As mentioned here the other day (What the People Said,  March 5), Instant Runoff Voting (IRV)  ”worked” in Burlington March 3 because the results showed that Mayor Bob Kiss, who won, would have beaten Republican Kurt Wright in a non-instant runoff, the kind  that would have been held a week or two later and cost more money.

But it turns out that Democrat Andy Montroll would have beaten Wright one-on-one, also. In fact, an examination of the results indicates that Montroll would have beaten the Progressive Party’s Kiss as well.

And there is at least a plausible mathematical/political case to be made that if Burlington had been using a more traditional, two-stage election system, Montroll might have been in that final round.

IRV ain’t perfect.

No voting system is, not even the straight-up , most-votes-wins system, used in most of the country. In the United State House of Representatives, for instance the Democrats now hold (as did the Republicans throughout most of the last decade) a much larger majority of seats than the party’s share of the popular vote. So a “majority (or plurality) rules” system does not always reflect the actual majority of the people.

But IRV is more controversial. (It is, for those who have not been paying attention, a system in which voters rank candidates in an order of preference, and if no candidate is the first preference of a majority, the bottom contenders are eliminated and their ballots redistributed to the remaining candidates according to the next ranking on each ballot, until one candidate gets a majority)

Almost two weeks after the Burlington election, advocates and opponents of the voting system are still turning the blogosphere blue with acrimonious debates which range from abstruse mathematical formulations to personal attacks which occasionally rise to the level of, “yah mudda wears army boots.”

Inside baseball, yes. But that doesn’t make it unimportant. This is an argument over how public officials are elected. Considering that one side in this debate wants Vermont’s governor to be elected though an IRV system, somebody ought to pay attention.

This effort to pay attention will avoid all references to the footwear orientation of anybody’s parent, and will delve into the mathematical minutia only enough to demonstrate that IRV has its problems.

That’s because neither the personal squabbling nor the math is at the center of the dispute. Politics is. The IRV debate is hostage to competing visions – mostly but not entirely on the left side of the spectrum – of political strategy and behavior.

But let’s start with one bit of math, to wit, with monotonicity.

With what?

It means that candidate A should not rank lower in the vote count if he gets more votes. If that happens to candidate A, non-monotonicity prevails. That’s a flaw in the system.

That’s what happened. According to perhaps the calmest assessment of the vote, by pro-IRV analyst Wes Hamilton of Middlesex (writing as wdh3 on the Integral Psychosis web site), “Kiss won, but if 753 Wright voters had instead voted for Kiss, then Kiss would have lost…”

In other words, Kiss might have lost had he received more votes.

(Confused? For details, see Hamilton’s full article).

This does not mean that IRV is a bad system. It has its advantages. It’s cheaper than a separate run-off. It renders personal attacks risky (if you enrage the supporters of another candidate, they’re less likely to rank you second or third). It eliminates the danger, in a straight plurality election, that a candidate most voters really don’t support (Wright, in this case) could get elected.

Furthermore, non-monotonicity is a danger in regular run-offs, also. But a smaller danger. In a standard runoff, the voters know who the two front-runners will be before making their final choice. In IRV, they’re guessing. (On the other hand, turnout usually falls off for the run-off election, a disadvantage IRV avoids).

And IRV has other problems. University of Vermont political science professor Anthony Gierzynski said that one of them is that it “discriminates against classes of voters by adding complexity the ballot.” Poorer and less educated voters, Gierzynski said, are more likely to be confused by this complexity.

Not so, said pro-IRV Progressive activist Terry Bouricious (pictured above right), a former Progressive Party Burlington City Council member and state representative, who pointed out that  ”voters in the low-income renter wards were as likely to use additional rankings as voters in the more affluent wards…There is no evidence of any class bias in the actual use of ranked-ballots.”

But this is inferring individual behavior from aggregates, a social science no-no. There are well-educated, affluent folks in those neighborhoods.

Not that there is conclusive evidence that Gierzynski is right, either. There simply haven’t been enough IRV elections to make that determination. But there is ample circumstantial evidence that IRV can be confusing, and obviously more confusing to the less educated. In support of this contention, allow me to present Exhibit A: Me.

I took me weeks of reading and interviewing to figure out exactly how IRV works. And I have covered some or all of 11 presidential elections, statewide races in 49 states (never got to Hawaii) and mayoral contests in every big city in the country. Granted, some will point out a paucity of intellect as a complicating factor here. But all that experience ought to mitigate that defect at least to some extent. So if IRV confused me, it’s likely to have confused others.

