predictably their performance declined when using heuristics.
The moral, I guess, is that if you really care about whom you vote for, you should stay politically informed, then follow your gut.