Michael McDonald actually looks at the evidence:
(3) Negative ads turn off voters and reduce turnout.
Don’t be so sure. The case on this one is still open. Negative TV advertising increased in the mid-1980s, but turnout hasn’t gone down correspondingly. The negative Swift boat campaign against Sen. John Kerry, D-Mass., apparently did little to depress turnout in the 2004 presidential race.
Some academic studies have found that negative advertising increases turnout. And that’s not so surprising: A particularly nasty ad grabs people’s attention and gets them talking. People participate when they’re interested. A recent GOP attack ad on Rep. Harold E. Ford Jr., D-Tenn., a Senate candidate, has changed the dynamic of the race, probably not because it changed minds or dissuaded Democrats, but because it energized listless Republicans.
We’ll have to wait to see whether the attack on Ford backfires because voters perceive it as unfair. That’s the danger of going negative.
He actually goes through five voting myths. The standard line that everyone hates negative advertising is somewhat true in that everyone knows they are supposed to dislike it so they tell reporters and pollsters that they don’t like it. It’s also effective and it’s not clear what it does to turnout.
Voters sort it out. They don’t do it by long hours of analysis, but by impressions and remarkably, those impressions serve them well over the long run even when they might miss it on specific elections.