More Hits From My Favorite Drug

I really can’t even explain it anymore. I know that Gallup’s Daily Tracker isn’t all that reliable, and as someone who spends a great deal of time sifting through polls, I know better than to give this one as much attention as I do. Still, I simply can’t help myself.

Seriously, I’ve explained once before my fascination with the GDT. It’s not so much the numbers that I find useful as the trends and how those trends could affect conditions on the ground. The Democratic party’s national numbers, for instance, are useless in that most of the primaries have already been held and each of the remaining contests will all have their own contours and idiosyncracies that will govern how they ultimately end up. But I believe that the trending in the GDT will have at a minimum a kind of ripple effect that will be transferred to some degree to the local contests.

Yesterday’s tracker tells an interesting story:

We are beginning to see what I had initially expected to see. There is a significant bump that occurs two days previous that Gallup attributed to a survey that greatly benefited Obama. Given that the tracker is a three day rolling poll, I predicted that we would see the bump, but then we would see the numbers gradually come back down.

Indeed, this day after bump almost perfectly mirrors the last day after bump data point with one crucial exception; Hillary’s numbers haven’t gone back up.

I also noticed something else with this snapshot that I think may become much more clear when we look at the entire history of the GDT since back in January:

Put together, this graph looks a little garbled, but when we look at each trend seperately, some interesting things jump out. The first is that Hillary has remained mostly constant. There was greater fluctuation in her numbers at the onset, but her numbers only dipped once, back at the beginning, under 40. Outside of that, she has maintained her polling numbers in the 40-50 range. The second thing is that Edwards’ retirement from the race seems to have benefitted Obama more than it benefitted Hillary.

The third and I think most interesting observation here goes to what Obama’s trend seems to be doing. If you look, you’ll notice that unlike Hillary’s which seems to oscillate in roughly the same ten point range, what Obama’s trend shows is that he surges, ebbs, but stabilizes out at a higher level. He then surges again, ebbs a little, but still stabilizes out better than he was prior to the bump.

The only aberration from this trend comes in Mid March where we can visibly see the hit he took from the Wright story.

We can also sort of possibly see where Hillary was weathering the Bosnia flap, though it’s a little more difficult to say explicitly that that the dip at the end of March was caused by the Bosnia story due to the fact that Hillary’s trend is less predictable than Obama’s.

The reason why this trend is interesting is that it gives us a possible indication as to how Obama’s going to end out the current bump in the trend he is enjoying now; specifically, by the time the survey that caused the bump is worked out of the system, Obama should actually stabilize out with a statistically significant lead unlike the last stabilization period where he was trending better than Clinton, but within the margin of error.

Another thing to think about is that at this point is that if Obama’s trend continues to maintain its pattern, the fluctuations that exist in both candidate’s trends will no longer be absorbed, so to speak, by the population of undecideds. This will, I believe, be the next thing to watch for in the coming weeks as it will indicate how well Hillary will be able to hold onto her support.

As for how the effect of the national polls on Pennsylvania, or at least analogous trending, I think it’s a pretty notable trend. As has been documented here continuously, and shown here on Pollster’s wonderful aggregated trend graph, we are seeing a rise in Obama’s Pennsylvania support:

08PAPresDems600.png

In this case, though, I don’t think either trend is driving the other, but instead they are occuring in tandem. On a national scale, the polls are being driven by a combination of news items, and of the narrative of the primary campaign which has shifted lately in the media towards being a little more critical of Clinton’s realistic chances of gaining the nomination. In Pennsylvania, this trend is essentially showing the struggle between two opposing forces; Hillary Clinton’s advantages with the local Democratic party apparatus and her name recognition against Obama’s campaigning efforts and grassroots organization.

While he’s started off with a huge deficit, though, the successes that he has enjoyed so far should not be ignored.

Moving away from the Democrats-only poll to look at the hypothetical match ups for the General Election, both Democrats enjoy something of a boost. Barack Obama enjoys his first numerical lead over John McCain in at least a month, and Hillary Clinton enjoys her first numerical tie. In both instances, however, the results fall well within the margin of error and are statistical dead heats.

and

It’s difficult to read these trends right now, though, and not exactly intuitive as to what has caused the shifts, if they could indeed be seen as shifts at all. While Gallup hints at the result of the hearings being held on the Hill, I’m not quite certain that they have had an effect or will have an effect at all.

On the other hand, in recent weeks violence broke out in Iraq twice and the two incidents put together could possibly be working to undermine a key factor in McCain’s campaign narrative.

And that’s enough of this drug for one day. Seriously, folks, moderation. You don’t want to get addicted to this one, it’ll kill ya.

(editorial blessing by DrGail)

Leave a Reply

Your email address will not be published. Required fields are marked *

Connect with Facebook