(Cross-posted at The Field.)
It's prediction time again and I begin with the usual disclaimer that a prediction is not an endorsement and my opinion doesn't represent RuralVotes or anybody else...
Learning the lesson from two of my past predictions that were off on statewide voter percentages but right about the resulting delegate counts out of each state, I've decided to call tomorrow's South Carolina - and future primary and caucus - results based on an analysis of how the delegates will be distributed. (In the delegate hunt, Obama tied Clinton for first in delegates in New Hampshire, and won by a sole delegate in Nevada even while losing the statewide vote tallies).
One of the problems in a country where folks are used to the winner-take-all Electoral College numbers in November presidential elections is that the media coverage of "who wins" tends to count by statewide vote rather than delegate counts in the primaries and caucuses, when it's the number of delegates that matters at the end of the day.
Full analysis at the jump...
There are 45 pledged (that is, bound to their candidate on the first two ballots) Democratic National Convention delegates at stake tomorrow in South Carolina: 29 by Congressional district, 10 "at large" delegates based on overall statewide voting percentages, and 6 "at large" delegates that are Democratic office holders but that will be pledged based on overall statewide voting percentages.
There are six Congressional districts in South Carolina with varying numbers (between 4 and 6 apiece) of delegates at stake. I've tabulated the estimated African-American vote in each, which should be helpful computing how the delegate count will go in each district if current polling trends hold.
Delegates per Congressional District:
- 31 to 33 percent African-American primary vote, 4 delegates = Obama 2, Clinton 1, Edwards 1
- 40 to 45 percent A-A vote, 5 delegates = Obama 3, Clinton 1, Edwards 1
- 32 to 35 percent percent A-A vote, 4 delegates = Obama 2, Clinton 1, Edwards 1
- 31 to 33 percent A-A vote, 4 delegates = Obama 2, Clinton 1, Edwards 1
- 50 to 56 percent A-A vote, 6 delegates = Obama 5, Clinton 1
- At least 80 percent A-A vote, 6 delegates = Obama 5, Clinton 1
District Delegate subtotal: Obama 19, Clinton 6, Edwards 4
10 at-large pledged delegates based on statewide percentages:
Obama 5, Clinton 3, Edwards 2
6 at-large pledged Democratic officials based on statewide percentages:
Obama 3, Clinton 2, Edwards 1
Total: Obama 27, Clinton 11, Edwards 7
Obama, who enjoys a one delegate lead over Clinton from the previous three contests, is thus likely to expand that lead in tomorrow's primary to a 17-point nationwide lead in pledged delegates.
That doesn't include national "superdelegates" (who are not pledged or bound to those they endorse and may switch at whim), or Michigan's so far unseated delegates.
Some other factors to watch in South Carolina:
- In the four majority white Congressional districts, Edwards may beat Clinton and if so, in each case, those Edwards-v-Clinton delegate estimates will flip. That's not likely to happen in the statewide numbers or in the two majority black districts because Clinton will get at least twice as many African-American votes as Edwards if current trends stick.
- The contest could narrow for a closer finish between Obama and Clinton, much closer than the 12 point average lead Obama has in the polls. Part of that would be due to a late Edwards surge, if it happens.
- Conversely, the Clinton campaign's radio ad and debate attacks on Obama for his words about Ronald Reagan could have the unintended effect of bringing more Independents and even Republicans into the Democratic primary and that would benefit Obama (remember that it snowed during the GOP primary last week and in some districts voting machine problems led to thousands being turned away from the polls, so they're still eligible to vote in the Democratic primary). Of Obama wins in a landslide, that will be one of the big stories out of South Carolina.
- Two recent polls by SurveyUSA offer cross-tabs for Hispanic voters, and both have Obama over fifty percent in that demographic group. The survey size is certainly too small to make presumptions from, but watch carefully the Hispanic (predominantly Mexican-American) precincts in Columbia, West Columbia, North Charleston, Goose Creek and other parts of the state to see if there's been any shift from the 64 percent Clinton to 25 percent Obama numbers in Nevada, as that could suggest some shifting to come between now and Tsunami Tuesday.
- Of course, we'll be watching, again, the rural vote to see if the trends of previous states continue. Current polls suggest a reversal of Obama's good fortune in rural communities from Iowa, New Hampshire and South Carolina, but the pollsters have been wrong before.
- Finally, if exit polls show Clinton losing among white voters to Edwards, that will serve as a major repudiation of race-baiting tactics and could give Edwards a healthy push heading into Tsunami Tuesday states: in Georgia and Alabama, where southern whites are demographically similar to those in South Carolina, as well as in Oklahoma and elsewhere. In that sense, the exit poll results may have a bigger momentum impact than the actual results, from which it will be more difficult to extract meaningful demographic information.
Two DKos related updates:
Neiman has posted a well thought out prediction about tomorrow’s delegate take in South Carolina: Obama 22, Clinton 14, Edwards 9. It’s also worth a read. The difference between his/her estimate and mine is partly because I weight Congressional Districts 5 and 6 as going at least five of six delegates for Obama, for the reasons explained above. That counts for four of the six delegate difference between our two totals.
Andre in Atlanta posted a blog earlier today about Georgia’s February 5 delegate take, and it was frontpaged over at MyDD, but, as I said in the comments there, I find that the African-American vote was undercounted by a factor of at least 2.25. (Overall registered voter percentages were used instead of Democratic primary voters.) Here, in this estimate on South Carolina, I did the kind of analysis that still awaits the Georgia contest, which I’ll do prior to February 5.