Warren Mitofsky, who more than anyone else created the exit poll, died on Friday. I have known Mitofsky since 1974, when I was an off-air election night analyst for CBS News, and kept in touch with him over the years. Often in foreign locales: I managed to see him when I was covering and he was conducting exit polls in the elections in Russia in 1996 and 2000 and in Mexico in 2000 and this year. He was always accessible, despite the hectic work of putting together an exit poll in a foreign land. I had lunch with him in Mexico City on June 30, two days before the Mexico election.
At lunch he told me about his one meeting with Mexican President Carlos Salinas, in the run-up to the 1994 election. There was great controversy over the results of the 1988 election, in which the government computer broke down at one point and then came back on line and reported a big margin for Salinas, the candidate of the ruling PRI. Salinas wanted Mexico to have a credible election system and believed that an independent exit poll would help to validate the result. He asked Mitofsky for other suggestions, and Mitofsky said, "It might help if your party lost some elections." Salinas was not amused. But PRI was declared the loser of some gubernatorial elections during his term in office.
Mitofsky was well versed in polling theory and was a stickler for what he believed was correct procedure. He was always ready to respond to serious criticism, and after the 2004 Edison/Mitofsky exit poll produced results that were off the mark, he conducted a serious examination of the firm's procedures, the results of which are accessible through Mark Blumenthal's mysterypollster blog. Here, on his new website, is Blumenthal's appreciation of Mitofsky.
Some people decry exit polls, and some wonder whether they are worth the expense to news organizations. I think they are, not so much to project the results-those will be known in time-but to provide information about what kind of people voted for which candidates. This is and will be invaluable to historians and future political scientists, as well as to the intended audiences. Warren died suddenly, while still busy and in the midst of his career. I hope he had the satisfaction of believing what was in fact true: that he made a significant contribution to our understanding of our times.
Reader Comments