MA-Sen: Town Benchmarks

The county baselines post has become a game day staple at Swing State Project for advanced elections-returns-watchers, and today’s no different. The basic idea here: find the bare minimum percentage in each major county that’s necessary to get the Democratic candidate over the hump at 50%. (That, of course, is predicated on all the counties moving the same direction as the presidential election the benchmarks are based on, which doesn’t actually happen in real life, but it’s a rough estimate.) As election returns come in, compare the benchmarks to the actual returns to see if we’re on track to win.

There’s one small problem here, which most of you are probably already familiar with: Massachusetts, and the other New England states, don’t report election results by county, but rather by town. With hundreds and hundreds of little towns, that’s a lot of ground to cover, so we’ll just look at the biggest, plus some towns in what seem to be the key areas to watch in this race.

In fact, let’s take a look at the state’s town-by-town map, created by our own DavidNYC (you can click on the map to see a full-size version):

This map is based on the relatively close 1996 Senate election between John Kerry and William Weld (a better choice here, because a map of the 2008 Presidential race would be almost entirely blue, and the 2002 gubernatorial race would be almost entirely red, with Romney winning the vast majority of towns). But it gives the general lay of the land in the state: the Democratic votes are heavily concentrated in Boston and its immediately surrounding cities. (There’s also a lot of low-density blue out in the college towns and arty enclaves of the Berkshires and Pioneer Valley in the west.) The red on the map is mostly rural and low-density too, so the real areas that we’re focused on today are the purplish and pinkish suburban turf to the north, west, and south of Boston.

Rather than just one model, I’m using two different models: one based on the high-turnout, high-Democratic-intensity 2008 presidential election (won by Barack Obama, 62-36), and the low-turnout, low-intensity 2002 gubernatorial election (won by Republican Mitt Romney, 50-45) — more generally, a best-case scenario for Dems and a worst-case scenario for Dems. With turnout projections high but not as high as yesterday (SoS William Galvin is predicting 40%), and weather mediocre, we’re probably looking at something somewhere in between, so consider these the bookends. Nevertheless, after making the necessary adjustments, both models, in most towns, point to very similar benchmarks.

Let’s start with the largest cities in Massachusetts (the ones that provide more than 1% of the state’s votes each). These aren’t really the places to watch, as they’re heavily Democratic (with the sort-of exception of Quincy) and the real question with them is whether turnout is keeping pace proportionately with the rest of the state.

Democratic performance in most of the state, as you can see above, stayed fairly consistent with the rise in the tide from 2002 to 2008. Statewide, Democratic performance went from 45% to 62% (a 17% gain), and, for example, Boston followed that closely, going from 61% to 79% (an 18% gain). Many other towns tracked that, too; for instance, the most conservative parts of the state (like Falmouth and Sandwich on Cape Cod, or the suburbs around Lowell) also moved about 16 to 18% in the Dems’ direction.

The interesting areas are the ones where the movement was much greater — these tend to be the wealthier areas in the state, fancy Middlesex Co. suburbs like Wellesley or North Shore towns in Essex Co., consistent with Obama’s overperformance nationwide among high-income voters — and where the movement was much less — mostly in blue-collar towns of the South Shore, as well as other blue-collar outposts in the state’s west. To me, these seem to be the swingy areas, and the ones most worth watching, especially since the trends may (or may not) continue to accelerate today — upper-middle-class voters may be attracted to Coakley’s technocratic image (especially those in Middlesex Co., where she was DA) or they may revert to liking the fiscal conservatism that they saw in Romney, while blue-collar voters seem likely to respond to Brown’s regular-guy shtick but may also be motivated by their ancestral Democratic loyalties and union or local machine GOTVing.

Let’s start with some well-to-do suburbs west of Boston:

And here are North Shore suburbs. (Lawrence is a little out of place here, as it’s working-class with a large Hispanic population, but it had the same large 02 to 08 shift that its wealthier neighbors did.)

Now turning to the more blue-collar locales where the trend seemed less favorable to Democrats, starting with the South Shore towns. (Looking up to the biggest towns list above, you can see that same trend happened not only in Quincy, but especially in Fall River and New Bedford, which are strongly Democratic but barely moved at all from O’Brien to Obama. In their cases, I’m not sure if that’s indifference to Obama, or particularly strong local machines good at keep turnout consistent.)

And finally, a mix of western mill towns and blue-collar suburbs around Springfield:

Now for the bad news… Suffolk polled several bellwether towns over the weekend, and found that Coakley is polling well below the level she needs to meet the benchmarks, in fact slightly below even Shannon O’Brien levels from 2002. They found a 41/55 race in Fitchburg (see the western mill towns chart), and a 40/57 race in Peabody (see the North Shore chart). I have no idea about the sample size or any of the other innards, but this suggests that for Coakley to pull this out — as has been more broadly evident for several days — the only way is for the pollsters to have been missing large swaths of heretofore unactivated voters who just got transformed into Democratic likely voters in the last few days.

Blue Mass Group has an interesting post that lists some other bellwether towns that you might want to keep an eye on. UPDATE: (And Cook’s Dave Wasserman has made available a Google spreadsheet doing more or less the same thing for every single freakin’ town in the state, albeit only for the 2008 model. Check it out.)