What was the Tribune thinking? (A clarification of IRV)

19 Jan

Last week, the Tribune printed a “My Word” opinion piece about instant runoff voting (IRV). I’m all for sharing varied opinions on any issue, including on IRV, but I think it’s important for editors to fact check opinion pieces to make sure the opinions shared aren’t based on assumptions that are patently false.

The piece starts off with this premise:

Undemocratic “instant-runoff” voting is a giant step in the wrong direction. Can you imagine the cries of outrage across the country if millions of Republicans and tea partiers had been forced to rank-vote for the Democratic candidate Sen. Barack Obama in the 2008 presidential election?

With instant-runoff voting or ranked-choice voting, I will be forced to vote for a candidate I do not like. In the coming Oakland mayoral election, I will be forced to rank-vote Jean Quan, whom I consider to be unqualified to become mayor…

The problem is that the author, James K. Sayre, is sorely mistaken on how IRV works. As explained this a couple months ago, via a talk the County Registrar gave, IRV does not force anyone to vote for a particular candidate. Here is what the ballot would roughly look like for the mayoral election in question:

Despite what Sayre thinks, nobody is going to hold a gun to his head and force him to vote for Jean Quan. Let’s say Sayre’s first choice is Paulette Hogan. He’ll mark Hogan under the 1st Choice column. And let’s say his second choice is Don Perata – he’ll mark Perata under the 2nd Choice column. And since we know he would never vote for Jean Quan, he’ll leave the third column blank.

And what if he hates Perata and Quan? Well, he’d just mark Hogan as his first choice and leave the rest blank.

That’s the beauty of IRV – it’s extremely flexible. You can either rank your choices, or if you really can’t bring yourself to vote for any of the other candidates, you can just mark one choice, just like on the traditional ballots.

Next time, the Tribune should check the facts in the opinion pieces they print and reject pieces like this, which only serves to further confuse people about an already controversial and misunderstood issue.


6 Responses to “What was the Tribune thinking? (A clarification of IRV)”

  1. Helen January 19, 2010 at 8:07 am #

    Thanks for helping set the record straight.

  2. mark January 19, 2010 at 10:30 am #

    It’s easy to understand how voters could be confused about the system, especially if these kinds of ballots are deployed as old fashioned punch card paper ballots. This kind of voting system begs to be delivered via a well designed electronic interface. (With an appropriate printed receipt.) An electronic system could provide user prompts that clarify the process and eliminate some of the confusion. The punch card system could be much better, too, if it were clear from the ballot that voting for a second or third choice candidate were optional, and if the ballot instructions were clear. Well defined information architecture, and good graphic design are not qualities I typically associate with the ballots I’ve seen in a voting booth.

    When the local paper obfuscates rather than clarifies the process it’s even worse. Instant Runoff Voting may not be the panacea that it’s supporters believe it to be, but since we have it, we might as well do it well. I suggest that a little time and money be set aside for interface design and user testing. It’s likely the failure of IRV in the communities where it was tried and then abandoned was due to poor implementation. We should avoid that here in Oakland, if possible.

    • Becks January 19, 2010 at 10:37 am #

      My understanding is that we’re using the same design as San Francisco uses, and their system has been well-tested in two actual elections. There will be a million dollars worth of voter education that happens between now and November. Also, poll workers will be trained to explain the system, in case anybody is still confused when they arrive at the polls.

      Punch cards are not used in Alameda. You either have to fill in with a pen (connect the arrow) or vote on a machine. So the arrow is nothing new and should be understandable.

      My mock up above was just an example based on a talk – I’m sure the actual ballot will have clear instructions.

      • mark January 19, 2010 at 12:33 pm #

        Not having voted here in the last 16 years I’m not familiar with what Oakland ballots look like now. I think we used punch cards the last time I voted here.

        I suspect that some of the problems other communities have had with IRV are due to the conceptual complexity of the process. Voting for a single candidate, even in a field of many candidates is essentially a binary decision making process. IRV introduces a dimensional array of choices. From an information architecture/user interface perspective, this is incredibly difficult to convey on a two dimensional interface (paper ballots of any kind). I haven’t seen any other ballot designs, but I suspect most paper based systems must be similar to what you’ve shown above. Regardless of how you mark the ballot (punch or fill in the dot or draw a line) the complexity comes with knowing which options are available after having made your first choice. On a paper ballot, it is easy to see how a person might vote for Jean Quan three times as first, second, and third choice. In an IRV system that’s the same as voting once for Jean and wasting the option to vote for a second or third choice. An electronic system could remove Jean’s name from the second choice column once she was chosen as a first choice, informing the voter about meaningful options for a second choice. Paper ballots are limited in this way.

        Another complexity with a dimensional array vs. binary system is the issue of understanding the value of each ranked vote. Mr. Sayre’s confusion is understandable, and points to a problem with a ballot showing three choices in a three way race. In a ranked system with three candidates, there are really only two meaningful votes, the first and second choice. Marking your third choice would never be necessary. (The remaining choice after you’ve selected your first two choices is your defacto third choice.) At a minimum the ballot should be designed so that the number of voting choices is one fewer than the number of candidates.

        • dto510 January 20, 2010 at 4:18 pm #

          I think the IRV Charter amendment actually says that there should be a minimum of three choices in an election with more than two candidates.

          How do write-ins work? Is there a space for multiple write-ins?

  3. JB February 1, 2010 at 3:57 am #

    Good to pay attention to these details, but voters have overall had very good experiences with IRV. There’s a collection of exit poll studies here –

    There’s a new one from Minneapolis –

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: