The Democratic primary is a delegate fight

When looking at Super Tuesday polls and considering who might win what states, don't forget that what will matter is as much delegate allocation than number of states won. Throughout January, only one state voted at a time (except for the GOP on 01/19) so it was a question of momentum. The day's winner was the person who won the state (which means that Clinton won Nevada). On Tuesday, 22 states are voting for both parties making it difficult to tell who comes out on top looking at the distribution of states.

The Democratic delegate allocation rules make it likely that Obama and Clinton will come out of February 5th with a comparable number of delegates and will remain viable. As long as they come within 5% in a state like Missouri, it matters little which of them comes out on top but for the sake of symbol and spin.

Delegates might not matter, of course, if one candidate sweeps most states. If Obama pulls together a string of victories in states that were supposed to be Clinton strongholds (New Jersey, Connecticut and even California) it will become very difficult for Clinton to stay in the race. The contrary is also true, if Clinton keeps her strongholds and pulls out victories in states that Obama is hoping to do well (Alabama, Colorado). But we are far from that for now, as most states are showing tight numbers with Clinton still very narrowly on top.

That means that both campaigns are strategizing to maximize their delegates. Some delegates will be allocated proportionally according to the statewide vote. No state appears ready to give a big win to any candidate at the moment -- including New York -- and that means that Obama and Clinton will remain close in this type of delegate.

Other delegates will be allocated proportionally at the statewide level. And this is where things get really complicated. For one, this could mean trouble for a candidate whose votes are concentrated in one area of the state. Let's say Obama and Clinton tie statewide in Alabama, but Obama gets more than 80% in districts that are predominantly African-American and that Clinton gets a more even 6o% in the other areas of the state, that could boost Clinton's chances.

Furthermore, this opens up a problem of odd/even-delegate districts. Districts have to distribute an even or an odd number of delegates and that will have an impact on the results. It is very difficult to get an extra delegate in an even-delegate district, while even a one-vote victory in an odd-delegate district creates gives the victor an extra delegate. This creates peculiar scenarios, one of which was described today in the LA Times:

A candidate who wins by a big margin in one district could end up with fewer delegates than a candidate who wins by a narrow margin in another.
For example, in a district with four delegates, a candidate who wins 62 percent of the vote would get two delegates — so would a candidate who wins 38 percent of the vote.
But in a district with three delegates, a candidate who wins 50.1 percent of the vote would get two delegates, and a second-place finisher with 49.9 percent of the vote would get one.

By organizing in even-numbered districts in which a candidate is supposed to be strong, a rival candidate can force a draw in the allocation of delegates. This explains why Obama is targeting the ares of New York City in which his supporters are active. The Politico's points out that the 5 districts he is organizing in are all 6-delegate districts. The same will be true of Obama in Illinois, who could be force to share an uncomfortable number of delegates in such districts.

The strange delegate-allocation rules also means that the candidates are not interested in campaigning in districts in which no candidate has a big lead but that award even number of delegates. In an area of California that awards 4 delegates but where neither candidate is likely to break ahead, it makes no sense for the campaigns to really engage since there is no chance they would break whatever number is needed (more than 62%) to get 3 of the 4 delegates. In the other hand, the districts that have an odd-number of delegates will be much more contested, though candidates will likely come out with a one-delegate difference out of them, not much of a reward for hard campaigning.

Overall, this will make out for a complex and very long Tuesday night, as the campaigns will wait for the breakdown of votes in all districts to know how they fared. And this will make the race particularly difficult for us observers to follow. If neither candidate breaks ahead by Tuesday, it could be very hard assessing how well each did come Wednesday morning.