It appears, looking at the numbers, that Patrick simply dominated in Western Mass., especially compared to Kerry’s performance there in ’04. Tiny towns in Franklin and Berkshire Counties such as Heath, Colrain, and Alford, as well as (slightly) larger places like North Adams and Orange saw significant swings towards Patrick (as compared to John Kerry’s performance here).
In Orange, for example, Kerry barely bested Bush in ’04, winning 50.1% to 48.6% — much worse than Kerry’s statewide victory of 62% to Bush’s 37%. Yet Patrick managed to nearly mirror his statewide average here in ’06, 55.7% to Healey’s 34.9%. Other cities and towns in the western part of the state were similar.
Where did Patrick lag behind? In the small to mid-size blue-collar cities mainly in the east — places like Revere, Quincy, Everett, and Saugus. For example, Kerry crushed Bush in Revere, 66.8% to 32.6%, a bit better than Kerry’s statewide numbers. Patrick, however, bested Healey by the more modest 52.3% to 38%, a bit worse than his statewide average. In Quincy, where Mihos did his best — getting about 27%(!) of the vote — Patrick outpaced Kerry 41.7% to 29.9%, far worse than Kerry’s 63.9% to Bush’s 35.4%.
Other places of interest — Patrick did especially well in Worcester, winning 69.0% to 23.2% (compared to John Kerry’s 67.8-31.3% margin), and pretty much ran even with Kerry’s performance in the large Democratic strongholds of Boston, Somerville, and Cambridge.
What does it all mean? Well, if we compare it to what we would “expect” given John Kerry’s performance in ’04, it seems Patrick’s message especially resonated in the western part of the state, but not as much in some of the smaller cities in the east. In these latter locations, however, it seems Mihos generally did his best — suggesting that those John Kerry voters not voting Dem in ’06 ended up going with Mihos rather than Healey. I had thought that perhaps the conventional wisdom was a bit off, and that Mihos would pull more voters away from Patrick rather than Healey — and these numbers suggest that this might be the case.
I have the Excel file with the numbers, though something like that is a bit difficult to post in its entirety. (Update by Charley: Here it is.) In any case, if you’re interested in any particular cities or towns, or would like to know more about the methodology in general, I’d be glad to share/post the numbers — and especially to get your thoughts on all of this.
Update by Charley: Here’s Hoyapaul’s notes on the spreadsheet: “I realize it may not be the most user-friendly thing ever, but most of the categories should be fairly clear except for the color-coded ones. Here’s the explanation for those:
The PVI columns, for Healey, Patrick, Bush, and Kerry, represent the total % better the named candidate did than the statewide average. Taking Abington as an example, Patrick’s 2006 PVI was negative 11.25%, because the 44.3% he received in Abington was 11.25% worse than his statewide average of 55.56%.
Likewise, John Kerry’s PVI for Abington was negative 7.6%, representing how much worse he did in Abington (54.5%) than his statewide average of 62.11%.
The three columns on the far right of the Excel sheet make an effort to compare the Patrick/Healey performance in ’06 with the Kerry/Bush performance of ’04. The “Bush-Hea” column simply subtracts Bush’s 2004 PVI from Healey’s 2006 PVI. The “Pat-Kerry” column does the same for the Democrats, subtracting Kerry’s 2004 PVI from Patrick’s 2006 PVI. A red number indicates that Healey outperformed Bush in that area (or Patrick underperformed Kerry), and a blue number indicates that Patrick outperformed Kerry (or Healey underperformed Bush). The “Tot Move” column attempts to show the TOTAL movement that city/town had away or towards the Democratic or Republican candidate. The more positive the number, the more movement towards the Republican in ’06; the more negative the number, the more movement towards the Democrat in ’06.
And I’ll put up the Excel sheet.
I’d love to see the full numbers. A few simple observations:
<
p>
-I think the third-party options for this Guv. election were more substantiative when compared to the presidential options in ’04.
<
p>
-Worcester picked up 70 percent thanks to Tim Murray. Those who aren’t from the area can’t usually understand the draw, but he is a favorite son here like no politician in my lifetime or the lifetime of my parents. Deval would have done well here regardless, but the extra ass-whuppin’ was provided by our Boy Mayor. It certainly made the area easier to work as a volunteer, I encountered plenty of folk who were lukewarm on Patrick but had the love for Mayor Murray that I do.
<
p>
As shown above, Central MA was the battleground, according to the state pundits. There’s a definite reason why we won it beyond the Patrick charm, and it’s one of the main reasons I wanted Tim on the ticket (besides his ample qualifications). Trust me, there was no small amount of folk who voted the ticket because of Tim Murray. If you think this was unimportant, look at the 2002 results for the county. This year we CRUSHED, even taking Reed Hillman’s hometown of Sturbridge.
<
p>
Absolutely, and this explains why there were cities and towns in which Patrick AND Healey both “underperformed” as compared to Bush/Healey — some of each’s votes went to Mihos (and Ross). The two main third party candidates in 2004 were Badnarik (Libertarian) and Cobb (Green), both of whom pulled very few votes from the Democrat and Republican. In contrast, Mihos/Ross pulled away about 9%.
