Saturday, May 31, 2008

Michigan's turnout

Harold Ickes keeps shouting about 600,000 votes in Michigan being thrown out. But as I've pointed out before, the turnout situation in Michigan wasn't remotely normal. According to Jay Cost's spreadsheet, turnout in Michigan was equal to 24 percent of John Kerry's vote in 2004. However, the average in other states with open primaries was 79 percent. In other words, turnout was only about one-third as much as it should have been. The judgment of two-thirds of the voters in Michigan was essentially that the primary didn't matter and wasn't worth their time.

Florida was a little bit more normal. Turnout was equal to 48 percent of John Kerry's vote; the average in other closed primary states was 59 percent.

-- Nate Silver at 5:54 PM  

Labels: , ,

19 comments

Charles Pluckhahn said...

Christ, I think WA State, where I live, had a bigger turnout for the caucuses. Clinton's arguments with respect to Michigan are flatly bizarre.

May 31, 2008 6:26 PM  
Anonymous said...

One of the problems with the rhetoric in this debate has been the ambiguity of the word "voters," and how to protect the rights of the "voters."

To the Clinton camp, "voters" has a narrow meaning: those people who actually cast ballots during the FL and MI events. Only the will of those voters matters to that side.

The real issue is protecting the rights of the citizenry, which is a different interpretation of the word "voters." That objective would include everyone, and especially the un-name-able and innumerable citizens who would have voted but did not (or, in MI, who voted differently because of an incomplete ballot).

Poblano's latest post shows us that Ickes' argument - invoking what we might call a "statistical voter" - cuts both ways, and IMHO ultimately justifies the RBC decisions.

May 31, 2008 6:46 PM  
Anonymous said...

Nate, you have a great way of visually presenting a compelling mathematical argument. This is a great graphic that conveys the problem with the turnout-as-validation argument. I hope that people see this.

May 31, 2008 6:59 PM  
Sensible Person said...

Wow... the RBC was actually a but harder on the Florida and Michigan than I would have been. I was always a big fan of the "wait until this thing is decided and then seat them all (possibly with some symbolic vote reduction) when they can can no longer have an effect on the result" theory. However, once it became clear to me that the RBC was gonna do something, I switched to the "seat all the delegates with half votes and joint vetting of uncommitted Michigan delegates by candidates off the ballot" theory, mainly because that way, if the Republicans tried to use it as an election issue, we could point out that there solution gave the states the same amount of presidential representation, and potentially less platform representation. (Joint vetting essentially means Obama, because Richardson and Edwards endorsed Obama, and Biden knows he's gonna be the nominee.)

I really thought the 59-69 thing was really an either-or with regards to half votes, not both. Of course, the advantage of this solution is it preempts any attempt by Clinton to claim to be ahead in the popular vote counting Michigan.

And, of course, he only needs a max of about 26 more supers to cinch it, including the three remaining contests. I just hope Ickes isn't serious about taking it to the credentials committee... we might really need to laugh these guys off the stage.

May 31, 2008 7:23 PM  
Another Mike said...

The chart is somewhat skewed and the turnout discrepancy would have likely been even greater if MI and FL's elections had been sanctioned by the DNC. Note that states holding elections on Super Tuesday tend to have much lower turnout than all others (presumably because more time to campaign in a state leads to higher turnout). If the FL and MI contests had been sanctioned, the more valid turnout comparison would be to non-Super Tuesday contests, which had a much higher turnout.

May 31, 2008 7:41 PM  
timfry said...

Note that states holding elections on Super Tuesday tend to have much lower turnout than all others (presumably because more time to campaign in a state leads to higher turnout). If the FL and MI contests had been sanctioned, the more valid turnout comparison would be to non-Super Tuesday contests, which had a much higher turnout.

Huh? That chart lists all the states that had primaries - pre, post and during Super Tuesday. They are all there.

The issue remains, why was Florida in the ballpark of other closed primary states when Michigan voters clearly understood how it would not count? I don't think intelligence is the answer - heck, many FL voters are retired Michiganders. I remain convinced that the ballot initiative was a big deal. But either way, the argument that the MI voter must be respected is silly - the MI voter stayed home.

May 31, 2008 9:27 PM  
darren said...

For closed primaries I find a better metric (though not easily accessible through an RCP) is turnout among registered Democrats. Comparison to Kerry voters overstates red states like Texas, Indiana, and West Virginia. The result would be about the same - Florida would still be in the middle of the pack.

May 31, 2008 10:38 PM  
Anonymous said...

anon@6:46 - Protecting the rights of non-voters is a fine principle, but since there is no hard data as to who was "disenfranchised" it is a slippery slope to imposing whatever result is desired.
Politically, the Rules committee made a colossally stupid mistake by assigning 4 Clinton delegates to Obama. Not only did they needlessly irritate the Clinton campaign, but they gave them a perfect excuse to file an appeal and win public sympathy, and to question how evenly the rules are being applied. The committee thus prolonged the problem they were trying to solve once and for all. And for what? A swing of 4 half-delegates to Obama? Whether their decision is "justified" by conjecturing what might have been means little in the public relations arena when contrasted with actual voters whose votes were arbitrarily assigned to another candidate. Obama has been hanging his hat on the "rules" for so long and criticizing Clinton for "moving the goalposts" that accepting an arbitrary reallocation of delegates like this raises the spectre of hypocrisy. Do two wrongs make a right? Does Michigan moving up its primary date justify bestowing unearned delegates on Obama, especially Clinton delegates? It doesn't matter what you or I think - it's a huge opening for the Clinton campaign to beat this horse into July and beyond.

