Tuesday, November 13, 2007

Why Ciommo Won and Arroyo Lost in the 2007 Municipal Election

Just before the polls closed last Tuesday night, I posted my predictions for the City Council race. I was pretty much spot-on for the Allston-Brighton District 9 seat (albeit missing the turnout numbers by a lot), but got the wrong odd-man-out in the Councilor-At-Large seat. Here is some analysis of what I thought happened on election day.

Allston-Brighton District 9 City Councilor

I predicted that Mark Ciommo would beat Greg Glennon by at least 900 votes and by at least 57-43% (possibly as high as 60-40%). My prediction couldn't have been closer: Ciommo won by 948 votes by 59.8-39.8%. The full precinct results can be found here.

That said, I screwed up on two points -- one moderate mistake (Wallingford Road) and one a big mistake (voter turnout) -- which magically cancelled each other out to make it look like I hit the lottery.

Turnout. I guessed 6500 voters would turn out, which was at the low end of previous municipal elections. It was far worse at 4745, which was barely over the 4521 votes in the preliminary municipal election.

Why was I so wrong? Not because of the weather. I think that the rank-and-file voters got so little information from the mainstream media (Allston-Brighton TAB excepted), that there was little motivation to vote. I also heard from many voters in the preliminary election who were dissatisfied with the two choices in the final election; I suspect many of them just stayed home, unwilling to be forced to choose.

Wallingford Road.
Glennon carried Ward 21, Precinct 13, full of elderly Russian Jewish voters at the Wallingford Road housing project, over Ciommo 303-142 in the preliminary election. I predicted that he would add to that lead by +200 votes, but the forces of arm-twisting pushed it the other way: Ciommo won 333-242, giving Ciommo a net +252 votes. I blew the vote difference by 452. Oops.

But this is a really, really big story in the neighborhood. Tim Schofield told host Joe Heisler on the Talk of the Neighborhood show that the Russians mostly listen to former State Representative Brian Golden and Mayor Menino; to that I add a likely third party, the Mark Alford-Pat Galvin-(and-his-brother) trifecta. Golden, Alford, and Pat Galvin all supported Glennon -- and Ciommo holds progressive social positions incompatible the conservative Russian voters, at least in the conventional wisdom -- so this means that the Mayoral power in A-B is now at the point that it trumps all else in specific cases of arm-twisting (like Wallingford in this election).

The other big feature of the Wallingford Road vote is that it actually didn't matter in the election. Ciommo could have won without a single vote from the Russians. In looking back on this year's election, the message we should all take away is that the Russian vote can impact a preliminary (or three-way) election in an important way by determining the second-place finisher, but their strength cannot dominate over the powerful base of a long-time resident.

Oak Square battleground.
I predicted that several precincts in the Oak Square neighborhood would represent battlegrounds between the long-term Brighton residents (Ciommo's base) and socially conservative Irish Catholics (Golden's base, and hence Glennon's). Many people in the neighborhood have one foot firmly in each camp, so predicting how they would swing was the parlor game of the election.

Ciommo carried the combined Ward 22, Precincts 7, 11, and 13 in the preliminary, with Glennon and Schofield in a virtual tie for second. Schofield's voters generally swung heavily for Ciommo, regardless of his endorsement, but this was the one neighborhood where that pattern might have been broken.

In the final election, Ciommo won 22/7, 22/10, and 22/11 heavily, and barely nudged Glennon 102-98 in more conservative 22/13. I predicted, "if Ciommo carries 7, 10, and 11 with big margins, then it's lights out." That it was.

My model was that Schofield's supporters went 80-20% for Ciommo, Selvig's 65-35%, and Hanlon's 50-50%. Using that model on Ward 22, Precinct 7, for example, my prediction was Ciommo over Glennon by 137 to 61 votes; the actual result was 131-67, off by six, but within the margin of error of 14 votes (for 198 cast).

Ward 21 Progressive Precincts. I predicted that you should forget-about-it without a strongly identified progressive in the case (like Schofield). Ward 21, Precinct 8 actually doubled their turnout over the preliminary election -- from 1.6 to 3.2%. In the end, the progressive precincts still didn't show up.

Summary. The bosses of Ward 22 were able to get Councilor Jerry McDermott elected in 2002 (he is the cousin to the Galvins, after all), even though McDermott claimed after the election that he was not the Mayor's candidate. (Ciommo told me earlier in the year that he disputes that characterization of the 2002 race, however, since the Mayor did not formally endorse a candidate.) In 2007, however, the Ward 22 bosses no longer appear to be able to throw an Allston-Brighton election to their favorite candidate -- although they can get him a second-place finish and into the finals.

Councilor-elect Ciommo enjoyed powerful support from middle-aged men who chose to stay in Brighton to raise their families here. He won out over the younger, more transient constituencies -- like former Representative Golden and lame duck Councilor McDermott -- who leave town with their young families.

Long-time Brighton residents and the Mayor's influence ended up winning over the ward bosses and families in transition through the neighborhood.

Boston City Councilor-At-Large Race

While many in the city thought that the race for the fourth seat among the Boston City Councilors-At-Large was between Councilor Stephen Murphy and challenger John Connolly, there were many soothsayers predicting Councilor Felix Arroyo to be the odd man out.

I called the race a toss-up until the last moment, when I thought that Connolly's strategic blunder of sending out the anonymous attack mailings would cost him enough votes to lose out on fourth place. I was obviously wrong, and the Arroyo predictions were right.

Why did Arroyo lose? The pundits have rained in with some pretty convincing reasons: low voter turnout working against minority candidates; Arroyo raised little money, while Connolly raised around ten times as much; Arroyo didn't campaign very hard or send out many mailings, while Connolly campaigned like crazy (and sent out quite a few flyers!); Arroyo's record on attendance at City Council meetings and hearings; etc. None of the reasons seems enough by itself, but, taken together, they seem plenty to explain why he got around 15% fewer votes than Connolly.

This year's City Council election is an important one for data mining purposes in the future, because it shows clearly how low voter turnout impacts voting from ward-to-ward. Connolly's West Roxbury Ward 20, for example, voted disproportionately higher than the rest of the city.

The election numbers, plus analysis on a ward-by-ward basis, for the Councilor-At-Large race can be found at Matt O'Malley's website.  He ran, albeit unsuccessfully, in 2005 for Councilor-at-Large, so he knows the ins and outs of all the different neighborhoods and their voting trends.

No comments: