Simple Arithmetic

I trust my readers understand the Constitutional, moral, and administrative problems with Instant Runoff Voting (IRV), starting with the fact that it's far from Instant.  But D. J. Tice really exposed its inherent flaw, the end of one man, one vote in his March 4, 2010 Minneapolis Star Tribune Commentary.   Allow me to recast his example of a three way race.

Candidate	First	Second	IRV Total
Tom	43	n/a	51
Dick	29	Tom=6, Harry=23	49
Harry	28	Tom=8, Harry=20	0

Tom wins, as he would have in a three way general election. But now D.J. makes a small change, where two of Dick's voters decide to vote for Tom instead.

Candidate	First	Second	IRV Total
Tom	45	n/a	49
Dick	27	Tom=4, Harry=23	0
Harry	28	Tom=8, Harry=20	51

Whoops! Harry wins this scenario even though Tom has even more first place votes.  D.J. notes:

IRV proponents say such outcomes are highly unlikely, and that the same ambiguities affect any election (including a primary) that narrows a field of candidates for a runoff vote. But in theory it does appear that votes could backfire under IRV.

I disagree, in that if these voting patterns took the form of a primary election immediately followed by a two candidate general election (my preference), Tom wins either scenario handily.  With time between these elections, the supporters of the losing primary candidates have time to more thoroughly consider their general election vote, which is a good thing.

I can appreciate the enthusiasm that "young Turk" voters might have for IRV, but they need to take some of the energy to fully and maturely consider the full situation.

Comments

[Miko]

A more visual tutorial can been seen here on YouTube:

http://www.youtube.com/watch?v=5ZGTnp3cgFY

and here:

http://www.youtube.com/watch?v=XEkrFkqIHHI

In no other voting system can getting more votes make you loose the election.

[Hiram]

I tend to view IRV as a solution in search of a problem. Voting is tough enough without requiring a hierarchy.

[The Big Stink]

You want people to think things through? This becomes more problematic than IRV.

[Brent Metzler]

I was convinced that allowing non-first place candidates to win the election was the whole point of IRV. No one wants IRV because Republicans and Democrats aren't winning enough elections, they want it because 3rd parties candidates just can't seem to compete when voters only get one vote.

[anonymous]

I honestly don't have a strong opinion on this one (and Tice is a writer I respect, even if I don't always agree), but there was an interesting counterpoint in today's Strib. With low numbers, IRV tends to fall apart, but once you get into the numbers that generally represent actual elections it's more solid. We're not talking about an election for class president, and in a real election the model holds up.

There's still a philosophical issue for many, but from how both parties are acting right now, I guess I don't have a problem with a system that encourages third parties.

--Annie

[Hiram]

The problem I had with the Strib response this morning is that we are often dealing with very low turnout elections. And I didn't see the problem with extrapolation to higher numbers in this area. Are the problems Speed is having above materially different if you add six zero's next to the numbers he is using?

What specifically is the problem IRV is addressing? That people who finish third in elections don't have much influence? Neither do people who finish second.

[Speed Gibson]

Take our last Senate race, Coleman and Franken nearly tied but neither over 50 percent. Now it would come down to small third party numbers where scenarios like the above could easily unfold.

But it shouldn't come to this. Any system where the front runner loses an election by gaining votes is undemocratic, i.e., subverts one man, one vote.