Improving Election Procedures: An Idea Whose Time Has Come?


Article I.G of the Academic Senate's Bylaws read:

3. To be elected, a candidate must receive a vote from a majority of those delegates present and voting.

4. In the event no candidate for a position receives a majority, the run-off will be limited to the top two candidates with the largest number of votes, including all ties.

This procedure mirrors the procedures used in most elections for political office in the United States. Some elections, like those for the electoral votes for President in most states, allow a candidate with a mere plurality to win an office or a state's electoral votes. Many other jurisdictions hold a runoff election if no candidate receives a majority of the votes cast. It can cost a school district, city, county, or state millions of extra dollars to hold a runoff election. This has sparked an interest recently in Instant Runoff Voting or IRV. Cities such as San Francisco, states such as Vermont, and countries such as Australia and Ireland have switched to IRV with successful results. So, what is IRV? Simply, IRV is a mechanism for ranking candidates during single election.


What happens under the current runoff system? If a candidate is unopposed or there are only two candidates, no runoff will be necessary (unless there is a tie, which I'm told has never happened in ASCCC elections). If there are three candidates and the one you vote for is eliminated, in the runoff, you get to vote for your second choice. If your candidate is still in the race, presumably you will vote for him or her again. With IRV, instead of marking an X next to one name, voters place ranks of 1, 2, 3 next to each name. Those who tally the votes count how many 1s each of the three candidates amass. If any one has more than 50% of the votes cast, that candidate is elected (exactly as in our current procedures). If no candidate has a majority, the one with the fewest 1s is eliminated and those ballots' second choices are distributed to the other two candidates (exactly as they would have been under our present system in a runoff), and the winner is declared. This would save everyone having to go through all the procedures of distributing, marking, and collecting all the ballots twice. For municipal elections and the like, this is where there are significant savings, and the results will be exactly the same.


Using IRV when there are more than three candidates on the ballot may actually result in a different outcome than under our current procedures. Let us say there are four candidates, A, B, C, and D, and assume that A and B are the highest vote-getters, but neither has a majority. Under our present system, there would be a run-off between A and B. Under IRV, if D had the least number of 1s, the 2s on those ballots would be distributed to A, B, and C. It is conceivable that after that distribution, C ends up with more votes than B, and B will be eliminated with 2s (or 3s, if a 2 is a vote for a candidate already eliminated) on those ballots going to A and C, and the one with the most votes after that wins. It is possible that C ends up being the winner, an impossibility under our present system. In the 2000 Presidential election, in some states the vote for Nader and other minor candidates prevented either Bush or Gore from getting a majority. With IRV, the second choices of all the voters for minor candidates being awarded to Bush or Gore may have made a difference in the outcome.

Under the present system, sometimes voters do not vote for their first choice because they fear that if too many others do the same, a candidate they really don't like will be elected. With IRV, voters are freer to vote their consciences, because by voting for a long shot, they can still place a more likely winner in second place.


Suppose there are 90 votes. It would require 46 votes to win an office. Of the four candidates, A gets 40 votes, B 22, C 18, and D 10. Since nobody has 46, under our present system, we would have a runoff between A and B. Under IRV, the second place votes for D would be distributed to A, B, and C. If all ten who voted D first gave second place to C, the votes would now stand: A 40, B 22, C 28, and the final runoff would be between A and C. Now, A would need only six of the 22 second place B votes to win, but it is still theoretically possible for C to prevail.


Does a voter have to give a rank to every name on the ballot? No. You can mark just a 1, or rank all names, or anything in between.

What happens to ballots with votes for a candidate being eliminated in the first round that don't have second choices? Then they will be eliminated entirely. This will change the number of votes needed to win. It is analogous to our present system when in the runoff you may decide not to hand in a ballot because you have no preference for either of the two remaining candidates. In the above example, if six of the 10 votes for D had no second choices marked, then only the other four would be distributed to A, B, and C, and that means there would be 90 - 6 = 84 still voting, so that a candidate would now need 43 votes to win.

What happens when there are three candidates left and the second and third vote-getters are tied? Who is eliminated? I am told that this has never happened in almost 40 years that the Senate has been holding elections, so it must be pretty rare. What would we do under our present rules? Have a run-off with all three, according to the Bylaws, and if everyone votes the same way again, then what? Under IRV, there are several ways to handle it. If ASCCC decided to switch to IRV, all that would have to be done is to decide which method to employ before IRV is used for the first time, realizing that it may never be necessary to invoke it.

Isn't IRV overly complex? For the voter, it is almost as simple as our present system: simply rank the names instead of marking an X next to one name. The votecounters will not have to distribute, collect, and count ballots for a second time. Instead, there will be rounds of re-allocating votes, a slightly more complex procedure than simply counting Xs, but after doing it once or twice, nothing too difficult to master.

There are a lot of sites on the Internet describing IRV. A search engine like Google will provide you with many sites. Try , the site of the Center for Voting and Democracy, to find out more. There is a link to FAQs from their site also: You are encouraged to read more about this innovative system.