[Update 2/27: After getting sucked into this discussion (where I can't comment because they haven't approved my registration) by way of this post, I went ahead and tried to come up with a rational estimate of popular support among registered Democrats for each candidate in the caucus states. But in trying to nail down the caucus ambiguity, I realized there is another, more
basic ambiguity that blows it out of the water and makes this whole
exercise pointless: Eight primary states and one caucus state do not
have registered party affiliations, at all! Over 38 million registered
voters!
Riddle me this: How many registered Democrats voted in the
primaries in Alabama, Arkansas, Georgia, Illinois, Missouri, South
Carolina, Tennessee, Virginia and Wisconsin? The answer is the same for
all of them: Zero, because there are no registered Democrats in those
states. So what the hell have I been adding up? Exit poll percentages -
which are subject to sampling error themselves, by the way - of voters’
self-reported party affiliation. If there is no way to objectively define what we mean by "registered Democrat", then trying to add up "registered Democrat" votes is impossible.
Just by the way, until yesterday I thought Obama fanatics were the worst in terms of being rude and divisive, but then I discovered that thread at corrente. Holy shit, they are crazy.]
Booman is very upset at this post by Jeralyn Merritt:
Who's really picking our Democratic nominee? If it's the Democratic
youth or African American voters, I'm okay with that. That's fair. If
it's Republicans, I'm not. We'll get trounced in November. Without
reliable stats to show Obama's support is from those who will for [sic] the
Democrat in November, I'd say the best way to ensure Republicans stay
out of our race and don't steal another election from us is for Dems to
vote for Hillary to be the nominee.
Her argument is really a non sequitur, because the article she points to says that Obama has won 64% of the independents, not Republicans, in the Democratic primaries so far. Rising to the bait anyway, Booman retorts:
In South Carolina Barack Obama won among registered Democrats 57%-28%-14%. In Virginia, he won among registered Democrats 62%-38%. In Maryland he won among registered Democrats 59%-40%. In Wisconsin,
he won the vote among registered Democrats 53%-46%. In spite of this, I
spent a good part of my day explaining to Clinton supporters (in email)
that Barack Obama is not winning the nomination on the backs of
Republican voters who won't show up for him in the fall.
It's a ludicrous assertion.
So just for completeness' sake, I went ahead and tallied all of the Democratic votes in the primaries so far. (I left out caucus states.) If you leave Michigan and Florida aside, Hillary has received about 7.4 million Democratic votes and Obama has about 7.1 million. If you include Michigan and Florida (and assume all "uncommitted" votes in Michigan would have gone to Obama), the totals increase to 8.4 million and 7.7 million, respectively.
On the merits, Jerlayn is clearly off base. Republicans have played a trivial role in the Democratic primaries. She's essentially arguing that winning a huge majority of independents is a bad thing for the nominee, which is absurd. Also, open primaries are open because the state Democratic parties want them to be. But the implication of Jerlayn's argument is that open primaries are illegitimate, and that we should only have closed primaries. Why? Finally, just like the issue of whether the Florida and Michigan delegates should be seated, note that Clinton supporters are arguing once again that the rules should be changed mid-game so that their candidate can win.
I relied on CNN for my vote numbers. Here's the spreadsheet I created:

Update: upyernoz says in comments: "why exclude the caucuses? i realize there isn't 'vote' numbers in the
usual sense, but there are exit polls that can be extrapolated into
something like a vote breakdown."
I excluded caucuses because I can't come up with a reliable way to correlate caucus votes to popular vote. For example, Obama won 75% of the caucus votes in Alaska, a margin he's not gotten near in any primary other than D.C. In Hawaii, same story: 76% Obama, 24% Clinton. Idaho: 79% Obama. Kansas: 74% Obama to 26% Clinton. Nebraska: 68% Obama - 32% Clinton. In Colorado, he got 67% of the caucus votes to 32% for Hillary, which is more believable but still really high. Maine had the most believable margin: 59% Obama - 40% Clinton. In contrast, the highest primary vote percentage Obama got - outside of D.C. - among registered Democrats is 57%. In the D.C. primary, Obama got 75% of the vote. To believe that caucus percentages are representative of the voters, you'd have to believe that coincidentally all the caucus states had Democratic voters who favored Obama as much as voters in D.C. (with 57% African-American population). Much more likely is that a combination of factors favored Obama in caucuses: a better ground game, a different demographic going to caucuses rather than primaries (say, fewer older voters), and possibly a "me too" effect of younger, more passionate Obama caucus-goers pulling people to their side of the room.
The point here isn't that caucus wins "don't count" or are illegitimate, the point is that while we can say confidently that Obama would have won the caucus states if they had held primaries, we can't confidently say by how much because caucuses are unrepresentative of the voters at large. You just can't multiply out caucus vote percentages by numbers of
registered Democrats and say that's the number of voters each candidate
would have won, just as you can't include Michigan and Florida.
Recent Comments