A partly-correct editorial about voting systems by Jack Lessenberry

In the below editorial, Lessenberry detects a fame-based failure of plurality voting probably about to happen in Michigan. But unfortunately he then draws the wrong conclusion from that, advocating IRV as the cure, whereas a superior cure, which he had not considered, would be score voting.

The error was Lessenberry, with zero direct evidence from any polls or any elections (but with some amount of logic), proclaiming the cure for the problem is IRV (instant runoff voting). Unfortunately, what Lessenberry did not know is: there is some actual evidence on this question, and it suggests IRV is somewhere between "a poor cure" and "not a cure at all" for this problem. A better cure – which that evidence indicates actually does work – is score voting. Also, the Burlington VT mayor election of 2009, conducted with IRV, is a counterexample to Lessenberry's title-claim that IRV gets us the "representatives we truly want." In Burlington,

All the above pairwise figures are counted from their official IRV election ballots.

And yet, despite all that, IRV announced the winner was not Montroll, it was Kiss. IRV also is capable of announcing that candidate X wins, i.e. he is the "best" candidate, whom the voters "truly want" but at the same time, with the same ballots, declaring that very same X is the worst candidate, whom the voters want most to lose! How can IRV announce both these things at the same time? The answer is that Lessenberry's title-claim that IRV elects the candidate "voters truly want" unfortunately is not always true.

Also, although Lessenberry correctly notes IRV is "widely used in Australia" (where it indeed, as of year 2015, has enjoyed far more use than in the rest of the world combined), he failed to mention that the Austrialians themselves want (by large margins in polls) to get rid of it.

Score voting is a superior system, which voters appear to actually want.


'Instant runoff voting' would get us the representatives we truly want

By Jack Lessenberry, Michigan Public Radio, 29 Sept 2015

When the Michigan House of Representatives finally got rid of its two disgraced members earlier this month, we thought that was that.

Nobody imagined there was much of a chance of them reclaiming their jobs.

Well, think again.

Todd Courser and Cindy Gamrat almost immediately filed to run for their old seats – as did way too many other people in their districts. There are 11 candidates running in the November primary for Courser's seat, and eight for the Gamrat seat.

I would bet that any single candidate would badly beat either of them. There's no sign a majority of the voters want them back.

But one thing is clear: Both are far better known than anyone else in their districts.

In fields this large, either could win with as little as, say, 15% of the vote. If that happens, I think there's a good chance any Democrat would knock off Courser in his Lapeer-based district, but Allegan County is so Republican it is quite likely whoever wins that primary will win the special general election in March.

Believe me, the one thing nobody in either party wants is to face a decision as to whether to seat these two if they actually win. That's mostly not because they had an affair, but because they are not team players, defiantly refused to learn how to do their jobs or even how legislation is passed, and spent all their time posturing and making statements.

There's not much we can do about this now, but is something we could and should do to avoid situations like this in the future, and get better representation.

Voters would cast votes ranking the candidates as they chose to.

We leave it up to IRV. IRV is not a person, but a system. It stands for "instant runoff voting," also sometimes called preferential voting, choice voting, or transferable voting.

Here's how it works. Voters would cast votes ranking the candidates as they chose to. If anyone got a majority of the vote in a primary or general election, that would be that.

But if nobody had a majority, the second-place votes from the last place candidate would be distributed among the others, and then the next to last, and so on until someone had a majority.

For example: Suppose Smith got 40 votes, Jones got 35 votes and Johnson got 25 votes. The counters, or more likely computers, would assign the second place votes of Johnson's supporters. Let's say 20 wanted Jones, and only 5 wanted Smith.

Democracy works best when those representing us are truly representative of the feelings of a majority of their constituents.

The result would be that Jones would win, 55 to 45, even though Smith had more first-place votes. Smith wouldn't have any right to feel cheated, since everyone would know how the system worked.

The main good thing this would tend to do is elect people who were, by and large, more representative of their districts, and lessen the odds of fringe candidates winning.

Preferential voting is slowly catching on, though I'm not aware of anyone in Michigan using it. It is widely used in Ireland, Scotland and Australia. They use it now in local elections in Minneapolis and Cambridge, Massachusetts, and a few other places.

Democracy works best when those representing us are truly representative of the feelings of a majority of their constituents. It would make sense to give this a try.

Jack Lessenberry is Michigan Radio's political analyst. Views expressed in his essays are his own and do not necessarily reflect those of Michigan Radio, its management or the station licensee, The University of Michigan.


Return to main page