There are three “macro” ways to approach the challenges faced by candidates in winning the unprecedented Democratic gubernatorial primary we are already in the throes of, and those three can be rougly broken down as numbers, message, and strategies. We’ll look at all of these, of course (and revisit them ad nauseum), but let’s spend a minute breaking out the numbers.
As I said in this recent diary, there are a lot of factors that will drive up turnout – I believe that all together, they could bump us to unprecedented levels – perhaps even an overall primary turnout of 60,000 voters. That’s assumption 1. Assumption 2 is that this race is going to sugar out to three leaders rather than two (or four or five, for that matter), and as such, candidates should be aiming their sights on a magic number of 21,000 votes to wrap this up. A lot of assumptions, of course, but not completely random ones. Check here to see some of the reasoning.
So where are those votes going to come from, and – just as interestingly – where will these specific candidates find the votes they need? There are no perfect models to look to for the purpose of making predictions, but everyone reasonably starts off looking at the 2006 Lieutenant Governor primary between John Tracy and Matt Dunne that generated a lot of interest among self-identifying Dems. Here’s what the turnout in that race looked like:
Before going forward, I want to point out that for purposes of comparison, I have a serious fudge factor in play, as we’re looking at (more or less) current voter registration totals to compare the turnout against. I think this is a fudge factor of quality more than quantity, as the mismatched data for comparison sounds like a bad thing, but I don’t think voter registration totals (or population spread) have changed that much.
So the turnout curve comes very close to matching the population curve by county, which we’d expect. Looking at the primary turnout as a percentage of the overall population for a county, we see GOP strongholds such as Caledonia on the low end (5.06%), and Dem-heavier counties like Windham on the high end (10.29%). The mean county value of turnout as a percentage of registered voting population was 7.31% in 2006.
If we look at the 2006 and 2008 Democratic Lt Gov primaries, we see that having a “native son” in the race does consistently drive up turnout, but it’s hard to say how much, given that each County in question – Chittenden, Windham, Windsor and Washington – tend to be higher performing counties for Democrats anyway. On the low end, when looking at the previously mentioned mean county value of turnout as a percentage of the registered voting population, you see John Tracy’s Chittenden in 2006 over that mean by a mere 1.03%, compared to Dunne’s Windsor on the high end at an extra 4.5%. Now, Chittenden’s results were virtually spot on the statewide mean in 2008 when the candidates were Washington’s Nate Freeman and Windham’s Tom Costello, so one could reasonably conclude that this 1.03% represents the safe “bump” in primary turnout in a county with a local candidate running, but Tracy only represented one Burlington district (compared to State Senators and statewide officeholders). That, along with the uniquely varied demographics of Chittenden, leave me inclined to take the mean of the “local bumps” from Tracy and Dunne in 2006, and Costello and Freeman in 2008, and call that mean (2.7%) the real “local bump.”
What’s the point? To put together a rough-but-not-too-rough “turnout curve” based on a gubernatortial primary with candidates living in Washington, Chittenden, Windsor, Windham and Lamoille Counties, then inflate that curve by whatever percentage necessary to get the total turnout to 60,000, which I believe will be close to the actual Democratic primary turnout. Doing so will give us a sense of how many votes will be available to the candidates from each county, and may then give us a sense of what numbers the candidates need to generate from where in order to take the nomination.
Here, then, is one turnout projection for the 2010 Democratic gubernatorial primary by county. Please note that the campaigns will be working with far more precise projections based on town-by-town analyses. This is a very fuzzy, broad estimate based on statewide averages projected to the county level. No way do I have time in my life to do this right:
What follows are possible “victory spreads” for each candidate assuming the above turnout by county. I have to say, it was often challenging to generate profiles that felt realistic that added up to 21,000 in light of the many candidates and their many geographic areas of strength. CHittenden has most of the votes, but Racine is going to win in Chittenden. Windham is another Democratic vote-mine, that candidates other than Peter Shumlin can only hope to finish second in.
