Today, for the first time in the eleven-month history of the Democratic nomination campaign, it seems to me that Barack Obama has become the frontrunner for the nomination. From December 20th, 2006, the first day when Obama, Edwards and Clinton were all included in Iowa, New Hampshire and national polls, through May 26th, the day that Clinton regained the national lead and didn't look back for a long time, the system I have used to chart the progress of the nomination showed several lead changes from Edwards to Clinton and back again. Obama, however, never secured the lead according to the system I used. In fact, with very few interruptions, he was in third place until August 5th, at which point the campaign took on the stable, familiar, first-through-third order of Clinton, Obama and Edwards. In terms of the sheer amount of coverage it provides, the national media has assumed this order pretty much since day one, but really election analysis really isn't a business the national media should dabble in, because they pretty much completely suck at it. For a long time, holding the lead in Iowa kept Edwards well clear of Obama in the nomination campaign, but few couple months ago that advantage became a thing of the past. Obama passed Edwards in Iowa polling averages at some point in August, and has never looked back since that time.
Now, it appears to me that Obama has done more than just pass Clinton. Today, for the first time ever, I now project Barack Obama as narrowly ahead in the overall nomination campaign. In the extended entry, I explain how I came to this conclusion.
Iowa Polling
Three-poll post-Thanksgiving average: Obama 27.0%, Clinton 27.0%, Edwards 23.3%
Four-poll two-week average: Obama 28.0%, Clinton 26.8%, Edwards 23.0%
Five-poll rolling average: Obama 27.2%, Clinton 26.8%, Edwards 22.6% Pollster.com regression line: Clinton 28.2%, Obama 26.2%, Edwards 21.1%
Is Obama ahead in Iowa? While it is difficult to make this conclusion with real certainty, an Obama lead is more likely than a Clinton lead. Even in the three polls after Thanksgiving that collectively show a tie, Obama has seen upward movement in all three of those polls. The favorable Obama trendline is emphasized further when one considers that the last nine polls with trendlines to come out of Iowa all show Obama improving relative to Clinton, and the last five polls from Iowa show him improving relative to Edwards. As I noted in the post below this one, more than any other factor in Iowa, I believe the trendline breaks all ties, and right now Obama has the trnedline in his favor. Obama also has an advantage on Clinton in what I believe is the second most important factor in Iowa, second-choice support. So, while it is a narrow lead, right now the odds are that Obama holds a lead in Iowa.
New Hampshire Polling
Four-poll post-Thanksgiving average: Clinton 32.8%, Obama 23.5%, Edwards 16.0%
Five-poll rolling average: Clinton 33.4%, Obama 23.2%, Edwards 15.4% Pollster.com regression line: Clinton 37.2%, Obama 22.9%, Edwards 14.0%
Average Iowa impact on New Hampshire
1st Place In Iowa: Plus 14.5 in New Hampshire
2nd Place in Iowa: Plus 3.2 in New Hampshire
3rd Place in Iowa: Minus 3.5 in New Hampshire
4th Place In Iowa: Minus 4.4 in New Hampshire
At this point in the analysis, we are assuming an Obama victory in Iowa. Given the average Iowa bounces candidates receive from Iowa, both the four-poll (Obama 38.0%--36.0% Clinton) and five-poll Iowa averages (Obama 37.7%--36.6% Clinton) project to a narrow Obama victory in New Hampshire. These are both narrow leads, but once again the trendlines are in Obama's favor. Eight consecutive New Hampshire polls that have trendlines have shown Obama improving in New Hampshire relative to Clinton. Further, the post-Iowa polling bounces for Bush, Gore, and Kerry all peaked between three and five days after the caucus took place. After five days, in all three cases the Iowa winner began to slide in New Hampshire polls. However, this time around, there will only be five days separating Iowa and New Hampshire, thus providing less time for any bounce to fade.
Now, the Pollster.com regression line, which is something of a lagging indicator and has not been updated to reflect the two new polls from New Hampshire that were released in the last twenty-four hours, projects that Clinton will still win both Iowa and New Hampshire (although it does project Obama to win New Hampshire in the event that Clinton finishes in third in Iowa). Further, past Iowa bounces are based on small sample sizes of only two past campaigns (1988 and 2004). Yet further, there is no telling what impact a narrow Iowa victory will have on momentum. Yet still further, there is no telling how much media time coverage of the Republican nomination will suck from the Iowa winner. In the past, the largest bounces have come from the campaigns when only one party faced a contested nomination: 1984, 1996 and 2004. So, just like in Iowa, there is a lot of guess work in these numbers, and nothing is guaranteed. However, once again, I think odds are that if Obama wins Iowa, he would also win New Hampshire.
Post-Iowa and New Hampshire
Nationally, Clinton leads Obama by between 19.2% (according to Real Clear Politics) and 20.8% (according to Pollster.com). Clinton also leads by 22-24% in Nevada (January 19th, RCP and Pollster), and by between 12.5% and 20.6% in South Carolina (January 26th, RCP and Pollster). Would such large advantages, coupled with a win in Michigan on January 15th, be enough to stave off two narrow losses in Iowa and New Hampshire?
According to research conducted by diarist fladem, almost certainly not. In his model designed to measure the impact of Iowa and New Hampshire results on national polls, Obama would gain 33% relative to Clinton in national polls if the junior Senator from Illinois sweeps the two early states. Not only is this enough for Obama to overtake Clinton in Nevada, South Carolina, and nationally, it is even enough for Obama to overtake Clinton in Florida, where she currently is projected ahead by 26.6% according to RCP, and by 29.7% in Pollster. When one considers the further momentum boosts Obama would receive from victories in all of these early states, February 5th would turn into a real rout, sealing the nomination then and there.
Put all of this together, and yes, I currently see Obama leading the Democratic nomination campaign.
Objections and Conclusions
The vagaries of projecting momentum were already discussed in the above sections, so I won't go over them again here. Suffice to say that this entire system is based on projecting momentum, which is influenced by several factors, from a small set of data points, and still only results in narrow Obama victories. A second objection to this view is that while Obama is more likely than Clinton to win Iowa right now, and while it is more likely that such a win would allow him to win New Hampshire than to not win New Hampshire, and while it is more likely that a victory in both states would catapult him to a lead pretty much everywhere else than to not do so, if the chain is broken at any single point, the entire system collapses. So, even if Obama is favored in all three of these circumstances, don't the odds overall actually favor Clinton because she only needs one of the three scenarios to go her way? Perhaps.
Still, it is hard to deny that the winds are clearly blowing in Obama's direction. Even if these objections make it difficult, if not impossible, to know whether Clinton or Obama is ahead, at the current rate of Obama improvement he will be clearly ahead in another week or two. If the next two polls from Iowa show Obama ahead of Clinton in both first and second place choices, and if the next poll from New Hampshire shows the campaign within single-digits, I personally will be convinced that Obama is in first place in the overall campaign. Anyway, even if he is ahead, as the post below this one notes, even with only 33 days until the caucuses, there is still lots of time for things to change. In this post, I just wanted to point out that today was the first time when I thought Obama might be leading, which I believe is a real milestone in the campaign.