9:00 PM. The thing that I find intriguing about Sirota's Race Chasm theory is that it implies a certain amount of non-linearity in the way that people vote. It may not be simply a matter of adding and subtracting constituencies but rather there are various sorts of tipping points and network effects.
It's hard to believe that only 5 percent of the Democrats in Floyd County, Kentucky came out of the womb ready to vote for Barack Obama. But maybe the number is 20 percent, and the 20 percent talk to their neighbors in the 80 percent, and before long, they become part of that majority as well.
Toward that end, one thing that may be particuarly relevant in Appalachia is that it tends to have very low rates of mobility. People from outside the region don't tend to move there, and people from inside the region don't tend to leave. As such, any sort of network effects might tend to be magnified there.
8:38 PM. The operative question tonight is not so much whether Clinton can get close enough to have a path to the nomination, but whether she can get close enough that she thinks she has a path to the nomination.
Her tone in her victory speech tonight was different than in West Virginia -- and seemed to suggest that she might think she does have such a path. Howard Fineman seems to think she might think so.
She certainly came pretty close to her best-case scenario tonight in terms of her popular vote gain in Kentucky -- not so much because of the margin (which everyone except my model seemed to get about right) but because of the relatively high turnout. Conversely, the magnitude of Obama's margin in Oregon might be fairly important in terms of disabusing her of that notion.
8:19 PM. Floyd County, Kentucky: Clinton 11,215 votes, Obama 653.
8:08 PM. When they're finished counting Kentucky's votes, Obama will hold a lead in the + Florida popular vote count of about 180,000 votes. Obama is liable to get anywhere from 50,000 to 150,000 of those votes back in Oregon. Erring slightly to the lower side of that estimate, he'll probably exit the evening about 250,000 votes ahead in the +Florida count.
I have absolutely no idea how to project the results in Puerto Rico. A 25-point win for Clinton on turnout of 1 million would get her those 250,000 votes. A 13-point win (the margin in the only public poll of Puerto Rico) for Clinton on turnout of 600,000 would net her 78,000 votes. Either of these results are plausible. An Obama victory or a very close result also is plausible.
The narrative issue that Clinton faces, I think, is that her argument loses moral force if it all boils down to what turnout in Puerto Rico is liable to be.
7:45 PM. To follow up on my previous thought. If you Google the phrase: "as _____ goes, so goes the nation", here are the states that come up most often.
1. Maine 2,870Maine was the subject of the original phrase, whereas Utah was the subject of a Mitt Romney joke. Nobody in the history of modern society has apparently used the phrases "As Oklahoma goes, so goes the nation", or, "As Alabama goes, so goes the nation". Until now.
2. Utah 1,100
3. Ohio 965
4. California 692
5. Iowa 352
6. Florida 221
7. Missouri 216
8. Texas 154
9. West Virginia 110
10. Michigan 103
7:28 PM. Clinton: "It has often been said: As Kentucky Goes, So Goes the Nation".
Actually, if you Google that phrase, it comes up just 16 times.
7:15 PM. To answer SPorcupine's question: you can infer those numbers from the exit polls, and 15 percent of Clinton supporters who stated a preference said that they'd vote for McCain in November.
6:42 PM. Chuck Todd clearly reads David Sirota. For what it's worth, I'm not completely convinced by the Race Chasm theory, but I think it makes for an interesting discussion.
FWIW, Obama has opened up a lead of a couple of points in Jefferson County (Louisville) and turnout there is enormous. Although, there's something funny going on with the reporting right now and I'm taking everything with a slight grain of salt at this point.
6:32 PM. It sure looks to me like Obama is not going to achieve viability in KY-5. He's losing some counties like 91-7.
6:16 PM. It's one thing for a campaign surrogate to spin information, but Terry McAuliffe just flat out made something up on Hardball, claiming that there was a general election poll in Kentucky that showed Hillary Clinton ahead of John McCain. If such a poll exists, there is no evidence of it anywhere on the Internet. I also heard him make the same claim about a week ago, so it wasn't any kind of misspeak.
6:02 PM. Kentucky called by all networks for Clinton. Exit polls suggest a big win, about 64-29. The exit poll also suggests that only about 9 percent of the electorate was black. Based on typical patterns in other states, you'd expect black turnout to equal about 150 percent of the state's African-American population, or 12 percent. So it could be that Obama's lackluster campaign in Kentucky meant that he didn't turn out his base. Or it could mean that the exit poll was a little off and Obama might slightly outperform those numbers.
5:53 PM. I don't want to call this a "liveblog" because it will be updated fairly sporadically, but if I have any thoughts worth sharing, they will go here.
Don't be fooled by the early results in Kentucky: almost everything is from Louisville. We had projected Louisville at 42/54/4 (Clinton/Obama/Other) and it's actually coming in at 48/49/3. So, Clinton is overperforming our estimates in this district by 11 points, which would imply about a 30-point win overall.