Half-Assed Vermont Delegate Projections

This is a sequel to my my Half-Assed Rhode Island Delegate Projections diary--which still appears to be the gold standard for Rhode Island delegate projections on DKos. (Admittedly that's not saying much.)

The idea of the series (of which this is the second and probably last installation) is to be a contrast to the work performed by Kossacks like MattTX, redvolution, and especially poblano. Those guys have spent hours poring though demographic data and conducting in-depth statistical analyses in order to produce thoughtful and well grounded projections of the delegate results in states like Texas and Ohio. In contrast, I find a poll, put together a simple spreadsheet, do absolutely no research whatsoever, and explain what I come up with.

So my process is proudly half-assed--but because the delegate calculations in the states I've dealt with are simple, I think my results are reasonably solid. (I'm not going to try this on Texas.)

[Note: I've had to tweak this on account of a few dumb math mistakes on my part that Scott in NJ pointed out below.]

According to the Green Papers' Vermont page, all fifteen of the pledged delegates Vermont will be awarding based on Tuesday's results are decided on the basis of statewide results. However, it isn't just a simple matter of apportioning fifteen delegates based on the candidates' proportions of the popular vote; that would be too easy. Instead, three different collections of pledged delegates will each be doled out, separately, according to the vote totals on Tuesday:

10 for the (single) congressional district
3 "at-large National Convention delegates"
2 PLEOs

As we'll see, this set-up--i.e., apportioning pledged delegates in three pools rather than a single 15-delegate pool--tends to favor the candidate running in second place.

The latest poll of the Democratic race in Vermont appears to be the Rasmussen poll conducted on Sunday, February 24, and published on February 25, one week ago today. That poll showed Barack Obama leading Hillary Clinton in Vermont 57% to 33%. For delegate selection purposes, this result is better stated as 63.3% to 36.7%--the respective proportions of Obama + Clinton voters alone, with undecideds and third candidates thrown out.

That's a substantial lead for Obama, and the overwhelming conventional wisdom appears to be that Obama will win Vermont, probably by a significant margin. Still, though, the at-large and PLEO pools of delegates help the runner-up: even if Obama wins by the 26-plus-point margin that Rasmussen shows, he will still split the PLEOs 1-1 and only win two out of the three at-large delegates. Indeed, even if he wins Vermont 74%-26%, those five delegates will still break for him only 3-2.

So, barring an unexpected surge for either candidate (pushing Clinton over 50% or Obama over 75%), those five delegates appear to be pretty much set in stone. The only question remaining would be how many of the ten congressional-district delegates the two sides will win.

The math on the district delegate issue is relatively simple (though a little less so now that I've installed the fixes called for by Scott in NJ's comment):

If Obama wins more than 50% but less than 55% of the Obama + Clinton vote, the ten district delegates split evenly, 5-5, and therefore Obama gets one net delegate (8-7) out of Vermont in total.

If Obama wins more than 55% but less than 65% of the combined vote, the district delegates go 6-4 for Obama, leading to a +3 for him overall (9-6).

If Obama wins more than 65% but less than 75%, it's 7-3 in district delegates and +5 (10-5) overall.

If Obama wins more than 75% but less than 83 1/3%, then the district delegates split 8-2. In addition, as soon as Obama rises over 75%, he picks up both PLEO delegates--so the grand total for this scenario would be +9 (12-3) overall.

As soon as he gets more than 83 1/3%, Obama scores a shutout in the three at-large delegates as well as the two PLEOs--and the ten district delegates still go for him 8-2. So for the slim band of vote percentages between 83 1/3% and 85%, that gives Obama a +11 (13-2) win.

In the unlikely chance that Obama tops 85%, then Clinton is in extreme trouble; she's not viable, and Obama wins all fifteen delegates. (Don't bet on this, my fellow Obama supporters.)

If the Rasmussen poll results hold up in the primary nine days later, Obama will win the Vermont delegates 9-6 for a +3 net. (Interestingly, this is the exact mirror image of the Rhode Island result that I think is by far the most likely; in that case, Vermont and Rhode Island will effectively cancel each other out. Oh, well.)

However, Rasmussen's result puts Obama very near to (1.7% away from) the level of support he would need to result in a 10-5, +5 pledged delegate finish. He has certainly overachieved his poll results before--to say nothing of his results in a poll nine days before a primary. So I think Obama's chances of getting five net delegates (more than overcoming his likely losses in Rhode Island) are not at all bad.

