Bob LeLievre's Blog – "The" source for Massachusetts election data

Here  is my summary of the Boston 2009 General election results.  I include analysis of the turnout, Mayor’s race, and at-large City Council race.  I summarize the results from a neighborhood perspective, not by wards, since almost everyone can identify with a neighborhood name more easily than a ward number.

The Turnout:

Here’s a link to the turnout counts and rates by neighborhood:

http://spreadsheets.google.com/pub?key=tizoE2JMGeIrVUBMilOXerw&single=true&gid=0&output=html

- The overall turnout was 111K voters.  This is 39% of active voters and 31% of all voters (active + inactive).  This is a 29K increase (36%) over the 2009 Primary, and a 14K increase (14%) over the 2005 General.

- Turnout had the usual distribution of whiter neighborhoods turning out better than liberal-voting and non-white neighborhoods.   Turnout was highest in Readville (62%), southern-white Dorchester (56%), and West Roxbury (54%).  Turnout was lowest in Allston (21%), Fenway (25%), and Back Bay (27%).  Non-white neighborhood turnout was in the mid- to high 30s%.

- There was a 10 percentage point increase in turnout from the Primary to the General (39% vs. 28%).   The increase was larger in the whitest neighborhoods.  The biggest increases were in Readville (+17 points) ,West Roxbury (+15 points), and Charlestown,  Beacon Hill, southern-white Dorchester (all +13 points).  The lowest increases were in Chinatown (+5 points), Allston (+6 points), and Brighton, Grove Hall (+8 points).

• This is somewhat surprising, since there is generally a “liberal/non-white boost” in the General compared to the Primary.  I expected the increases in the white-liberal and non-white neighborhoods to be relatively larger.   Maybe Flaherty was better at turning out his base here, and/or maybe Menino was trying to pull out more votes in the white neighborhoods???
• The South Boston (Flaherty’s home) change was up 10 percentage points to 45%, only 6 points above the city average.  The Readville (Menino’s home) change was 17 points to 62%, or 23 points above the city-wide average.   Perhaps some South Boston voters were not too excited by the Flaherty-Yoon alliance and stayed home???

- The chart has columns for both percentage point change and percent change for the 2009 General vs. 2009 Primary and the 2009 General vs. 2005 General.  To clarify – a neighborhood with 1000 voters that goes from a 10% turnout to a 15% turnout would have a 50% increase and a 5 percentage point increase.  A different neighborhood also with 1000 voters that goes from a 50% turnout to a 55% turnout would have a 10% increase and a 5 percentage point increase.   I like using the percentage point number because in the above 2 examples, the increase in actual voter turnout is the same (50 voters) and  the percentage point change is the same.  But if you look at only the % turnout change (50% vs. 10%), you might think there is a much bigger change going on in the lower-turnout neighborhood.

- This was the most contested mayor’s race in Boston since 1993, when Menino was elected to an open seat after Ray Flynn left for the Vatican.  The turnout was 112K in the primary and 118K in the general.

The Mayor’s Race:

Here’s a link to the mayoral candidate results by neighborhood:

http://spreadsheets.google.com/ccc?key=0ArYPkyBtYlUKdC1ocG5lWS1BRTI1dkx6YU82bjZJdXc&hl=en

- Menino beats Flaherty 57% to 42%, a gap of 15 points.  The gap in the primary was 26 points (50% vs. 24%).  So Flaherty closed the gap by 12 points.  I’ll explain below how did he did that.

- Menino did best in his home turf of Hyde Park and Readville with 75% of the vote.  He also did well in  the least-white neighborhoods like Mattapan and parts of Dorchester (90+% non-white population) with 70% – 75% of the votes there.  He also did well in large neighborhoods like Roxbury (67%), Roslindale (62%), East Boston (60%), West Roxbury (58%),

- Flaherty was strongest in the whitest neighborhoods.  Outside of South Boston (69%), he got 50% – 55% of the vote in Charlestown, and the white parts of Dorchester.   He got less than 20% – 30% in the least-white neighborhoods.  He also did very well in the white, liberal-voting neighborhoods like Jamaica Plain, South End, Back Bay, and Beacon Hill with 50% – 55% of the vote.

