Tuesday, November 13, 2007

Election Urban Legends: Checking Out the Facts

Rain Caused the Low Voter Turnout

Boston's 13.6% turnout was far less than the 23-32% turnout in all previous municipal elections without a mayoral race since 1985. Inclement weather can certainly reduce the turnout for elections, and November 6, 2007 had a good amount of it: half an inch of moderate rain coming down during the day. The rain stopped by 3:00 pm, and the Sun began showing by 3:30 or 4:00 pm.

Did this rain cause the low turnout? Last time I checked, Quincy, Brockton, and Fall River all have weather highly-correlated with Boston; those cities still had around 50% turnout among registered voters for their mayoral elections. You can't tell me that half the people stayed home in Boston, Quincy, Brockton, and Fall River because of the weather... unless the latter three can expect voting rates of 100% -- or more -- when the Sun is shining. Yeah, right.

More likely: voter turnout was depressed a bit by bad weather, but mostly by the blizzard-like-whiteout of horrendously inadequate coverage of the race and issues by the city-wide papers, the Boston Globe and the Boston Herald. Those newspapers could have written their campaign coverage via computer program: IF (not cover the campaign && rainy weather) THEN (write column about low voter turnout on election day and a news story the day after). I wonder if all their news coverage could be coded up like this.

Memo to Globe and Herald city news desks: you got the election you asked for. Brian McGrory is my candidate for local news editor of the year, since he appears to have been the most powerful influence in the election via his reporters' non-coverage in the race.

Allston-Brighton Residents Don't Vote

A staple of Boston election politicking is that Allston-Brighton residents don't vote. You can explain it in many different ways, primarily through the transient nature of much of the A-B population, but the bottom line result keeps coming up. A-B Wards 21 and 22 (which includes Ward 21, Precincts 1 and 2 which are not part of District 9) voted in the 2005 municipal election at 23.6% of registered voters, while the city as a whole voted at 35.6%. In the 2005 preliminary municipal election, Wards 21 and 22 voted at 11.5% while the city voted at 15.1%. Conventional Wisdom of city politics dictates that city-wide candidates shouldn't spend much time campaigning in A-B -- or delivering constituent services -- because the residents don't vote enough to make it worthwhile.

In the 2007 election, we have all learned a lesson: Allston-Brighton electoral interest is resurgent. For the first time in recent memory, voters in Wards 21 and 22 turned out at a rate of 12.6% -- a rate nearly matching the city-wide average of 13.6%.

Why? The open A-B District 9 City Council seat likely motivated voters here to turnout for their favorite between Mark Ciommo and Greg Glennon. This local race also brought to the fore important neighborhood issues on many voters' minds, particularly those of institutional expansion and the exodus of families from the area. The A-B turnout has its silver lining: A-B residents are showing that they want their neighborhood issues dealt with during the next two years. The City Council and Mayor are bound to take note.

Councilor Michael Flaherty So Encouraged by 2007 Finish That He Will Run for Mayor in 2009

Hold your horses! In 2005, Councilor Flaherty got 49,220 votes to take first-place in the Councilor-At-Large race, thereby beating second-place finisher Councilor Felix Arroyo by more than 5,600 votes. This time, Councilor Flaherty got only 25,847 votes, beating second-place finisher Councilor Stephen Murphy by only 2,206 votes. The difference between first and fourth place in 2007 was only 3,867 votes: the first four candidates ran pretty much neck-and-neck.

On a positive note for him, Councilor Flaherty managed to get a vote off of 55.9% of ballots cast in 2007, up from 50.7% in 2005, in a Councilor-At-Large race that allows voters to cast up to four votes. Nonetheless, in 2005, Mayor Thomas Menino got a vote out of 65.9% of the ballots cast (including blank ballots) in a Mayoral race where voters do not have the opportunity to cast more than one vote. Without a massive voter revolt over the horizon, it's hard to reconcile Menino's and Flaherty's numbers to make it look like the latter would have any chance in a head-to-head competition.

While Councilor Flaherty was bringing in the big bucks for his fundraising in an election year, Mayor Menino decided to do a little bit of his own fundraising in an off-year. During September and October, Mayor Menino raised $229,890; in the same reporting period, Councilor Flaherty only brought in $147,480. Does Councilor Flaherty have piles of his own money to throw into his own race like Maura Hennigan did in her failed 2005 Mayoral bid?

Councilor Flaherty needed to win big on November 6th to make his Mayoral aspirations look credible, but he only won marginally, not overwhelmingly. And the Mayor flexed his fundraising muscles to make it clear.

Bullet Voting in West Roxbury

John Connolly's campaign sent out two anonymous mailings attacking Councilor Stephen Murphy in the last week before the municipal election. There were three additional, anonymous mailings that were sent out around the same time (according to the Boston Globe's summary), but which nobody has stepped out to take responsibility for. Two of those were sent to voters in the Parkway neighborhood (i.e., West Roxbury) lobbying for "bullet votes" for a local candidate, presumably Connolly; one of those mailings also included the bulk mailing permit number previously used by Connolly's campaign.

Did Connolly's lobbying for West Roxbury bullet votes swing the election to him?

Ward 20 includes West Roxbury as well as some of Roslindale; Connolly was born in Roslindale but currently resides in West Roxbury. Ward 20 historically turns out to vote at rates well above the city-wide average: in the 2005 municipal election, the Ward voted at 48.4%, compared to the city-wide rate of 35.6%; in the 2007 municipal election, they turned out at 23.8% compared with city-wide 13.6%. In 2007, Ward 20 was therefore disproportionately voting at a higher rate even than is usual even for them. This is exactly the behavior expected in a low turnout election, where neighborhoods with white/elderly/wealthy residents turnout at higher rates than minority/young/poor neighborhoods.

What did this disproportionately high turnout in Ward 20 mean to the Connolly vs. Councilor Felix Arroyo race for the fourth Councilor-At-Large seat? If Ward 20 had voted at 18.5% in 2007 (matching their usual higher turnout) instead of 23.8%, Connolly would have lost only 635 votes out of his 3421 margin of victory over Arroyo. Ward 20's disproportionately high turnout in 2007 was not nearly enough, by itself, to elect Connolly.

Did these Ward 20 voters engage in bullet voting? City-wide in 2007 there was an average of 2.7 votes (out of an allowed four votes) for City Councilor-At-Large; in Ward 20, the average was 2.5 votes per ballot. Compare this to the 2005 municipal election which had 2.9 city-wide votes per ballot, versus 3.0 votes per ballot in Ward 20. While city-wide voters did a little bit more bullet-voting in 2007 than in 2005, West Roxbury and Roslindale voters did substantially more bullet voting in 2007 than in 2005.

Did this bullet-voting make the difference? We don't know for sure, because you would need to know how each and every ballot voted. But we can take two cases: an extreme case, where all the bullet votes took a vote away from Arroyo (unlikely); and an average (and likelier) case, where all the bullet votes took a vote away from one of the other four candidates, taken at random. In the extreme case, all of Ward 20's 0.5 fewer votes per ballot in 2007 than in 2005 removed 3142 votes from Arroyo, almost all of the margin between him and Connolly. In the more likely case, only 3142/4 = 786 votes were taken away from Arroyo, not nearly enough to produce the difference between Connolly and Arroyo.

OK, so the numbers back up that West Roxbury and Roslindale utilized the bullet vote more in this election than previously. But are they the masters of the bullet vote? No! Ward 6 (South Boston, Councilor Flaherty's ward of birth) averaged 2.3 votes per ballot, Wards 1 (East Boston), 7 (South Boston, Councilor Flaherty's home ward), and 22 (Allston-Brighton) all averaged 2.5 votes per ballot, similar rates of bullet voting to the West Roxbury/Roslindale recent converts.

What do we make of all this? Yes, Ward 20 (West Roxbury/Roslindale) voted at a disproportionately high rate and used significantly more bullet voting than in previous elections, but neither effect was sufficient to explain how Connolly beat Arroyo for fourth place. The combination of the two effects only explains around 40% of their margin differentiating the two.

Look elsewhere for bigger causes of Connolly's defeat of Arroyo, such as: Arroyo's failure to raise money; Arroyo's failure to get-out-the-vote; the mainstream media's failure to cover the race resulting in low voter turnout; etc. Those will likely be fingered as the real factors leading to Arroyo's ouster at the polls in 2007.

No comments: