Barack Obama's polling numbers have been extremely encouraging lately. Within the past 24 hours, we have seen polls that give him the largest Iowa margin (+7) that he's had in the entire campaign cycle, the smallest New Hampshire deficit (-6) that he's had since July, the smallest South Carolina deficit (-2) that he's had since July, and the smallest national deficit (-7) that he's had since early June.
I get as excited about polls as anyone; at the same time, I believe that people tend to misuse them, focusing more on the end numbers and not so much on the internals.
One particularly important distinction is between "hard" or solid support, and soft or tentative support. Really, there is something of a continuum when it comes to persuading voters to end up in your column. Using Obama and Hillary Clinton as examples, we might place voters into five broad categories.
Realistically, at any given time, it is probably only possible to move a voter into an adjacent column. For example, you can turn an undecided voter into a soft Obama supporter, or a soft Clinton supporter into an undecided voter, but it's much harder to turn a soft Clinton supporter directly into a soft Obama supporter. You need to move voters toward you one stage at a time. By the way, this is also an important principle in marketing. When Honda shows a car commercial, they are not expecting people to drop everything and buy a car immediately. Rather, they are planting an idea in the customer's head, so that the next time he comes into the contact with the product -- via another commercial, via word of mouth, by seeing a nice-looking Honda on the street -- he may be ready to make a purchase decision.
Let's briefly review these categories, and the political strategies that go with them.
- Solid Obama. These voters are his to keep even if he hits the bankrupt spot on the wheel. He does not need to spend a lot of time preaching to this choir, although these voters may be useful for purposes like fund-raising and volunteering.
- Soft Obama. These voters are waiting for Obama to seal the deal. If the election is held tomorrow, they will probably vote for Obama, but they are vulnerable to being picked off by another candidate, and they are also vulnerable to finding an excuse not to show up at the polls. The goal with soft supports is to motivate them.
- Undecided. Genuinely undecided voters. These are relatively few in number at this stage of the campaign; on the other hand, to score one of them is a big deal, since you're at once adding them to your column, but preventing them from going into your rival's column. The primary goal with undecided voters is to persuade them.
- Soft Clinton. Going after these voters is lucrative, but risky. For one thing, you may move them into the undecided column, but they may drift back to their original candidate on election day, or even to a third candidate. For another thing, moving shaking off these voters usually requires you to attack your opponents, something which may cost you support from your own undecideds and leaners. Generally speaking, it is probably only worth going after your opponent's soft supporters if there are a lot of them. For example, on the Republican side, I suspect that Mike Huckabee has recently picked up a lot of soft support, so I can understand if a candidate like Mitt Romney played attack politics.
- Hard Clinton. These voters should be considered off-limits to Obama. The only way that they're likely to desert Clinton is if something truly cataclysmic happens, like some major scandal.
One last category that we should not neglect to mention:
- Soft Edwards/Others. As you'll see, this is a pretty important group in both Iowa and New Hampshire. In certain ways, these voters should be considered as being closer to the Undecided group than to the Soft Clinton group. When you're dealing with a candidate who ranks third or worse in the horse race, a lot of their support is liable to be soft because their voters are concerned about their viability. Therefore, you may not need to attack their candidate to peel them loose; they may tend to follow your coattails as you pick up momentum.
With these categories in mind, let's see how these numbers break down in Iowa, New Hampshire, and South Carolina.
The numbers you see in the top half of the chart are the percentage of a candidate's supports that say they will definitely vote for him; these are his hard supporters. For example, according to the most recent Rasmussen poll, 68% of Hillary Clinton's overall 27% number say they will definitely vote for her, or 18% of the total voting pool. One thing you will notice is that these hard support numbers are quite consistent from poll to poll. This is because a lot of the variation from poll to poll depends on how hard they push leaners to disclose a preference, and leaners are by definition soft supporters.
The situation in Iowa is at once dynamic and stable. Both Clinton and Obama have a lot of hard support; about 18-20% of the voting population is locked in to those candidates. Edwards has about 15% hard support, while the other candidates collectively have around 8%. Moreover, there are relatively few outright undecided voters. So, while anyone has a chance to win, there aren't that many voters that remain up for grabs.
