Post Election Analysis: It’s the Turnout, Stupid!

Dash’s victory in last Tuesday’s election was driven by higher turnout, an analysis of election data shows. (Photo courtesy of Jeanne Mooney.)

Summary: Higher turnout correlated with larger margins for Adam Dash, the progressive candidate across Belmont last Tuesday. In fact almost all the differences in voting percentages for Dash and his opponent can be explained by voter participation, according to analysis by Belmont’s Nate Silver (aka Claus Becker). The implications for future elections in Belmont (and the Bay State?) are huge. 

So if you’ve read Blogging Belmont, you’ve noted the frequent references to Blue Belmont. That’s the name I gave to my observation (obvious though it is) that our Town of Homes is a pretty darned progressive community. In fact, I’d argue that its among _the_ most politically progressive towns in the Bay State.

I first noted this back in November 2014. Even as the rest of the country and even the Bay State lurched rightward, Belmont went in the opposite direction. With slightly more than 10,000 people turning out – 57% of the town’s electorate – Belmont backed the colorless Martha Coakley over Governor Charlie Baker 55 percent to 42 percent. And, in the race for Attorney General, Maura Healey captured 69 percent of Belmont’s vote to become the nation’s first openly gay Attorney General.

Down ticket, Belmont showed its true blue colors even more emphatically back in 2014. Fifty eight percent of Belmont opposed a repeal of inflation-adjusted gasoline taxes. We voted against casinos 56 percent to 44 percent, also bucking the rest of the state. Sixty five percent of Belmont voters approved the measure to extend paid sick leave to State employees – hardly the vote of an electorate that’s suspicious of government.

The same thing happened last November’s election, when 82% of registered voters went to the polls and more than 7 in 10 voted for Hillary Clinton, fully 10% higher than Massachusetts voters as a whole, while we shot down a charter school expansion at almost exactly the same (high) percentage as the state in general. You get the idea.

But what does this mean for local politics? Well, the crux of the “Blue Belmont” theory is that low turnout elections paint the town more “red” (conservative) than it really is. As we’ve seen, there seems to be a lot of evidence of that.

A corollary would be that higher turnout makes Belmont “bluer,” favoring progressive candidates and issues disproportionately. In other words: more voters going to the polls doesn’t just deliver (narrow) progressive “wins” but actually accentuates the town’s progressive bent.

Numbers don’t lie: Adam Dash’s margin of victory in a given precinct correlated almost perfectly with turnout in that precinct. (Graph courtesy of Claus Becker.)

Is that the case? The (tentative) answer is “yes,” based on an analysis of last week’s Town wide election, which saw progressive candidate Adam Dash ascend to the Board of Selectmen, and a raft of new Town Meeting members elected, as well.

An analysis by Belmont’s own Claus Becker comparing voter turnout across Belmont’s eight precincts suggests that higher turnout in all eight precincts correlated with larger margins for Adam Dash, the progressive candidate. In fact, Becker found, almost all the differences in voting percentages for Dash and his opponent, Guy Carbone, can be explained by voter participation.

“The big insight here is that turnout is everything for Dash, but not helpful to Carbone.” – Claus “Nate Silver” Becker

In fact, from the viewpoint of statistics, the correlation was almost scary. “If you know the percentage that voted in a precinct you can explain 7/8ths of the variability in Dash’s (vote) haul in that precinct,” Becker wrote. “As someone who deals with data all the time I would say this is an astonishingly high correlation.”

That’s not true of Guy Carbone’s votes, where the association between turnout and Carbone voters is not statistically significant, Becker found.  (See the diagrams.)

Apathy wins again: there were 2 no-shows at the polls for every voter who turned out in April, a big drop from November, 2016, when more than 8 in 10 voters turned out. (Graph courtesy of Claus Becker.)

What explains the difference? In two words, ‘Blue Belmont.’  “There are about 200 to 300 conservative voters in each precinct who vote, reliably. Then there are a lot more liberal voters who are less dependable voters,” Becker wrote in an email. “This helps explain why Dash’s home precinct was his best (relationships drive turnout) whereas Carbone’s home turf (Precinct 2) didn’t seem to help him especially.”

With a deep well of fickle, progressive votes especially in Precincts like 1 and 8, which deliver the highest vote totals, Adam stood to benefit greatly from even a modest boost in turnout. But there aren’t that many conservative votes to be had in Belmont, in any precinct, and those that exist and turn out reliably were already going to vote for Guy, never mind where he lives. “The big insight here is that turnout is everything for Dash, but not helpful to Carbone,” Becker wrote.

Alas: the real winner (yet again) was apathy, Becker concludes. “For every one person who voted, there were more than two who didn’t.” Turnout was 28% town-wide. But that wasn’t evenly distributed. Across the town’s eight precincts, voter turnout ranged from 19% to 38%.

What does this mean for Belmont politics or (stepping back) Bay State politics? Locally, I think progressives in our Town of Homes should pay very close attention to Claus’s analysis and go to school on it.

Liberals and progressives need to stop running scared in Belmont. The sooner, the better!

How so? For as long as I’ve lived in town, the presumption among Belmont’s progressive community has been that victory can be had by only by trying to play to both parts of a small (3,000 to 4,000 person) electorate and – if I had to put a phrase to it – “not ruffle feathers.” As a progressive, the Town is generally against you, so hope to peel off a few conservative leaning voters by keeping talk of the Town’s desperate need (capital projects, schools, roads) to a minimum. Who knows, you just might eek out a win (even if April after April that theory has been proven false.)

The data from this April’s election, as well as Belmont’s continued voting patterns in big-turnout November elections (and September Primaries) suggests otherwise. The path to victory in Belmont isn’t playing “guess what my politics are” with a pool of 3,500 voters who skew older and more conservative than the electorate as a whole. Rather, it’s by trying to engage ‘the electorate as a whole’ by embracing progressive issues and candidates and trumpeting those policies to the biggest audience you can. Inspire passion, drive up turnout and dip into that deep, blue well of votes, and you’ll win the day. In short: liberals and progressives need to stop ‘running scared’ in Belmont, and the sooner the better.

At the State level, I also feel strongly that Massachusetts should change its laws to give communities like Belmont the option of holding local elections in November along with State and Federal elections. There’s just no good argument for towns continuing to spend precious dollars to hold elections that fewer than 3 and 10 voters turn out for, year after year. The tradition of April elections is quaint – in the same way that horse-drawn carriages are quaint. And they should go the same way. Our best April election turnout falls short of even a below average primary vote, let alone a November vote. With a simple change, towns like Belmont can consolidate their election activity and give anywhere from 2 to 3 times the number of voters the chance to weigh in on local candidates and issues that are now mostly lost on them. This is a message to Dave Rogers and Will Brownsberger: get out in front of this issue!

  • Mike

    Interesting analysis!

    • Thanks, Mike! Lots of lessons here for future campaigns.