So what happened between the Primary and the General?  Where did the Yoon votes go?

• Menino gained 6.6 percentage points (57% vs. 50%) while Flaherty gained 18.3 percentage points (42% vs. 24%) in the General vs. Primary.   So Flaherty improved by 12 percentage points more than Menino city-wide.
• The best improvements for Flaherty vs. Menino came from white-liberal voting neighborhoods like the Fenway (26 points), Jamaica Plain (23 points), Allston and Back Bay (22 points), Beacon Hill (20 points), and the South End (18%).   These neighborhoods are where Yoon did best in the Primary.
• The worst improvements for Flaherty vs. Menino came from Readville (0 points) and Hyde Park (4 points) which are Menino’s home neighborhoods, West Roxbury (2 points), South Boston (2 points), and southern-white Dorchester (4 points).
• The non-white neighborhoods came out in the middle for Flaherty vs. Menino, in the 10 – 16 point range.

So it looks like the Flaherty-Yoon ticket idea had the result of:

• Boosting Flaherty a lot in the white-liberal-voting neighborhoods by giving better him “liberal credentials”.
• Didn’t make much difference in the non-white neighborhoods.
• Hurting Flaherty in the white-conservative-voting neighborhoods.
• My experience is that when a candidate tries to expand beyond the base, he/she risks alienating that base, which often results in lower-than-expected turnout and/or % of the vote in the base.
• The 29K new voters in the General vs. Primary came somewhat more from white-conservative-voting neighborhoods than I expected.   If I had to guess, I’d say it was the Menino machine pulling out voters there, not the Flaherty machine.  Any insiders know the strategy?

The City Council At-Large Race:

Here’s a quick summary of the results.  The top 4 candidates are elected.

City Council  At-Large Candidate	2009     General   Votes	2009 Primary Votes	2009 General  % Vote	2009 Primary % Vote	Candidate Race/Ethnicity	Home    Neighborhood
Connolly	51,308	35,115	46%	43%	White – Irish	West Roxbury
Murphy	50,962	30,311	46%	38%	White – Irish	Hyde Park
Arroyo, Jr.	45,099	25,821	41%	32%	Latino – Puerto Rican	Jamaica Plain
Pressley	41,847	16,846	38%	21%	African-American	Back Bay
Jackson	30,173	12,520	27%	15%	African-American	Roxbury
Kenneally	24,215	12,632	22%	15%	White – Irish	East Boston
Gonzalez	18,297	10,108	16%	12%	Latino – Puerto Rican	Hyde Park
Bennett	16,817	10,519	15%	13%	White	West End

- These percent results are different from what is published by the Election Department and the newspapers.  The way they do it is:  100 voters cast 300 votes (a voter can vote for up to 4 candidates) for various candidates.  Candidate X gets 30 votes.  The city calculates it as Candidate X got 10% of the vote (30 of 300 votes cast).  I calculate it as Candidate X getting 30% of the vote (30 votes from 100 people turning out).  My method is the way it gets calculated in all other elections, so it’s much more relevant when comparing with other election results.

- The order for the top 4 candidates stayed the same from the Primary to the General.  Connolly beat Murphy by only 346 votes for first place, after beating him by 4,804 votes in the Primary.  The only order changes were:   Jackson went from 6th to 5th, Gonzalez went from 8th to 7th, Kenneally from 5th to 6th, and Bennett from 7th to 8th.

- Pressley had the best improvement (17%) from the Primary to the General.  Jackson was next with a 12% jump.  Arroyo, Murphy, and Kenneally were next with 7-9% jumps.  Connolly, Gonzalez, and Bennett were up only 2-4%.

As in the past few elections, the 4th place finisher gets in the high-30%s  of the vote.  That seems to be the magic number to win.

