Reply #55: It's all in the numbers [View All]

[wyldwolf]

First, look at the turnout. Perot got 19,660,450 votes. The total turnout was more than 13 million higher than in 1988. So, even though Perot got a lot of votes, 13 million of those voters didn't vote in 1988. Clinton ran 3.1 million votes ahead of Dukakis, but Bush received 9.7 million fewer votes than four years earlier. The two party vote fell by 7 million. So, Perot only took 7 million votes from the two parties combined. If Perot had not been in the race, would those 7 million Perot voters who voted for Bush and Dukakis in 1988 have voted for Bush by a sufficient margin for him to overcome Clinton's 3.1 million vote lead. Those 7 million Perot voters would have had to favor Bush over Clinton by 5 to 2. Or, even if all 19.6 million Perot voters had voted for one of the major party candidates, they would have had to favor Bush by a 58% to 42% margin to overcome clinton's lead and tie the race. Was this likely in view of the fact that the other 84 million voters were favoring Clinton by 7%, 53.5% to Bush's 46.5%?

--- snip ---

If Perot cost Bush the election, the proof must lie somewhere else. On a statistical basis, it's essentially impossible to make a case for Perot costing Bush the 1992 presidential election. The election results show that Perot took many voters from Clinton among his supporters who demonstrated a low interest in politics by voting only for President and Governor, while taking marginally from Bush among those who demonstrated more commitment by casting ballots for Congress.

http://www.leinsdorf.com/perot.htm

See also graphical analysis:

http://www.swingstateproject.com/2004/05/all_state_votin.html