Again, all voting systems have their drawbacks and IRV has it advantages. But regarding it as the cure-all for electoral problems seems premature at best. Bizarre outcomes are as possible under IRV as under any other system. The Burlington election may have been an example of that.

In a sense, though, the technical back-and-forth about the pluses and minuses of  IRV misses the main point. As both Gierzynski and Bouricious acknowledge, the dispute is really about political philosophy.

Gierzynski’s preferred remedy is not so much a different election system as a different political culture, in which those with “a (mostly) shared ideology,”( in Vermont, that means Democrats and Progressives) coalesced in one party, fighting out their differences in primaries before presenting a unified front in the general election.

Bouricious thinks that’s a bad idea. He derides that as a “political duopoly,” and prefers a multi-party setup  such as prevails, he said, “in almost all the other democracies” in the world.

Well, not all. Britain has what is basically a two-party system; the third party Liberal Democrats are weak and likely to get weaker now that both major parties have moved to the center. Its de facto two-party system provides stability, in contrast to the multi-party Italian situation in which governments are lucky to last a year. Or the Israeli situation in which the minority (and in at least one case corrupt) ultra-religious parties exercise disproportionate power.

Besides, in most multi-party democracies, there are effectively two parties, as the several parties coalesce into two competing blocs.

“They do it after the election,”  Gierzynski  said. “Typically here in the U.S. we do it before the election so voters can see. We get to choose in our elections which coalition is going to run the government. They (voters in multi-party democracies) have no control over the coalition-building. It’s done behind closed doors.”

But to the many Vermonters who take a different point of view, IRV has another advantage. It protects a minor party such as the Progressives (I know, legally it’s a “major party.” But it isn’t) from being labeled a “spoiler” by taking enough votes from their sort-of allies, the Democrats, to throw a race to the Republicans. That’s what happened in 2002 when Progressive Anthony Pollina diverted enough votes from Democrat Peter Shumlin to elect Republican Brian Dubie.

In other words, whatever its advantages, in this state IRV is among other things a Progressive Party Preservation Act. Whether that’s a plus or minus depends on whether one thinks Vermont is best served by two major-party coalitions or by several independent parties, a matter over which reasonable people may differ. And do.

Those who have been loyally following this web site from its inception might remember (but no problem if you don’t; there will be no exam) that the very first entry (December 5, 2008) outlined a schedule of  one post every weekday, with perhaps some exceptions.

As in: (1) every once in a while a Newsguy needs a day off; (2) Some stories require too much reporting to leave time for writing, so on some days, to allow for elaborate research and reporting, no post would be written.

This is one of those days. The first one, actually. Not because anyone needs the day off (that will happen by and by), but because the proprietor is working on not one but two (count ‘em: two, as in ‘2′) intricate and complex stories. Rather than rush either into print (if ‘print’ is the right word for on-line documents) before it is properly cooked, the decision has been made to report them both out fully and write them for later in the week.

One tomorrow. The other, we hope, Thursday.

So that’s it for today except for the following two announcements:

1-Those who read Friday’s post before noon are encouraged to scroll down to it to see the update added in the middle of the day, correcting the assertion that no newspaper bothered to inform its readers that there will be a contest for Democratic State Committee Chair when the Dems meet this weekend. The Brattlboro Reformer did.

2-Starting tomorrow, readers who want to comment on our posts – or on other comments on our posts – will be able to do so without registering on the site and getting a password.

Simply email your comment to me (jon@vermontnewsguy.com). I will accept the heavy burden (about a minute per message, I’m told) of registering you, complete with getting you a password, and will post your comment onto the web site for all to read (and comment on should they choose).

Dispensing with registration completely seems not to be a good idea. It attracts spam and other horrors of modern life. But this should make it easier for some readers to join our discussion, which is all to the good.

Happy St. Patrick’s Day

Is Vermont a dangerous place to be a woman?

Or, for that matter, a man?

It’s dangerous to be a person anywhere. But a new report suggests that Vermont might be one of the states where a person is more likely to be murdered by his or her significant other, present or former.

The new report by the Domestic Violence Fatality Review Commission reveals that eleven people, nine women and two men, were murdered in domestic violence incident last year.

That’s a substantial majority of the 15 adult homicides from all causes in 2008. Five children were also murdered last year, but the Commission’s “protocols” do not cover child deaths, and it referred the child cases to the Child Fatality Review Team.

From a national perspective, all these are small numbers, and one reason domestic violence murders are such a big percentage of overall homicides is that homicide in general is relatively rare in the state.