Results by Neighborhood:

Here’s a link to the candidate results by neighborhood:

http://spreadsheets.google.com/pub?key=t3l6oAMckxpzAP4PIhuYtLw&single=true&gid=0&output=html

As usual, white candidates do best in the whitest neighborhoods and non-white candidates do best in white-liberal and non-white neighborhoods.  The % white voting-age-population (from the 2000 Census) of the neighborhood is included in every chart to make this clear.  So it’s often the relative turnout of each neighborhood that plays a major role in the final results.

You can look at the chart at the link above to see all the details.  The precinct-by-precinct results are also available at a link below.  Here are some general observations:

• Connolly’s best results (46% citywide) came from West Roxbury (66% – his home neighborhood), Charlestown (63%), southern-white Dorchester (60%).    His worst results came from Roxbury, Grove Hall, Blue Hill Ave. / Washington St. Corridor (17% – 19%).  He also got mid-50s% in white neighborhoods like West End, Brighton, Beacon Hill, and Back Bay.

• Murphy’s best results (46% citywide) came from Readville (69%), south-white Dorchester (63%) and South Boston (61%).  He did worst non-white neighborhoods like Roxbury, non-white parts of Dorchester (25% – 28%).  Murphy did a lot better than Connolly in these non-white neighborhoods.

• Arroyo’s best results (41% citywide) came from Jamaica Plain (67%), Fenway (52%), Roxbury (51%), and Grove Hall (50%).  He did worst in South Boston (20%), southern-white Dorchester  (27%) and Charlestown (28%).

• Pressley’s best results (38% citywide) came from Melville/Ashmont (58%), Mattapan (54%), Roxbury (50%), Blue Hill Ave. / Washington St. Corridor (50%).   She did worst in East Boston (15%), South Boston (18%), Readville (20%), and Charlestown (23%).

• Jackon’s best results (27% citywide) came  from Roxbury (56%), Mattapan (53%), non-white parts of Dorchester (51% – 56%).  He did worst in South Boston (9%), East Boston (10%), Readville (11%), and Charlestown (12%).

• Kenneally’s best results (22% citywide) came from West Roxbury (44%), East Boston (43%), and South Boston (40%).  He did worst in Mattapan, Roxbury, and non-white parts of Dorchester (5% – 9%).

• Gonzalez’ best results  (16% citywide) came from Jamaica Plain (28%), Mission Hill (26%), and Hyde Park (25%).  He did worst in Charlestown (9%), South Boston (9%), West Roxbury (10%), and southern-white Dorchester (10%).

• Bennett’s best results (15% citywide) came from West End (39%), South Boston (29%), Charlestown (27%) and the North End (27%).  He did worst in Jamaica Plain (8%) and least-white neighborhoods like Roxubry, Mattapan, non-white Dorchester (5% – 8%).

Candidate Result Changes from 2009 Primary to 2009 General:

Here’s a link to a chart showing how each candidate’s vote changed from the Primary to the General:

http://spreadsheets.google.com/pub?key=t3l6oAMckxpzAP4PIhuYtLw&single=true&gid=3&output=html

- Example:   if a candidate went from 30% in the Primary to 40% in the General, there was a 10 percentage point increase.  For the 8 candidates, the overall increases ranged from +2 points to +17 points.   For individual neighborhoods, candidate results varied from -4 points to +31 points.

- The non-white candidates gained a lot more votes than the white candidates in the General vs. the Primary (+17, +12, +9, and +4 points for non-white candidates vs. +9, +6, +2, and +2 points for the white candidates).  This is the “liberal /non-white surge” that we often see.  It’s complicated to compare it to other years since there have typically been only 1 or 2 non-white candidates in the general.

