That title sounds like a punk band. I always tell my macroeconomics students that one of the problems with democratic societies is that we elect people to do something even though there are many situations when doing nothing is either correct or desirable.
I submit that this makes the concept in Bryan Caplan's EconLog post entitled "Voters As Mad Scientists" even more frightening. He asserts that unselfishness amongst voters is worse than selfishness if those voters have biased viewpoints.
If you find it hard to believe that unselfish motivation ever makes the world worse, think about a mad scientist. He imagines he's got the cure for what ails you, but all he's got is a syringe full of cyanide. If the mad scientist were selfish, he'd demand payment for his "treatment," and you'd be safe. "Thanks, but no thanks!" The real danger is the unselfish mad scientist. He'd insist on helping you whether or not you paid. Indeed, he'd probably help you even if you screamed "No!" "You'll thank me once you're cured," he'd insist.
Adding my observation gets us mad scientist voters whose criteria for successful politicians is that they not only carry out their potentially harmful prescriptions, but further, that they have no litmus test for whether action is warranted or not.
Caplan goes on to remark that he "can't help but think of Dr. Frankenstein" when evaluating certain policies. I'll up the ante a bit: electorates appear to toss out politicians who don't build enough Frankenstein monsters.