<
p>
I agree that Murray’s presence on the ticket helped in central MA. Though not by overwhelming margins, Patrick outperformed John Kerry in places like Worcester and Sturbridge. This boost from central MA, along with the massive shift to the Democrats in western MA and the fact that Patrick did about as expected in the eastern part of the state, all equalled the blowout we saw last week.
and it affected the surrounding towns as well. That’s why we put “Deval Patrick and Tim Murray” in all of our phone bank scripts around here, and emphasized Tim Murray’s “familiarity with issues facing Central Massachusetts.” All of the canvassing and phone banks after the primary were directed toward surrounding towns and other areas in Worcester county; none in Worcester itself, due to a strategic decision made by the campaign. We did do GOTV in Worcester on Tuesday. In 2002, all of the towns in Worcester County except Worcester went to Romney. Patrick and Murray prevailed in all of them by a good margin. So the strategy worked quite well, along with all of the other factors in play this year.
Are you looking just at % of the vote, or also at how many people came out? You get a different population who votes in a presidential election than votes in a gubernatorial election.
I’m simply comparing total % of the vote. I recognize that all of this should be taken with a grain of salt, as you suggest, since I’m comparing a Presidential year with a mid-term election. (It will be quite interesting to compare ’08 numbers with ’04 to get to political trends within individual municipalities, by the way). Nevertheless, given that all of our Gubernatorial elections are during mid-term years, I think it’s still an interesting and maybe worthwhile comparison.
<
p>
Interestingly, about the total votes cast in this election equalled about 76.7% of the total votes cast in 2004, which was a bit higher than I expected. Turnout tends to drop off significantly from Presidential years, but in MA it didn’t drop off too badly. It’s hard to say how the fact that this was a midterm election affected this race. It seems that MAYBE Patrick did a bit better in those districts where there was slightly lower turnout than in ’04, but it’s really all over the map so I’m not sure what conclusion could be made there.
say, compared to the last Gov election, that’d be swell to see. One thing that so many folks including me found quite inspiring in Patrick’s campaign is his ability to get loads of people to “check back in” as he calls it.
<
p>
So how many more checked back in as compared to the last Gov election as measured by turnout? (I was really sick all last week so it might have been in the papers but I missed it). Thanks.
If only someone had looked at turnout patterns. That would be interesting, wouldn’t it? đŸ˜‰
thanks
and analysis. Thanks for taking the time to break that down and post it. To me, it suggests that some of the arguments made during primaries about the importance of appealing to independents rest on an exaggerated sense of the unenrolled vote. (But that’s probably a moot point, anyway, as i don’t expect that we’ll be seing a hotly contested open primary next time around…)
That John Kerry won in 2004. That’s 1/3 of the towns, and the bulk of the population, as Patrick took the sparsely populated outer Cape towns, while Kerry Healey took Barnstable.
Didn’t Mitt win all the cape except Provincetown?
….I don’t think he won anything east of Eastham.
<
p>
And I thought we were comparing Patrick and John Kerry.
Here’s my home blog post on Cape numbers and that when you add the total count of votes up in ’06 versus ’02 for the Governor’s race, there were only about 21K more that voted this year, in raw numbers.
http://www.capecodto…
<
p>
Indeed, Romney did win everything from Eastham, back to the bridge, and Nantucket, but not the Vineyard. Overall, you went from Romney winning the Cape by 15 to Deval winning it by 5.
So he and Kerry both took the mid-Cape.
Can you email me the file to statehouseguy@yahoo.com
<
p>
I am doing a little project as well. I would be happy to talk to you about it in a less-public forum.
<
p>
Thanks
We were interested in the below 128 numbers. No one ever mentions the cluster of towns along I-95 down to the RI border. My area fought hard for Deval with some success in getting a few towns to turn blue. Any help in doing some research is appreciated.
Don’t forget, you’re not comparing a 2 man race to a 2 man race… so your PVIs may have to be normalized somehow. Consider this — which is stronger:
<
p>
Scenario A:
Joe Dem 50
Sally GOP 30
Bob Indyy 20
<
p>
Scenario B:
Joe Dem 60
Sally GOP 40
<
p>
In both cases, Joe Dem won by 20 points, but its not clear to me that both scenarios are just as partisan.
I can’t wait to see the numbers. However, when I was doing my own analysis on Arlington precincts, I steered away from the 2004 Presidential. There was a tremendous draw of people (at least in my part of the world) with pure passion to vote Bush out of office, and the lines at the polls were significant. The anti-Bush vote, along with this being John Kerry’s home state, makes it an optimal year.
<
p>
I like baselines where people revert to form. The 2002 Governor race and the 2000 Presidential had no incumbents, and I think are better baseline elections.
…but I remember 2000 as being pretty anti-Bush here too.
I’ve posted the results from the Berkshire, Hampshire & Franklin Senate District, including a comparison with 2002, on a town-by-town basis.
<
p>
I did not include the PVI (since I hadn’t heard of it), but if time permits maybe I’ll go back and add it in.
<
p>
Thanks for mentioning Alford!
