Reader Posts
« previous | TPM CAFÉ READER POSTS HOME | next »
Projecting the Popular Vote
Disclaimer: Serious post with some math in it (calling flyonthewall and urbinato)
TPM Election Central recently posted the results of a poll suggesting that most voters believe that super-delegates should back the winner of the popular vote. We all know that a pledged delegate win is all but impossible for Clinton, but leaving aside the importance of the popular vote in the minds of superdelegates, what are Clinton's chances of actually winning it?
Before answering the question, a few caveats:
1) The popular vote is not sanctioned by the Democratic party as a way of choosing a nominee. Nor do I endorse it. I address the question only because it may affect the decisions of some superdelegates.
2) The vote described here is not truly "popular" because voter turn out is much lower in caucus states, so their representation is diminished. IA, NV, ME, WA have not even released official totals.
3) Leaving aside the question of FL, there is simply no good way to truly gauge the popular vote in MI. According to the vote count, Obama has zero supporters in MI. I assume that the supers know this, and I will let them decide for themselves how to weigh vote counts from FL and MI.
All the numbers that you need can be found here:
http://www.realclearpolitics.com/epolls/2008/president/democratic_vote_count.html
Counting only official results, Obama leads by 717K out of a total 26M. Estimating the caucus states of IA, NV, ME, and WA would improve his lead to 827K. Counting FL would bring him down to 533K, and counting MI would bring him down to 204K.
So what is Clinton likely to accomplish in the remaining states? Due to a surge in new registrations, there are now 4.2M registered Democrats in PA, which has a closed primary. Not all registered voters will turn out, of course. It's difficult to gauge that number. In OH, 45% of registered voters participated in the primary, though the percentage of registered Democrats is probably higher. To be conservative, let's assume a very high 60% turnout for registered Democrats. Polls show Clinton leading Obama by an average of 6%. Simple calculation: 4.2M x 60% x 6%= 151K. Not enough to surpass Obama's lead even if you were to count MI.
But now let's look to NC with 2.6M registered Democrats and Obama leading in the polls by 16%. NC allows independents to vote, so the percentage of votes to registered Democrats will likely be even higher than in PA, but to conservative, let's use the same formula: 2.5M x 60% x 16% = 240K. If the polls hold, Obama will actually extend his lead after the NC primary.
The polls for the remaining states are too limited and variable to come up with a calculation, but roughly speaking, Obama's gains in OR will likely neutralize any narrow losses in KY and IN. And for some perspective, all the other states put together, including OR, KY, and IN, have fewer registered voters than PA and NC combined.
In short, it seems very unlikely that Clinton will make much headway in the popular vote if she makes any headway at all.
Comments (5)
Good math. And I HATE math.
April 16, 2008 7:03 PM | Reply | Permalink
For what it is worth, I am not expecting Obama's loss in KY to be "narrow." Other than that, I think that you get it exactly right.
April 16, 2008 7:05 PM | Reply | Permalink
It's a great post, Genghis. But I'm afraid I've got to take issue with your conclusion.
I'd be shocked if Hillary carried Pennsylvania by just 6%, even though that's roughly the margin shown by the current polling. The issue is that the consensus also leaves 10% undecided, and I'm skeptical that they'll split evenly between the two candidates. Past rust-belt contests have shown Hillary claiming the bulk of those who remain undecided until the eve of the election.
So let's assume 10% as a reasonable margin. In an optimistic scenario for her campaign, Hillary could win by as much as 250k votes. Enough for her to claim at least one version of the popular vote lead. Or even if it's not quite that large a margin, say she gets enough to make it look pretty close. All of a sudden, there's a new race on which the media can focus - the drive for the popular vote!
That's the danger of buying into any version of this argument. Count Florida, discard Michigan, ignore the caucuses and - hey! - you've got yourself a race. Just drawing close enough in this count to claim she's still contesting something serves Hillary's purpose, as she continues to hang on in the hope that Obama will suddenly falter. And if she can divert attention away from the only metric that actually matters - the delegate count - and toward a more contested measure of success, she can score that a victory.
So I think that Clinton will actually make the race for the popular vote quite close. But it doesn't matter. Because she can't win the delegate count. The numbers just don't add up, and they'll get a little worse with every remaining election.
April 16, 2008 8:21 PM | Reply | Permalink
The day after the PA primary you look so prophetic, dear FotW. I tip my hat to you.
April 23, 2008 2:12 PM | Reply | Permalink
Thanks for weighing in Fly. The total delegate count is the only metric that matters. But the superdelegates are part of that total. I suspect that many of them will consider the popular vote in coming to their decisions, and they probably should. If it becomes clear that Clinton can't win the popular vote, I expect that the supers will commit, and the race will end. If she does well enough to stay in range of winning the popular vote (whatever that means), I think that the supers will continue to hold off. So the duration, if not the outcome, of the race really does hang, indirectly, on the popular vote.
PS This debate is dull. Don't know why I expected otherwise.
April 16, 2008 8:56 PM | Reply | Permalink
Post a Comment