Monday, February 11, 2008

Convention Math: Who leads the superdelegate chase this morning?

As the dust settles on the weekend's electoral battles, media organizations and amateurs alike are trying to get a fix on the delegate count. CBS News is out with a startling claim - it's Obama by a nose. The New York Times, meanwhile, shows Clinton by 171. So who's actually in control on this sunny Monday morning? Well, that depends how you count.

Superdelegates: This is where there's the greatest room for fudging the numbers. CBS is basing its story on a fresh survey of the superdelegates,which found 137 supporting Obama, and 211 for Hillary. There's reason to be skeptical of both those numbers. DemConWatch, the blog that's tracking publicly announced endorsements, puts the current totals at 128 to 223. Other media organizations have different totals, ranging upward to the AP's 156 to 243. And, of course, there's the Obama campaign's own claim from last week to a total of 170. Not to mention, it's a moving target - the tragic passing of Rep. Tom Lantos (a tireless champion of freedom) should remind us that neither the total number of superdelegates nor their names is etched in stone.

The Ruling: The problem here is that everyone's measuring something different. The media groups are conducting surveys, and the responses they generate depend on who answers their phone, and just how hard the reporters press for commitments. So the AP tally is a fine indication of which way 37 supers who haven't publicly announced their preferences are currently leaning (they split 28-19 in favor of Obama). That speculative trend is well worth noting, but let's not mistake it for the hard currency of actual, public endorsements. Until these folks announce their support, let's leave the tally at 128-223.


Pledged Delegates: After the chaos and uncertainty of counting superdelegates, tallying the pledged delegates should be easy, right? Well, maybe not. Here are three key disputes separating some major tallies:

1) How do you count caucuses? Voters at precinct caucuses elect delegates to county or congressional district caucuses, who elect delegates to the state convention, who elect delegates to the DNC. At every stage of that process, there's room for slippage - not everyone's going to show up, folks may change their votes, and candidates in the earliest caucus state subsequently dropped out. The New York Times, in its effort to produce an iron-clad count, excludes caucus states entirely until they hold their state conventions. That position is a reductio ad absurdum of its generally cautious approach to reporting totals. Most organizations count the caucus delegates by projecting the results of the state conventions.

The Ruling: For heaven's sake, count the delegates. We know that things may change - bear in mind, even pledged delegates are (almost all) perfectly free to change their minds when they show up in Denver. We're looking for a snapshot that gives us the most accurate impression of where things stand right now. Excluding the caucus states makes that a hopeless project.

2) How do you deal with incomplete returns? This one is a little trickier. CBS, for example, is showing only 14 delegates for Obama from yesterday's Maine caucuses, perhaps because only 99% of precincts have reported. The AP shows 15, awarding the final delegate to the man who is now statistically certain to win it. This one's not hard to settle - the Maine Democrats award the final delegate to Obama, too. But I wouldn't care to hazzard a guess about New Mexico, or to say for certain that the allocations in LA won't change when the votes of independents are tallied. A dozen little disputes produce a range of numbers:
CNN - 986/924; CBS - 997/920; AP - 952/893.

The Ruling: Here, I'm inclined to embrace the maximal interpretations, even at the risk that we'll have to adjust the results by a delegate or two when all is said and done. It seems fusty to leave 20 delegates from Washington unallocated, when they're virtually certain to split 11-9 for Clinton (although three counts do exactly that). Better to have a snapshot of where the race almost certainly stands at the moment than to produce a result - like that of the NYTimes - which is woefully disconnected from reality. So we'll go with CBS - 997/920. But I'd note that the spreads, as opposed to the tallies, aren't all that different: 62, 77, 59.


Unpledged Add-on Delegates: Yes, I'm back to the UADs again. No major media organization or website is tracking the 76 UADs who will be selected by state committees, conventions, or delegations. There are two ways to count them: you can award them to the candidate who carried each state, or award only those who will be chosen by a group known to be pledged to a given candidate.

The Ruling: Here, for once, I'd favor a more minimal approach. We really don't know how state party committees are likely to act; they're not bound to follow the results of primaries or caucuses, and their composition is often unreflective of the primary electorate as a whole. We know, for example, that in several states the key institutional players favored a different candidate than the one who prevailed at the ballot box; will they be able to resist the temptation to award the UADs in a contrary fashion, or to hand the slot to a 'neutral' party elder as a compromise? By my count, the UADs to be awarded by conventions and delegations stand at 15-7 in favor of Obama.

Totaling it up:
So I'd say that a fair count this morning puts Obama at 1,140 and Hillary at 1,150. So you could say it's Hillary by a hair. But it'd be more accurate to say that the two are, for the moment, effectively tied. A ten-delegate gap, given the complexity of counting, is too close to call.
What's not really in doubt is that Obama will gain enough delegates on Tuesday to surge into the lead. My guess would be that he won't release too many superdelegate endorsements in the next 48 hours. He'd rather claim the lead in most of these counts by winning elections, given his rhetorical posture on the legitimacy of superdelegates. To do so (assuming networks don't re-count their supers or resolve a large number of pending results) he'll need:
ABC: 18 delegates to take the lead
AP: 28 delegates
CNN: 27 delegates
CBS: 3 delegates ahead already
Given the projected size of his margins in the Potomac Primaries, he should easily clear all of those hurdles. And, since superdelegates in the last week or two have been breaking for Obama, and he seems poised to claim a majority of the unallocated delegates from previously held contests, it's even possible that one or more networks will follow CBS's lead in announcing Obama ahead before then. After all, networks hate to the be the last to a story. One way or the other, by Wednesday morning, everyone but the New York Times will show Obama firmly in the lead. And that's both a triumph and a major headache for a campaign that will be looking ahead at Ohio and Texas (and Vermont and Rhode Island), and the very real probability that their (overall) lead will vanish as quickly as it appeared, allowing Hillary to regain momentum.

No comments: