Gallup Daily Tracking Poll: Obama’s First Lead

I’ve been keeping my eye pretty close on the Gallup Daily Tracking Poll lately.  I know, I say it time and again, individual polls aren’t all that important, especially if you are trying to pin down an accurate prediction for a specific race.  And I still hold true to that–aggregated trends are far more reliable than individual data points.

This is doubly true when we are looking at national vs. state specific polling in a contest that happens a few states at a time.  Still that doesn’t mean that there isn’t a use for polls, and Gallup’s tracking poll, which at least provides a steady and stable indicator has become one of my two favorite tools.  My other is still, however, because of the lack of polling that we’ve seen in a lot of the post Super Tuesday states, the trends at pollster are not quite as accurate as they once were.

Specifically, Gallup’s daily tracking poll is useful in determining the direction of momentum, and acts as a kind of a barometer in the general public’s overall mood.  This barometer can then be applied to data that we have in specific states in order to correct them for time passed.

For instance, when I took a look at MattTX’s analysis of how Texas will go, I pointed out that much of his assumptions were based on demographic data available on Super Tuesday.  This creates a kind of conservative buffer in that campaigns are fluid, and how a demographic behaves on one day is not ever locked in.

A great example of this is the African American vote before and after Iowa.  Before the Iowa caucuses, Senator Clinton had a significant lead over Senator Obama amongst African Americans according to polling.  After Iowa, however, that demographic shifted significantly the other way.

If one were to try to predict South Carolina, for instance, based upon pre-Iowa data, that person would be so grossly off the mark that they would never be allowed to even casually mention who they think would win in a political contest ever again.

This is going to be true for Texas where Hillary is outright depending on the Latino vote and the blue collar white vote to halt Obama’s momentum.  In order to predict whether that defense will hold up, one needs to look at demographic breakdowns of contests that have occurred since major Latino populations helped Clinton to victory, and one can look at the daily tracking polls to update the assumptions one works under.

Now the most reliable data we have in terms of exit polls and demographics come from Maryland and Virginia where I understand Obama won the Latino demographics as well as most other demographics as well.  But there is an interesting dynamic that stems from observing the Hispanic demographic in that there are many different parts to it.  It’s not a homogenous group (no group is homogenous, to be fair, but what I mean to say is that the Hispanic demographic is far more complex than a single group would typically indicate, this a result of regional differences).

As a result, it’s difficult to say that because Obama did well among Latinos in the Beltway, he’s likely to not be quite so hindered by the same demographic in Texas.

So, there’s a possibility, though unlikely one, that Obama is making progress among Hispanic voters.  Now, Obama also won women in the Beltway, and there have been other indicators out there that show that he is cutting into the woman vote, as also seems to have been the case among blue collar white people.

So, in general, it is reasonable to believe that Obama has made progress among a lot of vital demographic groups, but because of the regional differences between the Beltway and Texas, it is next to impossible to accurately predict how this change will be reflected in the Lone Star state if at all.

So it’s on to the polls.  Pollster is showing that Obama is suffering from a twelve point defecit in polling trends in Texas, but on the other hand, he’s enjoying a massive surge in support, picking up seven, maybe ten points since the beginning of the year.  By contrast, Hillary has dropped a couple of points from the beginning of the year, meaning that they are narrowing, but there is still a sizeable gap between their support.


Without looking at the daily tracking poll, we can also apply a few other things we know.  We know that Obama has yet to really campaign in Texas, which means that much of this data is based upon name recognition.  We also know that Obama has a high success rate as far as winning over voters when he actually campaigns in a state, and he’s going to have three weeks to campaign in Texas, two weeks if he focuses all of his attention on Hawaii and Wisconsin until this upcoming Tuesday.

In other words, in the coming weeks, we can expect the already steep upward trend that Obama is enjoying right now to increase at an even faster rate once he begins campaigning there.

Now, let’s take a look at that National poll:

Remarkably, this is the first time that the Gallup Daily Tracking Poll has shown Obama with any kind of lead over Hillary Clinton, albeit one that is within the statistical error.  Regardless, the picture we see here is very clear: momentum.

Especially since the Super Tuesday contests have concluded, we see a rapid narrowing of the gap between the two candidates, and tomorrow we should be able to see if the blowouts that occured this past Tuesday will result in even greater distance gained by Obama over Clinton.

Thus, what we are seeing is a significant uptick in Obama’s support, and there’s something else to take into consideration, that name recognition factor.  Hillary has led Obama in polling from the moment pollsters started taking polls measuring the Democratic candidates.

One can reasonably assume that in the beginning, much of this has been a result of name recognition; compared to the rest of her candidates, Hillary Clinton had a much higher profile.  Her profile was arguably even higher than John Edwards who had ran in 2004, and served as John Kerry’s running mate.

Obama, by contrast, had comparably little name recognition going into this race.  Thus, the movement in the two graphs can be viewed this way: Hillary Clinton started off with the most supporters because she was best known, and she held on to a great deal of them because she honestly convinced them.  Barack Obama, on the other hand, has had to introduce himself before gaining supporters.

The reason I bring this up is because, if the changing of the guard shown above is not a statistical error and holds to be true and even widens, the fact that now both candidates are relatively well known, with Obama still to introduce himself to voters in states yet to be campaigned in, if he developes a lead in the polls over Hillary, he is far less likely to lose that lead as Hillary has over the course of the campaign.  This effect can’t be repeated.

How is this going to affect upcoming races?  There is no definitive way to predict at this point, but when we look at Texas, and we take all of this information into account, and we assume that MattTX’s analysis was for the most part correct, we can take that three delegate lead that Obama could get in Texas and we can talk on anywhere from five to fifteen delegates to that.

Of course, all of this is going to hinge on three things.  Will the Beltway Blowout result in actual momentum in the polls?  Will Obama win Wisconsin and by how much?  And how effective will Hillary’s campaigning be in Texas.  Right now when it comes to head to head campaigning, the Obama team appears to be a juggernaut, but we also have yet to see Hillary Clinton throw all of her resources into a state since Super Tuesday, and it’s more than feasible that when she does so, she can receive appreciable returns.

Now, the Republican race is pretty much sealed, however, that doesn’t mean there aren’t some things that can be gleaned from their daily tracker as well:

One thing that strikes me as interesting is that both candidates in the race are trending closer together.  The gap between them is far too big for Huckabee to make that up, nor is it likely to be transposed upon the individual races we have coming up, though if one were to take a look at the Texas trends, it is quite possible for Huckabee to win the incredibly delegate rich state:


But the other thing to think about is that this daily tracker also doesn’t accurately show the effect of the Potomac Primary either, and it’s possible that McCain may be on a surge.

If he’s not, however, the daily tracker spells out a bad trend for the Arizona Senator, one in which he’s continuing to lose support, something we see happening in opinion pieces throughout the political sphere; conservatives still aren’t behind him the way he needs them to be.  Thus, while the daily tracker is almost useless in predicting who will win the Republican nomination, what it does do is provide a snapshot of how McCain is doing amongst his own base.

As of this poll, he’s losing the battle; Huckabee, who would more aptly represent the base, continues to rise as McCain continues to drop.  If this trend continues into the General Election, McCain may find himself in trouble, not necessarily because the base is all of a sudden going to vote for the Democratic ticket, but because a lot of them are just going to stay home come election day.

2 Responses to “Gallup Daily Tracking Poll: Obama’s First Lead”

  1. Hey man, thanks for the love, but im back, got it out of my system with the last post, friday it will be like the regular. and thos plots look like the stuff i teach in my statistics class LOL. but still stole my thunder with that post on November

  2. No problem, and I’m glad that you got it out of your system RDB. You’re right, she lived to a ripe old age and I can only hope that I live that long.

    As for this stuff, I feel like Ferris Bueller, “I didn’t even take lessons”!


  1. Mitt Romney Will Endorse John McCain » Comments from Left Field - [...] little bit about this yesterday when we were taking a look at the Gallup Daily Tracking Polls–Mike Huckabee’s rise…

Leave a Reply

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

Connect with Facebook