Let’s stipulate that nobody can win with less than a 15% turnout in any county. It’s also clear that each candidate may be able to emphasize different facets of traditional campaigning to get to these numbers (field, paid media, earned media, etc), but all these facets will have to be in play. We’ll look at such strategies, given the likely financial pictures faced by each candidate, in another diary.
Susan Bartlett: I’ll admit, it was most challenging to make these numbers work for Bartlett, primarily because Lamoille County is a relatively lightly populated base of support. Bartlett will need to take Lamoille outright and push that momentum into the adjacent rural counties. She could also package herself as an alternative to the other “interstate corridor” candidates in counties like Rutland and Bennington – but she’ll be hamstrung by low name recognition. Like everyone, she needs to make a good showing in Chittenden County, and perhaps she can use her geographic proximity to push towards a second place finish to Racine, especially if she works the suburbs.
To bring up numbers around the state, she’s going to have to work to counter the anytime-now official endorsement of Deb Markowitz by Emily’s List, which will give Markowitz a sort of anointed “woman” candidate status.
Matt Dunne: It’d be easy to say that Dunne just needs to repeat his 2006 performance and eke out only a few more votes, but it’s not going to work that way. Dunne did show, however, that one can lose fairly decisively in Chittenden County (where he only took one town against John Tracy) and still win the state.
Dunne will win the highly populous Windsor County, and could use his popularity and geography to take two out of three of the other southern counties, and come in a strong second to Shumlin in Windham.
Dunne has maintained a good field network, which will give him good numbers everywhere. If he can tap into that loyalty – particularly in counties like Orange with well-established Democratic networks that know him well – he can generate votes in every county. Again, a second place showing in Chittenden is necessary.
Deb Markowitz: Markowitz is popular across the state among Dems. She isn’t the favorite among many constituency groups (well, labor), but those groups have been hesitant to dive into primaries, and not been terribly effective when they do, so her popularity should stay strong. With her campaign office already staffed by Emily’s List folks, the actual endorsement is only a formality. Coming into this stage with a lot of money and the EL endorsement will give her an across the board spread among the primary voters.
Not only will she take Washington County, but will be in a strong position to make a strong showing in the other interstate corridor areas – particularly Chittenden and Windsor.
Doug Racine: There’s nothing like starting with a win in the most populous county, but it won’t be enough on its own, as such he’s going to need to work his unique strength among Dems in the more conservative burbs such as Essex and South Burlington to win Chittenden big.
Racine is likely to struggle in the northeast and have a harder time in the south, so he’ll need to maximize his vote in neighboring Addison as well as take a strong second in Washington. Emphasizing his business and Party connections could serve him well in conservative Franklin/Grand Isle, and may give him a pitch in the Springfield area and into Windham.
Peter Shumlin: Shumlin takes Windham and needs to take a strong second in Windsor. He also needs to translate some of this two county momentum into the other two southern counties.
Shumlin is well known and well liked among most primary voters, and will need to tap into that goodwill to compete in rural areas with strong Dem organizations like Orange, but will not do as well in Franklin, where he may be the only candidate to suffer in the conservative area from being identified with marriage equality. However, that same identification could help him mine votes in Chittenden and make a strong showing in Addison. Interestingly, he’s already been touting his appointment of Susan Bartlett to the Chair of the Appropriations Committee in campaign contexts, so that could translate to a second place finish in Lamoille for those who like Bartlett, but don’t see her as a winner.
The caveat to all of this? None of this will be easy, and looking at these possible victory profiles, I feel fairly certain that whoever wins will probably have a different spread when the dust settles. Still, as a beginning point of discussion, this is as good a place to start as any…
With 60,000 voters and 3 viable candidates, a spread like this is well within the realm
Candidate A 24,000
Candidate B 21,000
Candidate C 10,000
Others 5,000
21,000 might win it if the top three are closer, or if the others stay stronger till the end. But I wouldn’t count on it if I was doing the numbers for a particular candidate. I’d want to aim for 25,000.
I don’t think the vast majority has made their mind up yet…and each of the candidates has a certain level of name recognition. The fact that they are spread around the state also helps them secure a base and fight for the big counties. So, 20,000 votes might be enough.