I took the votes Kerry and Obama received in each city and town, and compared it to the overall margin of victory statewide to get the PVI. (The numbers came from the Globe’s cool 2008 and 2004 town-by-town breakdown.) Thus, while the town of Winthrop is R+1, this does not necessarily mean that it is a “Republican town” – Obama and Kerry won handily there – but that it was slightly more Republican compared to the state as a whole. These are the types of towns that Republicans (like Scott Brown) should be winning if they want to succeed in statewide elections – indeed, Romney won Winthrop by exactly the same amount he won by statewide.
I’ll note that off-year elections (i.e. for Governor) are not included in the PVI, only Presidential elections. However, I also looked at the non-competitive Kennedy-Chase race in 2006, the ’06 Governor’s race, and the Kerry blowout against Jeff Beatty in 2008. Interestingly, there was little difference between a town’s PVI in either year – a place like Worcester (D+6) tended to give the Democrat right around 6 points more than the state as a whole, regardless of off-year or presidential year status (at least in the Kerry/Kennedy Senate races). There are a few towns that buck this trend, a couple of which I mention below.
Bellwether Cities and Towns
The following towns tend to track the overall statewide results. Obviously the turnout next Tuesday will be much lower than a Presidential or gubernatorial election – but the relative performance of Coakley and Brown (as well as who they are turning out) should be captured by comparing the early reported returns with the town’s PVI.
Hull (D+0) – Not only is Hull consistently among the best bellwethers in all of Massachusetts, but it is in Plymouth County, where Brown should do well. If he’s trailing here, that’s a good sign for Coakley.
South Hadley (D+0) – One of the question marks will be what turnout will be like in Western Mass. South Hadley is a bellwether and also in Hampshire County, which is small in population but nevertheless an area where Coakley should rack up some votes.
Waltham (D+1) – Of course, Middlesex County will be key in this contest, like it always is in statewide races. Waltham typically tracks statewide results, with a slight Democratic lean. Melrose (R+1) is another Middlesex bellwether, though one with a very slight Republican lean.
Winthrop (R+1) matched the Romney/O’Brien results perfectly, and was very close to the state average in the last two Senate races – though it was slightly more for McCain (+3.5%) in ’08.
Wellesley (D+1) isn’t really representative in terms of its underlying demographics, but perhaps surprisingly it is pretty representative of Massachusetts as a whole in Presidential races. However, Wellesley is one area that displays a different dynamic in off years – Romney and Chase did about 8 points better than the state as a whole (Brown also represents a sliver of Wellesley). Sudbury (D+1) is very similar.
Chicopee (D+1) in Hampden County is another politically representative town, albeit one that has trended more Democratic in off-year elections.
Falmouth (R+3) – Barnstable County should be Brown’s best area, and Falmouth is the biggest swingy town in the county. If Coakley is winning by a decent amount in Falmouth, it bodes very well for her.
Westport (D+1) – Bristol County had awful turnout in the Democratic primary compared to other areas, so it will be a county to watch on Tuesday. Westport is also a bellwether, though interesting because it trends more Democratic in off-year elections.
Other key cities/towns to keep an eye on include Milton (R+0) , Quincy (R+1), and Fitchburg (R+1).
For anyone interested, the most Democratic town in the state is tiny Aquinnah on Martha’s Vineyard (D+29). The most Republican is Charlton in Worcester County (R+22), which gave Bush a whopping 65% in 2004.
This will come as no surprise, but Coakley must get a good chunk of votes from Cambridge (D+25), Somerville (D+20), Brookline (D+18), and Boston (D+16). If turnout is reported as particularly low in these areas, or Coakley isn’t hitting at a margin at least as big as the D+ number here, she’s in trouble. Conversely, one of the benefits of the Obama visit will be to activate some of these liberal base voters. If turnout is OK here, and she’s exceeding her targets (say over 70% in Boston, with decent turnout), that’s a good sign.
Worcester (D+6) – is both a population center and has been very consistently at about D+6 the past several years. If Brown is close to even in Worcester, look out.
Most of Plymouth and Barnstable Counties should be Brown’s best counties, so track the turnout there. Of the decent sized towns, watch places like Hanover (R+17), Middleborough (R+15), Bridgewater (R+13), and Barnstable ((R+9). Outside of these two counties, also keep an eye on Walpole (R+13), North Andover (R+12), Weymouth (R+7). If Brown is getting his numbers in these areas, he has a good chance of the upset.
Another is Plymouth (R+9) – if Coakley running even or better in Plymouth, she’s in very good shape.
Both candidates were raised in towns that reflect their partisan leanings. Coakley was born and raised in North Adams (D+14), and Brown’s hometown is Wakefield (R+6). Of course, he represents Wrentham (R+14) now as a State Senator, where he should be beating Coakley handily.
… this is great stuff. Really great.
I’ve finally figured out how to post the underlying data in Excel format, if you’re interested. The link as at the top of the post.
Plymouth also has a ballot question on Tuesday regarding the local option meals tax. While that won’t generate as much interest as a general Prop. 2 1/2 override, it will bump up turnout, and not in a way favorable to Coakley.
influence in Plymouth. Should that not counter some of Brown’s base?
but it doesn’t always translate into actual voters. We have a lot of angry white guys and a lot of apathy down here. It’s one of the reddest areas of Mass. I will say that Sen. Murray is doing her damnedest and she’s working closely with the unions who love her.
p>I agree about the restaurant tax vote. It will bring out the Republicans in force. They hear the word “tax” and they turn out big time. All the restaurant, golf course, hotel owners will be pushing their workers to get out to vote against it.
p>If Coakley wins Plymouth, I’ll be on cloud 9, but very surprised.
Yes, if Coakley wins Plymouth, then Coakley will not only win, but win in a landslide compared to what the polls are saying.
p>I’m looking for her to keep it within 5 points. That would be a good sign in the heart of Brown country.
This is excellent analysis and data – I don’t have any time to do this sort of thing and just wanted to say “Thanks”.
one correction, though: Brown would actually have to do much better in the R+ areas or Coakley would have to do much worse in her base areas, if Brown is to pull off the upset, because those are the numbers Republicans were getting in the Republican-base towns when they were getting stomped across the state. Additionally, if either party does control the areas that tend to be split evenly, that could be enough, too.
there may be some of this effect going on, though I was surprised that the PVI numbers are pretty close even when compared to the 2002 Governor’s race (which was closely contested).
p>For example, Walpole is R+13, and Romney won with 64% in Walpole (compared to 50% statewide, for a R+14 PVI in 2002). North Andover is R+12; Romney recieved 65% there. Barnstable is R+9 — Romney got 59% there. Hanover gave 64% to Romney, pretty close to the R+17 in the two Presidential races.
p>I’d have to go through the numbers more systematically to see if this holds for other base cities/towns. It does seem logical that fewer Republicans would show up when they are getting crushed statewide, but surprisingly it doesn’t seem to be the case (at least in the subset of towns I’ve looked at).
Hoyapaul, thank you!
p>My simple question: will Coakley get more, or fewer, than 669,000 votes next Tuesday?
p>In the primary, the Dem candidates combined got 669K votes; the GOP combined, 165K votes; if I read the Sec’y of State’s tables correctly. I’d like to think that ALL those 669K voters who cast ballots in the DEMOCRATIC PARTY primary will turn out again, and vote for Coakley. If they do not, I will have a serious problem with continuing to let just anybody vote in “my” party’s primaries. If you’re NOT going to vote for the party’s candidate in the general, says I, then the party does itself no service by letting you help pick its candidate in its primary.
p>Meanwhile, I will just ASSUME that Cosmoboy will get many more than the 165K votes that were cast in the GOP primary. Four times as many would get him even with Coakley. Based on historical experience, do we expect 4 times as many people to turn out and vote for Brown, in the general, as turned out to vote at all in the GOP primary?
There’s probably not a whole lot exactly comparable to this Senate race, but consider the special election for MA-05 (the one Tsongas won, but closer than expected). Assuming the numbers on this Wikipedia entry on the race are accurate (which I believe they are), about 1,000 fewer people voted for Tsongas in the general election than in the entire contested Democratic primary. Ogonowski, on the other hand, received a little over 3 1/2 times as many votes in the general as were cast in the Republican primary.
p>So even though it seems logically unlikely, it definitely is possible that 4 times as many people could turn out and vote for Brown.
The Republican primary wasn’t seriously contested. Why would Republicans or right-leanings Unenrolleds show up under those conditions? Nothing interesting there.
p>But for a contested general…
Neither the Ogonowski nor the Scott Brown primary was seriously contested, which is probably why you’ll see considerably more people show up for the general in both cases than in the lightly contested primaries.
…that being a DA in Middlesex will help or hurt Coakley in that county?
Correlates to anything else in particular (income, ethnicity, % of Roman Catholics, % college graduates, or something). What makes Falmouth relatively conservative? Interesting stuff.
If you want a pretty interesting analysis of this sort of thing — at least one a county-by-county level, check out this analysis put together back in ’06 by MassINC. It speaks to a few of the demographic issues you mention.
as opposed to the presidential vote. Mass voted for Kennedy, Kerry and Romney. I think the Romney vote is more important here…,
Just a terrific piece of work. I’ll have it close at hand Tuesday night.
David, Bob, or Charley,
p>Any way you can frontpages this and use it as the open thread for this evening? Turnout #s would be a lot more interesting if readers had this close by to compare to.