Don't count Dean out

[Nym90]

Exit polls in 1992 showed that the Perot voters would have split about evenly between Clinton, Bush, and not voting. So all of you who say that Perot cost Bush the race are basically saying that you believe the voters lied in the exit polls, or that the exit polls were way wrong. But they were right on everything else, so that's a pretty tough argument to make.
And to say that Perot cost Dole victory in 1996 is even more patently false, since even if every Perot vote had gone to Dole, he still would have lost in the popular vote to Clinton.
It's a complete myth that Perot cost Bush victory in 1992, and in fact he may have hurt Clinton's legitimacy by costing him a majority of the popular vote which he would have gotten in both 1992 and 1996 if Perot had not run.

[Nym90]

True, but since the exit polls were right on in determing the actual Clinton Vs. Bush result, there wouldn't seem to be too much evidence to suggest that they were inaccurate. The exit poll that showed Perot voters splitting evenly was also right in determing in the Clinton-Bush-Perot result. So people weren't lying and saying that they voted for Clinton just to say they were with the winner.
Of course no one knows what would've happened if Perot hadn't run. The entire dynamics of the campaign would've been different. But I can't see any solid evidence to suggest that Bush would've won twice as many of the Perot voters as Clinton did, which is what would've had to happen in order for him to win. And that's assuming that all of the Perot voters had still voted, which many of them wouldn't have. If, say, a third of the Perot voters had stayed home, then Bush has to win 3/4 of the Perot vote in order to win.