What Should the RBC Do With those Michigan Delegates?

“It’s impossible to discern the mind of each of the 238,000 people who voted for ‘uncommitted.’”
Harold Ickes, 5/30/08

Is Harold Ickes correct? Are the intentions of the "uncommitted" Michigan voters really all that much of a mystery? Well, not really. In fact, we can get a pretty good sense of the preferences of Michigan voters from the exit poll that was conducted in that state on the day of its primary in January. In fact, one of the questions on the exit poll asked which candidate the respondent would have voted for if all the candidates had been on the ballot. We can look at these figures in two ways:

1) The simple way to read the exit poll is to look at how these voters broke down in terms of their preferences. When asked who they would have voted for if all the candidates had been on the ballot, 46% of Michigan Democratic primary voters said that they would still vote for Clinton, 35% would have voted for Obama, and 12% would have voted for Edwards. Let's say you split the Edwards vote evenly between Obama and Clinton; then the vote would be 52-41% in favor of Clinton. ("But wait," you are probably thinking, "we all know that Clinton got 55% of the vote in Michigan, not 52%. Why would she have gotten less of the vote?" Good point, I'll get to that in a second.) Turns out, this is the most favorable reading of the exit polls as far as the Clinton campaign is concerned. If you look at things this way, giving Clinton a small majority of the delegates makes sense. (Alternatively, you could give Clinton 46% of the delegates, Obama 35%, and keep the rest (19%) truly uncommitted.)

2) Another way to look at the exit polls is to use them to get a sense of how many uncommitted voters wanted to vote for Obama (or Edwards) and whether any Clinton voters would have voted for a different candidate if they had the opportunity. This information is presented in the figure below, which shows the composition of Clinton and uncommitted voters depending on which candidate those voters actually wanted to vote for:


Now you can see why Clinton's support dips down to 46% when the exit poll asked respondents who Clinton voters would have voted for if all names had been on the ballot. A non-trivial share of her voters actually preferred Obama or Edwards. Of course, Edwards is not in the race anymore, so we'll let her keep those votes. But, based on the exit polls, about 6% of the electorate voted for Clinton, but would have voted for Obama if they had been given the opportunity to do so. This means that Clinton's vote would have been right around 50% if Obama's name been on the ballot. Thus, it is not simply the case that Obama didn't get any votes because his name wasn't on the ballot, but it is also evident that Clinton got more votes than she would have because some Obama supporters went ahead and voted for her. (As an aside, I'd love to know how many of those Obama supporters would do the same thing now.) If you look at the exit polls that way, then a 50-50 split of Michigan's delegates doesn't seem so far off base.

Two last thoughts on this issue, because it really will be the thorniest one the Rules and Bylaws Committee (RBC) has to tackle (halving the Florida delegation's votes is a pretty obvious and easy call at this point). First, a good share of the delegates who have already been selected to fill the "uncommitted" slots have already stated that they intend to vote for Obama if seated. So, even if those slots aren't dedicated to Obama specifically by the RBC, it is likely that Obama is going to get the lion's share of those votes. Second, what the exit poll cannot tell us is how many people stayed home and didn't vote in the primary because their favored candidate's name was not going to be on the ballot? Most likely, many supporters of both Clinton and Obama failed to turn out to vote because they didn't think the primary was going to count for anything, but the drop off might have been even greater for Obama supporters. Ickes is overstating things when he claims that we can't possibly know who the uncommitted voters really supported; the exit poll gives us some pretty solid evidence that most of them supported Obama (along with some of Clinton's voters). But what we can't really know is how many people might have gone out to vote for both candidates if all the candidates' names were on the ballot.