- Some highlights:

• Pressley had big gains in Mattapan (31 points), non-white parts of Dorchester (22 – 25 points), Roxbury (23 points).  She also had big gains in neighborhoods like Jamaica Plain, Allston, Brighton, and Hyde Park (21 – 22 points).
• Jackson had gains of 18 – 25 points in very non-white neighborhoods.
• I’d guess that the Jackson and Pressley gains were due to them having low name recognition in the primary as first-time candidates.
• Arroyo Jr. started out with better name recognition (thanks to Arroyo Sr.), so he did far better in the primary, and still had a decent 9 point gain in the General.
• Murphy’s 9 point gain is somewhat surprising since he started with good name recognition as a 12 year incumbent city councilor.  His gains were highest in the Back  Bay, Beacon Hill, Fenway and West End (14 – 16 points).  He also had a good gain in Mattapan (14 points).  He had the lowest gains in the white-conservative-voting neighborhoods.   I don’t have a great explanation for this.
• Of the 7 candidates who didn’t make it to the General, the 2 white candidates got 9,780 primary votes and the 5 non-white candidates got  29,994 primary votes.  So there were a lot more potential votes going to the remaining non-white candidates.
• For lots of reasons, the power of incumbency made it hard for the 2 incumbents, Connolly and Murphy, to lose under any turnout scenario.

Votes per Ballot (a.k.a. bullet voting):

- Here’s a link to the VOTES PER BALLOT tab with a chart that shows the average # of city council candidates (up to 4)  that a voter voted for, broken down by neighborhood.  Using only one of your four votes is referred to as “bullet voting”.

http://spreadsheets.google.com/pub?key=t3l6oAMckxpzAP4PIhuYtLw&single=true&gid=4&output=html

- A voter casts one ballot and can vote for up to 4 city council candidates in the at-large race.  In the past 10-20 years, voters typically use an average of about 3 of their votes per ballot.

- In the General, voters used an average of 2.5 votes per ballot.  This is up from 2.4 votes per ballot in the primary.   The neighborhood pattern in the General is similar to that in the Primary, where the whitest neighborhoods are a few tenths of a vote above the city average and the least-white neighborhoods  are a few tenths of a vote below the city average.  The white-liberal-voting neighborhoods are also a few tenths above the city-wide average.

- Bullet voting actually works against non-white neighborhoods’ clout.  There are less votes per voter coming out of those neighborhoods, compared to the whitest neighborhoods.  Given that there were 4 strong non-white candidates, there was no good reason for this to happen.   A neighborhood using an average of 2.3 votes per voter vs. a neighborhood using an average of 2.7 votes per voter has the same affect as decreasing the turnout by 15% (.4 divided by 2.7), or for example, reducing a turnout of 40% effectively to 35%.

Strategy for future City Council candidates:

The opportunity is there for future open seats to be won by liberal / non-white candidates.  The winning strategy for them is:

• Lots of quality candidates producing very-contested elections that generate a high city-wide turnout.
• Concentrate campaign resources in the liberal-voting / non-white base to generate high name recognition and turnout.
• Discourage bullet voting.  Encourage voters to use 3 or 4 of their 4 votes per ballot.
• Ongoing demographic and voter registration changes in the city are shrinking the white-conservative base.
• High-turnout elections in even-year state and federal elections increase registered voter count and increase the pool of frequent voters.

Get all the raw data here:

- In the interest of making public records more public and elections more transparent, here’s a link to the raw precinct results in a spreadsheet format.

http://spreadsheets.google.com/ccc?key=0ArYPkyBtYlUKdFVCVUtraVpJVTl4dERoSXY2cWtPVEE&hl=en

- Here are the raw results from the Boston Election Department, as of 11/4/2009:

http://www.cityofboston.gov/elections/currentelections/pdfs/MunicipalElection3Nov09UnofficialResults.pdf

Notes:

- % Turnouts are based on the active voter counts, unless otherwise noted, instead of the total voter count (voter status = active + inactive).  The inactive voters are usually 20%-25% of the total voter count and represent voters who are mostly gone, but for a variety of reasons, can’t be officially purged from the voter list for 3 years.  If I included the inactive voters in the % turnout calculations, it artificially drives down the real % turnout, since the total voter count is padded with non-existent voters.  In Boston, the inactive voters as a group typically vote at one-tenth to one-twentieth the rate of active voters.

-  Numbers may not add up to 100% due to rounding issues.  Percentage point comparisons may appear off due to data after decimal point.

- Since Dorchester is by far the biggest neighborhood (66K total voters) and has diverse sub-neighborhoods, I also break out those sub-neighborhood totals.