By Stephen H. Unger, professor of EE/CS at Columbia University
We should be careful about drawing conclusions from the IEEE use of Approval Voting. It was initiated by the established power group in order to reduce the chances of Irwin Feerst being elected president. The situation was as follows: The IEEE Board of Directors generally nominated 2 candidates for president. Candidates could also be nominated by petition (and there were cases where petition candidates won). Feerst was a smart, hard hitting man, who often hit targets who deserved to be hit, and who was popular with rank-and-file IEEE members. As a write-in candidate, Feerst once came very close to winning an election and really scared the established group. They feared that he might win because the anti-Feerst vote was being split between the two B.o.D. nominees. Approval voting allowed many people to vote for both B.o.D. nominees thereby reducing Feerst's splitting advantage. This tactic did seem to work – although the fact that some of Feerst's most vicious attacks were on other rank-and-file leaders who he considered as competitors was also a major factor in his failure to be elected IEEE President. The real reason Approval Voting was terminated, was not because of any problem with it, but because Feerst died so that the in-group felt it was no longer needed.
WDS: I've heard this theory from several IEEE members, not just Unger. So this, if not the truth, is at least widely believed.
Further note by Steven J. Brams (professor of politics, NYU):Actually, I worked closely with Jack Nagel to get the IEEE to adopt AV. We tell the story in
Brams and Nagel: "Approval Voting in Practice," Public Choice 71, 1-2 (August 1991) 1-17; and see alsoWhat you and Stephen Unger report about the IEEE's 2002 decision seems to me accurate. I add in my book, however, the following: "There may be other reasons for the abandonment of AV, but I was not privy to such information."
Brams and Fishburn: "Going from Theory to Practice: The Mixed Success of Approval Voting," Social Choice and Welfare 25, 2 (2005) 457-474.
Compare with the IEEE"s official story.
Return to main page