When we talk about changing election rules, we should keep things simple. We should only make changes if they clearly give voters more influence and remain fair and easy to understand. If people do not see the logic or trust the results, even the best reform will not help. If something already works, we should not change it just because a new idea sounds interesting.

This year, the Charter Commission received 281 proposals, including several about changing how we vote in Honolulu. It can get complicated quickly, so let’s focus on one piece of the puzzle: our ballots.

Currently, we use a “pick one” ballot and hold a separate runoff if no one gets 50%. At least three new proposals would change this to a ranked-choice ballot, combining the runoff into one election.

A Special Commentary Project

Civil Beat is focusing on transparency, accountability and ethics in government and other institutions. Help us by sending ideas and anecdotes to sunshine@civilbeat.org.

Read More

How City Council Elections Work Now

Today, each of the nine council members represents a geographic district. In the first round — the August primary — voters see a list of names and fill in one oval for each name. If a candidate gets more than 50% of the vote, the race is over. If no one gets a majority, the top two move on to a second, head-to-head election — the November general election.

There are strengths in that system. We do end up with majority winners. The ballot is simple and familiar. Most voters understand how it works.

But a winner-take-all, choose-one system is not perfect. Turnout often drops, holding two elections costs more and the “pick one” rule encourages people to vote strategically. In a crowded first round, voters worry about wasting their vote on someone who cannot make the top two, or splitting the vote between similar candidates and helping someone they like less. Many people end up dropping their first choice and pick someone they think can win.

Here’s how that dynamic plays out at my family’s Christmas dessert table.

The Dessert Problem

Let’s say my family is responsible for bringing one dessert for Christmas dinner with our extended relatives. Before he heads out to buy it, my dad gives us a short list of favorites: Liliha Bakery Coco Puffs, UBAE Ube Cheesecake, Ted’s Chocolate Haupia Pie, Hawaiian Pie Co. Pumpkin Custard Pie, and Paalaa Kai Bakery Snow Puffies.

Now imagine my dad says, “We can only bring one. Choose one and write it down.”

Under that rule, I start gaming it out. I love Ube Cheesecake. My dad loves ube too, but Pumpkin Custard is his favorite thing in the world, so he is probably going to write down Pumpkin. My niece doesn’t want pastries, and my sister has already announced that we must have a pie for Christmas. My mom will almost always pick the chocolate option.

I’m not just thinking about what I want to eat. I’m thinking about what other people don’t want to eat. Clearly, the pastries don’t have the support. The Ube cheesecake is a long shot. And, we’re still polishing off the giant Costco pumpkin pie from Thanksgiving, and I really don’t want more pumpkin. So I write down Chocolate Haupia Pie as a “safe” non-Pumpkin choice that I think can win.

Around the table, everyone is doing a version of the same thing. People are voting mostly to block their least-favored option, not to support the dessert they actually think is best.

For pie, that’s fine. When we’re choosing a council member who will vote on a multibillion-dollar city budget, rail extensions, zoning and public safety, it matters more whether people feel free to choose their true first choice without feeling like they are throwing away their vote.

How A Ranked Ballot Would Work

Now, back at the dessert table, my dad could ask us to rank our top three choices instead. Now we all write down our honest rankings. Let’s say it looks like this:

• Dad: 1) Pumpkin Custard, 2) Ube Cheesecake, 3) Chocolate Haupia
• Niece: 1) Pumpkin Custard, 2) Chocolate Haupia, 3) Ube Cheesecake
• Sister: 1) Chocolate Haupia, 2) Pumpkin Custard, 3) Snow Puffie
• Mom: 1) Chocolate Haupia, 2) Snow Puffie, 3) Coco Puffs
• Me: 1) Ube Cheesecake, 2) Chocolate Haupia, 3) Coco Puffs

There are five voters, so a dessert needs three votes to win a majority.

Round one:
We count everyone’s first choice.

• Pumpkin Custard: 2 (Dad, Niece)
• Chocolate Haupia: 2 (Sister, Mom)
• Ube Cheesecake: 1 (me)

No dessert has three votes yet. Ube Cheesecake is in last place, so it is eliminated. But my ballot is not thrown away. My vote moves to my next-ranked choice, which is still in the running: Chocolate Haupia.

Round two:
We count again with Ube gone.

• Pumpkin Custard: still 2
• Chocolate Haupia: now 3 (Sister, Mom, me)

Chocolate Haupia Pie now has three out of five votes, which is a majority. It wins.

This voting experience is completely different. I could safely rank my favorite first, even though it was a long shot. When Ube dropped out, my vote still helped choose between Pumpkin and Chocolate in the final round. And all five of us put Chocolate Haupia Pie in our top three, so the outcome feels more representative of what we want.

The Case For Changing Our Ballot

Research from places that use ranked choice voting shows some consistent benefits. The decisive round happens in a single, higher turnout election instead of a lower turnout preliminary. Voters can express their full preferences without worrying as much about “spoilers,” and new or underdog candidates can compete without being punished for splitting the vote. Campaigns move away from negative messaging, since candidates want to be the second or third choice of people who prefer someone else first. It also saves money by eliminating runoffs.

There are trade-offs. Counting and explaining the results is more complex, even if ranking candidates feels natural. Good voter education helps, but it doesn’t happen by accident. New York City used practice ballots and light-hearted “bagel elections” so people could try ranking everyday choices before they saw a real RCV ballot.

Evaluations there and elsewhere suggest most voters found ranked ballots easy to use and often prefer them to their old system, but getting to that point took deliberate outreach.

We should also be clear about what a ballot change does not do. Winner-take-all, ranked choice voting still elects one person per district. It makes the choice more expressive and efficient, but a large minority group can still end up without a council member of their choice if they are in the minority everywhere. Fixing that will take additional change.

But, for now, let’s come back to the test.

Does moving to ranked choice voting clearly give Honolulu voters more say? And can we roll it out in a way that feels fair and easy to understand?

Reasonable people can answer those questions differently, but I think the answer is yes. Our current “pick one plus runoff” system mostly works, but it has real weaknesses. Ranked choice voting offers clear advantages and clear trade-offs.