The key exception to this is the soft supporters of John Edwards and the minor candidates. In fact, the 17% figure that I've provided in this column may be an underestimate, because some of the "hard" supporters of the other candidates may find that their candidates fall below the 15% viability threshold on caucus day. This number may well be close to 20% or even 25% of the voting pool.
How can Obama reach out to these voters? Well, a lot of this will depend on his ground game -- particularly the ability of his precinct workers to convert non-viable supports of Richardson, et. al. into his column on caucus night. I think it may be worth his drawing some very subtle distinctions with Edwards, such as making some references to his voting record.
The important point, however, is that Obama has almost no reason to go after soft Clinton supporters, because there aren't very many of them; most of what soft support she had has already been picked off. So it's imperative to keep the campaign positive and behave like the frontrunner.
I have started to believe that New Hampshire, rather than Iowa, is becoming the key focal point in this nomination. I would guess that of Obama wins Iowa, the candidate that wins New Hampshire may win the election as much as 80% of the time.
There is good news and bad news for Obama in New Hampshire. Although he's been closing in the polls, his support is relatively soft. In fact, we should expect these two things to go together. The definition of momentum in a political campaign is having recently converted a lot of voters into your column, but at first these voters are likely to be soft rather than hard supporters.
I think that Obama needs to be spending more time on the ground in New Hampshire. You can move voters from undecided into soft supporters from afar, but to turn them into hard supports probably requires more in the way of retail politics. In recent weeks, Obama has spent almost four times as much time in Iowa as he has in New Hampshire, whereas Hillary Clinton's ratio is closer to 2:1.
Campaign Events Since 9/1/07
Iowa NH
Obama 74 20
Clinton 60 27
Edwards 91 21
Additionally, the
emperical analysis on New Hampshire primary results that I did yesterday suggests that, while there is definitely an Iowa bounce in New Hampshire, the effects of Iowa are not quite as important as the pre-Iowa New Hampshire polls. I think that Obama should consider spending as much as 40-50% of his time to New Hampshire from here on out.
Let's also take a quick look at South Carolina.
These numbers should also be encouraging for Obama. Unlike in New Hampshire, it's Hillary who has a lot of soft supporters, whereas 65-70% of Obama's supporters are firmly committed. As a result, even a poll like the Pew poll that showed an nominal 14-point edge for Clinton, in fact boils down to only a 4% edge in hard support.
I suspect that a lot of this has to do with black voters, who I'm guessing may tend to be either soft Hillary supporters or dedicated Barack supporters; once they embrace Obama, they tend to embrace him fully, but for now they may be keeping him somewhat at arm's length. Either way, the numbers in South Carolina are good enough that I don't think Obama needs to be spending a lot of his pre-Iowa time campaigning in that state.
Summary
So, let's review some things that Obama needs to accomplish in the next 30 days.
1. Stay Positive; Behave Like the Front-Runner. Hillary has already lost most of her 'soft' support in the early states; this is what probability accounts for the polling discrepancies between these states and her national margins. Moreover, the numbers are now such that Obama can probably win the election by persuading undecideds, Edwards learners, and his own leaners, without having to worry about Hillary at all. Thus, there is very little reason to go after her directly, especially since Hillary only seems to have harmed herself with attack politics.
2. Devote More Resources to New Hampshire. Obama may want to consider spending a higher percentage of his resources in New Hampshire, especially with only five days for retail campaigning between Iowa and New Hampshire.
3. More Red Meat to persuade supporters of Edwards and other non-Hillary candidates. Broadly speaking, I would guess that voters who like Obama but prefer Edwards/Richardson/Dodd et. al. tend not to see Obama as substantive enough. I think he needs to give them some read meat, for example, in the form of a jobs/minimum wage proposal, or an outline of his goals for his first 100 days in the White House, or a major policy speech on the environment. He may also want to provide some tactful reminders about Edwards' voting record, but in general most of this is highly compatible with his goal of staying positive.
For Obama supporters, how would you like to see him spend his next 30 days?
For Edwards supporters, what would Obama need to do to persuade you that he's your guy?
For Hillary supporters, you can be assured that I'll be just as annoying about going after the Republicans in November as I am in going after Hillary now, regardless of whom our nominee might be.