May 31, 2008 10:55 PM  
DC Economist said...

To the Anonymous Clinton troll above.

I suppose you think a primary where all candidaes said wouldn't could at the time it was held, that was known not to count, would have similiar turnout and voting patterns to a primary that counted in the same state?

Please.

May 31, 2008 11:13 PM  
Sensible Person said...

DC Economist -

I actually agree with the anon that it was a stupid PR move for the DNC. they should have just done half delegates all around, with joint vetting (which, as I said above, means uncommitted = Obama). And I'm an Obama supporter. I'm just worried that the RNC will have a field day (or dozen) with it. I didn't like the 59-69 solution when it was full votes, and it's even worse with half votes.

May 31, 2008 11:20 PM  
Anonymous said...

Another Obama supporter here who thinks that uncommitted should go to Obama (since Edwards, Richardson, and Bioden essentially consented to it), but that it should have been 73-55.

May 31, 2008 11:39 PM  
Ben said...

The real issue is protecting the rights of the citizenry, which is a different interpretation of the word "voters."

Actually, the real issue ought to be protecting the rights of Democrats (and I say this as a registered Independent). The primary process is supposed to be about a party determining its nominee. Much muddies these waters, especially "open" primaries and caucuses.

But if I were in the leadership of the Democratic Party, in the long run I'd be worrying about the rights of other members of my party. (Though I suppose one needs to, in open primary states, also be concerned with the rights of non-Democrats who might vote in the primary...talk about a murky number!).

Finally, all the speculations about the PR impact of this far overestimate how much the public cares about the RBC. This is real inside baseball (pardon the metaphor, Nate). And to the extent the public does care, validation of the Florida and Michigan results as "hard numbers" in any meaningful sense can be used by the Clinton campaign to play Calvinball on the popular vote issue. So, from a PR-angle, to the extent that anyone cares, it's damned if they do, damned if they don't.

Moreover, nothing but nothing is going to get Clinton to concede or move toward "unity" before Denver. So coming up with clever formulas for appeasing her campaign is a waste of Obama's and the RBC's time. The willingness of even neutral RBC members to hold the line is one price that Clinton is paying for her refusal to negotiate or argue in good faith.

June 1, 2008 1:16 AM  
vosh said...

anon@20:55, the votes cast in Michigan's illegitimate Soviet-style straw poll were not "arbitrarily assigned" anywhere. That is Clinton propaganda. The RBC made clear that they were trying to give voice to the est. 1.4 million Michigan citizens who would have voted in a fair primary. That is what they did in reassigning delegates. No one who voted for Clinton had their vote "stolen" or "reassigned." Nor did anyone who voted for uncommitted in support of another candidate Obama. Rather, the RBC refused to play the Clintonian game of ignoring the 70% of the electorate that didn't show up. They made an attempt to reflect those citizens who would have voted.

It was a blatantly unfair attempt. Too much weight was still given to the other 600,000. Nate's work here has clearly demonstrated that there is no basis for assuming Clinton would have any advantage over Obama in a fair Michigan election. If the state was going to be seated at all, in absence of a free and fair election, the split should've been 50/50. Sham elections should never be recognized. Clinton herself said "This election they're having is not going to count for anything."

As Chuck Todd reported, Obama had the votes for a 50/50 split. He gave Clinton a "win" in Michigan as an olive branch.

June 1, 2008 2:40 AM  
hosertohoosier said...

If your argument is that Michigan's primary was unrepresentative and abnormal - an argument I accept - how do caucuses survive on those same merits?

This is not to say that Clinton should win - by the rules, she lost. Insofar as super-delegates have some subjective ability to weigh the rules, the popular vote, and winnability, they have, and again, she lost.

However, I would think that there is a clear need for reforms of the primary process.

June 1, 2008 6:01 AM  
vosh said...

Great point. Reform is imperative, but only for reasons of fairness, which is why I doubt much will change. I loathe caucuses and hope the Democratic Party abolishes them in the future, but this is very unlikely because they are cheap and help with party building.

Caucuses indisputably suppress the vote (which is one of many reasons why the "popular vote" is not a legitimate metric under this primary system). However, they are also indisputably superior to elections that don't have major candidates on the ballot. They were also agreed to under the rules, were an even playing field for all the candidates. As much as I'm against caucuses, I'm even more against changing the rules after the game starts.

Here's a superb legal analysis shredding Ickes' fraudulent arguments: http://dailykos.com/storyonly/2008/6/1/72643/08080/210/526590

June 1, 2008 7:57 AM  
Anonymous said...

You all have the nerve to argue as if Obama was forced to take his name of the ballot, Obama pandered to the voters of NH & Iowa by volunterily taking his Name off the ballot... To give him delegates based on the Exit Polling is absolutely absurd... What about Caucuses.. Not everyone can make it the caucuse site..Should we take polling data from Utah, Kansas, Nebraska because Obama doesnt Really have a 35+ lead in those states if they actuaklly held primary,,, THE FACT THAT THEY REWARDED OBAMA FOR PANDERING TO NH & IOWA AND JUST HANDED HIM DELEGATES HE JUST DIDN'T EARN!!!!!!!!!!!!!! JUST SAD!!!

June 1, 2008 10:20 AM  
Anonymous said...

the RBC did t