On the other hand, it's not inconceivable that the trend could be in the other direction, as it was in neighboring New Hampshire in the run-up to the primary there. I don't think it's likely, but it seems possible that Clinton could hold Obama under 55% of the O + C total and thus limit him to one net delegate.

So I'll project Obama to pick up three net delegates in Vermont, with +5 as a real possibility and +1 as a somewhat less likely outcome.

We'll close with a rousing round of everyone's favorite Shary Bobbins song:

If there's a task that must be done,
Don't turn your tail and run,
Don't pout; don't sob;
Just do a half-assed job....

If you cut every corner,
It's really not so bad.
Everybody does it,
Even Mom and Dad.
If nobody sees it,
Then nobody gets mad;
It's the American way!

The policeman on the beat
Needs some time to rest his feet:
"Fighting crime is not my cup of tea!"
And the clerk who runs the store
Can charge a little more
For meat (for meat!)
And milk (and milk!)
From 1984....

If you cut every corner,
You'll have more time for play;
It's the American way!

•  Cabot Cheese Tip Jar

  I miss living in New England, where every grocery store was well stocked with white Vermont extra-sharp cheddar....

  by Rieux on Mon Mar 03, 2008 at 07:49:01 AM PDT

  •  Extra-cheesy tips!! nt

    by vertexoflife on Mon Mar 03, 2008 at 08:11:21 AM PDT

  •  Half-assed is Highly Adequate.

    In fact, it's markedly better than paying for the truly puzzling results produced by consultants and national companies.

    by Man in the Middle on Mon Mar 03, 2008 at 08:20:37 AM PDT

  •  I say 10-5, & the Cabot yogurt is yummy too!

    I end every night with a bowl of the Cabot fat-free plain mixed with a pumpkin/flax-seed granola. Living large here in the Green Mountains!

    The Cabot cheddars are divine, and every grocery store carries a smorgasbord of them. My faves: the 50% and 75% reduced fat varieties taste just as divine as their regular cousins. Perfect for the VTer making herself "less significant," if you know what I mean.

    My half-assed prediction: we'll hit 70% in the popular vote for Obama! Polling data from my office: 100% - 0% BO-HRC.

    by earicicle on Mon Mar 03, 2008 at 08:26:03 AM PDT

•  Great Explanation

  very clear, with equally clear predictions.

  Given the time zones, state sizes, and likeliest outcomes in play Tuesday, I think facts versus spin favor Obama:  We're likely to see VT called for Obama about 3 nanoseconds after the polls close there.  If RI isn't called equally quickly, that's an early meta-point for Obama.  And if RI drags on, and we're hearing about surprising strength from Obama there, it'll color all the traditional hogwash about whether or not this little state is setting the tone for the evening.

  Next, if Obama seems to be winning TX clearly, whereas Clinton is either not winning OH or not winning by a substantial margin, we'll start hearing about a two-two split, and whether or not that would give Clinton enough momentum coming off of 11 straight losses.  Add in the commentators who actually bring up pledged delegates, and we'll hear about how a narrow victory in TX may nevertheless yield a big delegate advantage for Obama, which Clinton will be hard-pressed to make up in Ohio.

  And if Clinton is clinging to the barest of leads in Ohio, whereas Obama clears the threshold articulated above in Vermont and emerges with a five-delegate advantage, we'll hear about how cute, quaint Vermont may actually cancel out whatever delegate advantage Clinton manages to pull out in Ohio -- with the big Texas delegate bonanza becoming the effective story of the night.  And we'll hear more about the big state-small state silliness -- with Clinton's people trying to argue that only big, blue states like Ohio matter, and they therefore don't campaign in places like Vermont -- with disastrous consequences in the one place where it matters.

  by deminva on Mon Mar 03, 2008 at 09:01:00 AM PDT

•  If Obama beats Hillary 3:1

  then he wins both pledged PLEO's.

  If he beats her 5:1, he wins all three at large delegates.

  by Scott in NJ on Mon Mar 03, 2008 at 09:13:09 AM PDT

  •  D'oh!

    You're right. Clearly I didn't look at the lower reaches of my own spreadsheet.

    I've revised the delegate counts accordingly. Obviously this issue only matters if Obama gets more than three-quarters of the O + C vote, but (a) that's not out of the realm of possibility and (b) regardless of probability, my earlier statements about the 75%+ area were simply wrong.

    So thanks for the correction.

    by Rieux on Mon Mar 03, 2008 at 10:02:30 AM PDT