“There are not a lot of stranger danger offenses,” in Vermont, said Assistant Attorney General Amy Fitzgerald, who coordinated the 15-member commission made up of state officials and representatives of private organizations who work to combat domestic violence.

But if the numbers are small, the percentages seem to be comparatively large. Two years earlier, when six Vermonters were slain in domestic violence incidents, the state ranked sixth in the nation, per capita, for homicides committed by men against women.

That study did not deal with crimes by women against men. In fact, it was called “When Men Murder Women: An Analysis of the 2006 Homicide Data.” It was  conducted by the Violence Policy Center, a non-profit educational organization.

According to its report, only Nevada, South Carolina, Alabama, Oklahoma, and Louisiana had higher per capital rates than Vermont of such murders. Vermont’s rate was 1.90  per 100,000 population. The national average was 1.29 per 100,000.

All the other New England states had far lower rates of these murders in 2006. New Hampshire had none.

With five female victims of domestic violence murder last year and no big increase in the state’s population, Vermont “may find itself ranked again as having a high rate of female homicides per capita,” according to the Commission report.

“We still unfortunately could be in the to ten,” Fitzgerald said. “It’s not such a great honor to have.”

Statisticians warn against reaching broad conclusions based one a limited number of cases. And the Violence Policy Center appears to be engaged in advocacy as much as research. Furthermore, there appear to be no statistics showing that that Vermont has a higher rate of domestic violence in general, as opposed to domestic violence leading to fatalities. So the evidence at hand does not in itself prove that Vermont has a worse domestic violence problem than other states.

But perhaps neither should two “high” rankings in three years be casually dismissed. In 2007, when there was no Violence Policy Center report, seven adult Vermonters were murdered in domestic violence incidents. Only two and four, respectively, were killed in the preceding two years, but ten were killed in 2003, three of them by a partner or ex-partner and four by a family member.

“I don’t think your numbers necessarily small, considering the size of Vermont,” said Neil Websdale, a professor criminology at Northern Arizona University in Flagstaff and one of the recognized experts in the field.

The numbers are big enough, Websdale said, “to raise the question of whether there is something in Vermont in terms of service delivery, say in availability of shelter services, or in (emergency medical technician)  response times in remote rural communities. That can make the difference between a bleedout that becomes a death or a bleedout that is stemmed.”

What makes the Vermont situation more mysterious – and possibly more troubling – is that its demographic profile indicates that it should have a lower-than-average rate of domestic violence. Websdale said domestic violence is more common where people are poor, especially where the poor are crammed together in crowded urban neighborhoods, and where there are a disproportionate number of males between the ages of 18 and 34.

But Vermont has a low poverty rate, no urban slums to speak of, and a lower than average percentage of males or females between 18 and 24.

There were actually more male than female victims of domestic violence homicides in Vermont last year. Six men and five women. But nine of the “responsible parties” were men, and only two were women. According to the report, “five males killed five females, four males killed four males, and two females killed two males. Three of the male killing male cases involved an estranged partner confronting his estranged partner and her new partner. In one of these…the new partner came to the defense of the female and tragically killed the estranged partner.”

One of the murders was in Washington County. Four were in Rutland County, two each in Chittenden , Windsor, and Essex, the first murders there since 1973. Five of the victims were killed by a partner and three by a family member. Most of the murders were committed in or right near the victim’s home.

Seven of the 11 domestic violence victims were killed by gunshot, the report said, four from handguns and three from rifles. In two of those cases, Fitzgerald said, the shooter was a “prohibited person” who was not permitted to have a firearm under the Federal Gun Control Act.

“One of highest risk factors that make a difference between  homicide and assault is the availability of handguns,” Websdale said. “If enforcement around the Brady Law is weak, that could be a significant factor.”

In Vermont, the enforcement is weak. Unlike New Hampshire, where Fitzgerald said there is, “a state statute setting up a process for relinquishing guns,” Vermont has a hodge-podge system in which judges handle different cases in different ways, sometimes handing the weapons over to a family member.

In one recent case, Fitzgerald said, the judge allowed the family member to return the “prohibited person’s” rifle to him for hunting season. The man killed himself. In its report, the Commission recommends that “defendants not be allowed access to guns…even during hinting season (and that) family members not be given the responsibility of storing the defendants’ weapons.

“Firearms are a huge issue in terms of the percentage of homicides,” Fitzgerald said, but “we don’t know whether lack of gun control here has a correlation (to the high murder rate).  We go back and forth on the gun control issue because it’s a political hot potato.”