Friday, February 24, 2017
My Account  | Login
Nashua-BoireFieldAirport;41.0;;2017-02-24 05:15:34
Monday, December 5, 2011

Participate in our online alternative-voting ballot test (no sharp pencils needed)

David Brooks

Would American history be different if the New Hampshire presidential primary used non-traditional ballot methods like ranked-choice or approval voting, instead of the usual single-vote-per-ballot, plurality-wins method?

That question probably hasn’t kept you awake much at night, but it generates frenzied debate among those pondering the ins and outs of voting processes. (Judging from the e-mail I get, that includes more people than you might think).

They say different types of ballots would produce different results from the same group of voters. But is that really true?

Experimenting with real voters in real elections isn’t possible. So, in a desire to throw a little light on this intriguing question, GraniteGeek has created an online version of the Republican presidential primary, to provide an entertaining and insightful, although wildly unscientific, look at the effect of alternate ballots.

This issue came up recently in this column because Portland, Maine, decided to use a method called instant-runoff election to choose a mayor. Subsequent discussion at the GraniteGeek blog led to this online experiment.

In Portland, voting day there went smoothly despite the novelty of the complex ballot, ballot-counting took an extra day but was otherwise unremarkable, and nobody has challenged the results. So it has to be judged a success by fans of alternative voting methods.

On the other hand, the man who won would almost certainly have won anyway, judging from the results, so why bother?

That question led to our little experiment.

We have established a fake GOP ballot online, in which we ask people to rank some or all of the Republican hopefuls from first to last, rather than merely voting for one as we can do Jan. 10.

This is known as ranked-choice voting. It lets voters express their opinion about more than one candidate, but the ballots are big and cumbersome, and voting is a lot more work.

We’re using it because the same ballots can also be counted as if they were approval-voting ballots and traditional plurality ballots, which lets us compare methods. (They can also be counted as instant-runoff ballots, but that is so complicated I’m not sure I can do it. We’ll see.)

Anybody can participate, regardless of political affiliation, age or state of residence. Telegraph readers in Massachusetts – at last you’ve got a chance to vote in a primary!

The ballot will be kept online for a couple of weeks, to get as much participation as possible. I’ll discuss the results later in this column, probably around Christmas.

Aside from the intriguing geeky aspect of the issue, why does it matter?

Well, the advocates (including several legislators who tried unsuccessfully to get New Hampshire to consider approval voting) say alternative methods better reflect voters’ wishes and even lead to less partisan, extreme politics because they force candidates to appeal to a broader swatch of Americans.

Opponents say ... well, there aren’t many opponents because most people haven’t thought about it, but the general reaction is that they’re too complicated and unnecessary.

Knowledgeable opponents can point to Arrow’s Theorem, which helped win a Nobel Prize for its inventor, economist Kenneth Arrow. It says, roughly, that no voting method is perfect when it comes to reflecting a population’s wishes.

Arrow’s Theorem doesn’t mean America can’t improve on traditional plurality voting, it just means that any alternative has flaws.

Please note two things:

We are not trying to predict the outcome of the GOP primary.

We’re using that race only because we need a high-profile race to draw interest, and we need a multicandidate race to test whether different methods of counting votes leads to different results from the same ballot.

We are not trying to predict which voting method is more accurate.

Fans of ranked-choice voting say its complexity prevents people voting “strategically” – deliberately voting for somebody they don’t like to indirectly help their candidate – and produces “sincere” results that better reflect voters’ wishes.

They claim that people approach different types of voting with different mindsets and therefore mark ballots differently, depending on how they’re going to be counted. Hence, counting the same ballot in different ways doesn’t reflect what would actually happen with different types of ballots.

Still, it’s an interesting experiment.

One thing for sure: There won’t be any hanging chads.

Granite Geek appears Mondays in the Telegraph, and online at David Brooks can be reached at 594-6531 or