Long-time listeners may remember that late last year I posted two diaries which displayed in chart form the year's trends in the
national polls and in the
key early nominating states.
Now that the Des Moines Register's Iowa Poll has been released, the polling before the first votes which count for delegates are cast is basically over. There'll be one more day of Zogby tracking, but really? I don't care. As such, I thought it might be interesting if we revisited my charts and had a look at how the last twenty-odd days have changed the landscape, in Iowa and nationally.
As always, these graphs are being hosted by the spectacular, superdelegatariffic Adam in Mass. Kudos to hizonner. These are big graphics, so you'll find them behind the extended copy link.
(If you have any problems with parts of these charts being cut off, click the image. It's a link through to itself on its own.)
First things first. Here's the fundamental graph, of polls in Iowa from 1 January to yesterday:
But that graph is missing a few things. It does not incorporate the Zogby tracking poll results, only the snapshot polls. This is in part because the tracking polls aren't really cut from the same cloth as more conventional polls, and in part to see how the race looks without the day-to-day horserace which Dr. John Zogby has so kindly been providing to us.
The other reason is that the tracking poll mucks up my favourite chart, the moving averages. See my previous diaries for an explanation of why these are important...
Now, last time I wrote about this that three dynamics were immediately noticeable:
One. Dean's line always went up or stayed flat; it never seriously dipped. Obviously that's now over -- he started a precipitous drop in late December and just went on dropping. Whether or not it's true at the national level, then (see below) it's certainly safe to say that Dean peaked in Iowa in December. But -- only if you include the Zogby tracking can you say he doesn't command the #1 spot.
But look at the shape of the lines. In December, Dean is going slowly up, Gephardt's flat, and Kerry and Edwards are sharply up; this means everybody's gains are coming from undecided. But at the end of December, Kerry and Edwards rocket up and Dean takes a huge dive even as Gephardt starts a slump -- meaning Kerry and Edwards are now pulling their voters directly from Dean, and to a lesser extent from Gephardt.
Two. Gephardt is basically trapped in a band between 20 and 25 points. He can't break out of it. That remains true. Paradoxically, the fact that the other two candidates have closed up the race helps Gephardt. If it's a two-man race between him and Dean with Dean sitting around the 30-35 mark, he can't outhustle Dean enough to generate the extra votes he needs to pass him. But if it's a narrow four-way race, he quite possibly can outhustle the field to the tune of the two points or so which will decide the whole shebang. Take their slogan seriously -- fear the turtle.
Three. Kerry has a similar problem in a band around 15%. This is obviously not true, and along with the precipitous rise of Edwards, is the dynamic reshaping the caucuses. Kerry finally figured out how to drag his polls upwards. The problem, of course, is that in a four-way race with the four candidates about equal, organization wins -- and Kerry's organization is, if we're being generous, only the equal-second best in the race, and quite probably the third-best.
So those are the big trends. Let's take a look at the small ones -- the polls since January 1st, and this time, including the Zogby tracking.
Honestly? This doesn't tell me anything. The candidates are bunched so close that any of them could feasibly win. But it's fun to look at. I haven't done a moving-averages analysis of this poll because with so many data points so close together -- and so regular -- the moving average looks almost exactly like the raw polling.
And then, finally, there's the national polls. Take a good look at these -- they will probably start looking like a seismometer's output the day after Iowa:
(See my original diary for an explanation of why graphs of raw national polling are useless.)
It's quite clear at this point that there are no less than four tiers in this race: first there's Dean, then there's Clark, then there's Kerry, Lieberman and Gephardt, then there's Edwards, Sharpton, Kucinich, and until recently Moseley Braun. Clark took quite a sizeable step up over the holidays, but the few polls that have come out since about the 9th have shown him fairly steady. Whether this is just a short breather or a new plateau remains to be seen.
As for Dean, tempting though it will no doubt be for some people to call that mid-December mountain range his high-water mark, I'm not at all sure. What it does represent, I think, is the apex of his media buzz. With the Gore endorsement, he was pretty much walking on water and certainly cemented his front-runner status; then the press piled on the negative stories. This happens. While naturally the most desirable situation is for the media buzz never to peak, if it did, December wasn't a bad time. There was but a short sprint to Iowa, and at that point -- all going to plan, of course -- he can start earning buzz with tangible victories. It's a peak. It would be foolish at this point to assume it's the peak.
As extra in-flight entertainment, feel free to predict who'll place first in Iowa tomorrow with the aid of the handy poll below.
