I'm happy to report that the ASSU (Stanford student government) has approved my bill to use IRV for election of its president. The ASSU has been plagued by runoff elections that most students find annoying; grad student turnout in the runoffs hovers between 3 and 5 percent. I had to face several challenges in this effort that you might find instructive, or at least interesting. First, the ASSU constitution spells out a detailed presidential election procedure: one vote per student (counting blank ballots), a majority requirement, and a runoff if that requirement is not met. Amending the constitution is practically impossible. Fortunately, the constitution has no provisions governing who gets to participate in an election. So the bill implements IRV by defining all but the final round as a series of 'instant primaries' that narrow the field of candidates to two. The final round is defined as the election itself. To guarantee a winner, I had to contrive an additional instant runoff that discards exhausted ballots. Second, I had to face a tough crowd - last year's presidential election had only two candidates but was spoiled by none-of-the-above votes. The loser of this election went on to win the runoff and is now the ASSU president. So the current president and his supporters are now quite fond of runoff elections. They raised objections about the value of extra campaign time to educate voters, and claimed that ranked ballots would have a disenfranchising effect for those who didn't understand them (specifically, voters need to know that voting for fewer candidates makes it more likely for their votes to end up in the 'exhausted' pile). I counterargued that voters need to be thoroughly educated before the FIRST election for maximum legitimacy and accuracy. After noting that more people cast decisive votes in an IRV election than a low-turnout runoff, I promised to help with a voter education effort. The bill squeaked through the undergraduate senate by a one-vote margin. The Grad Student Council was very excited about the bill once I convinced them that it was constitutional. I enjoyed seeing GSC members who were unfamiliar with IRV discover its implications as the discussion went on. One member was from Australia and had good things to say about his experience with it there. It passed unanimously. [Fast forward to April 2001] Stanford just had its first IRV election for president of the student government. It was a five-way race (not counting write-ins). In the first round, the front-runner had close to a 2-to-1 margin over the second place candidate, but he didn't manage to muster a majority of all votes until all candidates but those two were eliminated. A runoff election, along with two additional weeks of campaigning, was avoided. The winner was very skeptical of IRV when I first proposed it, and he voted against it. He has now reversed his opinion and is very relieved to be done with his campaign. Perhaps best of all, a few thousand more people are now familiar with IRV.