<
p>
While I agree that “Patrick’s message especially resonated in the western part of the state” I also believe the results had a lot to do with where we had an active grassroots presence. At least, that is true within my District — in towns where we did not have an active local organizer, the results were generally sub-par. We had a huge network of volunteers working hard in the western four counties, and N2N had a big presence in the Pioneer Valley as well. I have to believe that had a lot to do with the outcome, though since I’m not really sure what was going on in some of the places you mention, it’s hard to be sure that was the critical variable. My guess is that it was.
Hoyapaul for the large overview and MFW for the local breakdown. Both very illuminating.
<
p>
As a town coordinator in Williamstown, I find our 5% increase over last time mildly disappointing. We always vote at a high rate (my rough sense–without the precise numbers at hand–is that typically about 70% of registered voters show up), which of course makes large percentage jumps more difficult. But, still, I think we could have amped up our turnout by at least another 5% (as I think we had a better surge for the primary). Here, in the final two weeks, it was a bit hard to fight the confidence in victory. But I’ll take our 80% support for Patrick any day.
Good work, and good analysis, Hoyapaul. I would point out that there is a discrepency with Mihos’ numbers in Quincy. The Globe and the local paper, The Patriot Ledger show different results for Mihos:
QUINCY RESULTS
Healey
Mihos
Patrick
Ross
<
p>
The Globe
10,684 (30%)
9,479 (27%)
14,914 (42%)
687 (2%)
The Patriot Ledger
10,684 (38%)
2,479 (9%)
14,914 (53%)
n/a
<
p>
<
p>The Ledger’s numbers for Mihos are probably the accurate numbers, if only because the 9% number is closer to the rest of the state. (The Globe made a similar mistake after the primary when it pronounced Lakeville as Tom Reilly’s best town; he finished last.)
since 27% seemed far too high for Mihos. Correcting the information indicates that Patrick still underperformed in Quincy as compared to Kerry, but by not as much of a margin.
<
p>
The only other town that jumped out at me as having incorrect numbers in the Globe’s count was my hometown of Holliston, which was the only town that was a true outlier as far as turnout went — while in most towns the turnout compared to ’04 was high-60s to low 80s, Holliston’s turnout (according to the numbers) was only 19% compared to ’04. Obviously this is not correct, but it’s the only other town that the numbers looked off.
<
p>
Thanks for the updated info on Quincy.
How about the town of Montgomery, which was reported as 35% Healey, 32% Patrick, 29% Mihos? That one seemed odd to me since Mihos didn’t see anywhere near that popularity anywhere else (he was mostly single-digits, with a couple of 12-14% towns). Sure, there were just around 500-600 votes cast, but that is a pretty big outlier.
Good catch..I saw the 29% number and figured that (unlike Quincy) maybe this was because Montgomery (a Republican-leaning town, which gave Bush 53%) was so small and thus the numbers easier to skew.
<
p>
However, after looking at it again, I realized that if the vote count was accurate, fully 97% of the voters in ’04 returned to vote in ’06, far higher than the statewide percentage. I think the problem is that the Globe misreported Mihos’ vote total as 134, rather than just 34. If that’s the case, Montgomery’s turnout would be the more average 76%, and Mihos would have received the much more reasonable 9.5% of the vote.
I was questioning that Montgomery result on election night, as I was looking through the results from greater Westfield. If you check the similar surrounding towns (Blandford, Russell, Chester) the numbers would parallel what you suggest as the correct result.
and wondered how Gabrieli could only get 3 votes in the Town. The Globe had it as his worst, if I remember correctly. I couldn’t imagine that with young Steve Canessa working for him he couldn’t get more than 3 votes in the town. After all, he plus his parents and sister make 4!
There is no way those Globe numbers made any sense. I live in the Q, and there was no Mihos presence and are not that many Greeks.
The more appropriate comparison is probably with 2002, when the same offices were mostly available. The 2004 Libertarian presidential campaign, in the state, had several hundred lawn signs, and a Libertarian State Committee (now thankfully out of office) that viewed its mailing list (not its donor list, its mailing list) as a state secret. The 2002 campaign for Governor had more like a hundred thousand dollars, not to mention the income tax elimination referendum that close but not quite passed.
After the discussion of the Quincy and Montgomery numbers above, I crunched all of the race data available from the Globe, comparing the total number of votes on all six statewide races.
<
p>
Montgomery is the worst offender, percentage-wise.
<
p>
Quincy is wrong, as noted.
<
p>
Petersham has Healey numbers that are too low; perhaps 242 instead of 42?
<
p>
Holliston is missing 5000 votes for both Governor and Senate, but the proportions look believable. I’m guessing that the AP only has one of the four precincts.
<
p>
With respect to other races:
<
p>
AG: Coakley is low by about 1000 in Mattapoisett, and Frisoli is low by about 3000 in Braintree.
<
p>
Auditor: DeNucci is about 6000 votes too high in Malden.
<
p>
Senate; as noted above for Holliston
<
p>
Secretary and Treasurer look like any mistakes are minor.
<
p>
The only other town that was a statistical outlier was Dover; upon closer examination it appears that there were a lot of blanks in the three races with no Republican candidate.