Interesting spreadsheet (how often do I say that.Fireball1244 wrote:Defiant is correct -- Perot drew disaffected voters from both Bush and Clinton in about equal shares. Moreover, Clinton's lead in the states that he won was so significant that Bush would have had to take a supermajority of Perot's voters to win. None of the exit polls indicated that was possible. From the available data, it appears that had Perot not been in the race, Bush might have won Georgia, Montana, New Hampshire and Ohio, worth 41 more electoral votes than the 168 he won with Perot in the race.
Here's a spreadsheet that I built that allows you to plug in percentages of Perot voters voting for Bush, Clinton or Other, and view the recalculated results. You'll find that it's hard to create a realistic scenario where Perot voters put Bush over the top:
http://s195335120.onlinehome.us/files/1992noPerot.xlsx

That said, Perot's candidacy no doubt changed the campaign itself (one more candidate to deal with, attacks from Perot on the other candidates, the political engagement of the electorate, etc). But there's no evidence to suggest a dramatic change in outcome in a two way match up, especially for a president with low approval ratings.