Monday, March 31, 2008

Will May 20th Be the End of the Democratic Nomination Race?

If you think that calls for Clinton to get out of the race are prevalent now, wait until May 20th. Based on earlier projections from the Obama campaign, May 20th will be the day that Obama clinches a majority of the pledged delegates (minus MI and FL). The magic number is 1,627 and Obama's count currently stands at 1,408 (based on MSNBC's projections). Based on the delegate totals that Obama's campaign thinks he will win in the upcoming states (from the memo that the campaign inadvertently sent to a Bloomberg reporter), Obama would pass this milestone on May 20th, after picking up 28 delegates in Oregon and 23 in Kentucky. (NOTE: the Obama campaign's estimates have generally been conservative, they predicted he'd have accumulated 100 fewer delegates at this point than he actually has).

If he does clinch this on May 20th, might Clinton call it quits at that point? Certainly she would be facing increasing calls to get out of the race if she stayed in. In addition, on the 20th, she will likely have won Kentucky but lost Oregon. Thus, she would be able to leave the campaign on a day that she carried a state. The only states voting after the 20th are Puerto Rico and then Montana and South Dakota. She could win Puerto Rico, but she is likely to lose Montana and South Dakota and none of these states will do much to change the delegate count. If she left the race on May 20th, she could do it on a relatively high note and on her own terms, whereas anything after that may make it look like she was forced.

In addition, by May 20th, the Michigan and Florida argument may be largely moot. If you include Michigan and Florida into the equation, then a candidate needs 1,784 to clinch a majority of the pledged delegates. If you give Obama the delegates he won in Florida (67) and the uncommitted delegates from Michigan (55), the projection would put him at 1,771 pledged delegates on May 20th. There are 86 pledged delegates up for grabs in Puerto Rico, Montana, and South Dakota, which means that Obama would easily get the 13 pledged delegates he would need to pass the more difficult milestone (1,784 pledged delegates) even if he lost each remaining state by huge margins.

So, if the delegate count goes as projected, would Clinton get out on May 20th?

Friday, March 28, 2008

How Common are Defections in the General Election?

One of the stories attracting attention this week is a Gallup finding that a significant percentage of Democrats would vote for McCain if their candidate was not the nominee. I've blogged a couple of times already this week (here and here) in an attempt to put these findings into perspective. One question I have not yet been able to address, however, is the extent to which voters actually do defect from their party's nominee after a contested election.

Fortunately, the National Annenberg Election Study (NAES) can provide some insight into this question. In 2000, the NAES conducted interviews throughout the entire year focusing largely on the presidential election. Fortunately, even in their Fall interviews, the NAES continued to ask respondents who they had voted for in the presidential primaries. This isn't necessarily an ideal measure; after all, the Gallup poll is asking respondents who they support, not who they voted for. In 2000, the nomination campaigns were over by early March, meaning that many never had a chance to vote in the primaries when they were still meaningful (59% of NAES respondents reported that they did not vote in the presidential primaries). Nevertheless, the results are interesting, and worth taking a look at. The figure below presents the intended presidential vote for four different groups: those who voted for Bush in the primaries (15% of respondents), those who voted for McCain (5%), those who voted for Bill Bradley (4%), and those who voted for Gore (16%). This is based on over 10,000 interviews that the NAES conducted between September 4th and October 2nd.

How common were defections? Fairly common, actually. Even in September, only 49% of former McCain voters intended to vote for Bush and 29% were planning on casting their ballot for Gore (in March of 2000, a Pew Survey reported that 51% of McCain supporters planned to vote for Gore). McCain supporters were also far more likely to be undecided late in the race as 11% of this group reported that they still did not know who they intended to vote for.

Former Bradley supporters were also divided. While 52% of this group planned on voting for Gore, another 28% intended to vote for Bush.

Perhaps unsurprisingly, both former Bradley and former McCain supporters appeared more likely to prefer 3rd party candidate Ralph Nader. About 10% of Bradley supporters and 7% of McCain supporters expressed their intent to vote for Nader.

Overall, this suggests that it is not necessarily uncommon for voters to vote for the other party's nominee when their favored candidate has lost the nomination. In 2000, about three out of every ten McCain or Bradley supporters did so. Of course, the relative size of the former McCain/Bradley voters was not very big. Thus, ultimately, the defectors did not make up a large share of the electorate. The figure below presents the makeup of those who reported voting for Bush and Gore in the general election.

The largest part of Bush and Gore's coalitions were citizens who had not voted in the primaries. This is not surprising since turnout in primaries is usually far less than it is in general elections and in 2000 the nomination races were over after less than half the states had voted. What is really important here is how big a share of Bush and Gore's support the McCain and Bradley defectors accounted for. Based on this survey, former McCain voters accounted for 4% of those who intended to vote for Gore while former Bradley supporters accounted for 2.6% of those who voted for Bush. (Interestingly, citizens who voted for Gore in the primaries accounted for 1.8% of those who intended to vote for Bush and Bush primary supporters were 1% of those who intended to vote for Gore).

Ultimately, the relatively small size of the former McCain and Bradley supporters meant that the defectors did not account for a sizable share of either Bush or Gore's general election constituency. In 2008, two things will be different. On one hand, the size of the group of citizens who voted for the candidate who loses the Democratic nomination will be much larger than it was in 2000. This will make the potential group of defectors far larger and more significant. On the other hand, the 2000 election, while close, was not terribly polarizing. Many voters did not seem to think that it would make a major difference who won the presidential election. In fact, according to the National Election Study, 24% said they didn't care who won the election in 2000, compared with just 15% who said the same thing in 2004. If 2008 continues to be as polarizing an election as 2004, it seems less likely that a large share of Democrats or Republicans would defect since a polarizing environment generally means that partisanship will trump other concerns. Nevertheless, this analysis does suggest that defections do happen and there may be some cause for the eventual Democratic nominee to be concerned.

UPDATE: I should note that I also looked at the distributions for interviews NAES conducted after the general election. In these interviews, respondents told interviewers how they voted in the general election, not how they intended to vote. Essentially, I found very similar percentages of McCain and Bradley voters did, in fact, defect. This is not surprising since most voters have made up their minds by October. I used the October interviews in this post because there were vastly more interviews conducted in that month (providing a large sample size to work with).

UPDATE 2: Upon checking the data from my post, I discovered that I had been reporting that the data above came from October interviews when, in fact, it was from September interviews. I have now changed the information in the post to match the data that I was analyzing. I should note that there was little change in the October interviews. For example, McCain voters were only slightly more likely to say that they were going to vote for Bush (53% rather than 49% in September). My apologies for the initial mistake.

Thursday, March 27, 2008

More Context on the Democratic Defectors Survey

The Pew survey released today provides some more context on the Gallup finding that a significant percentage of Democrats would vote for McCain if their candidate was not the nominee. In the Pew survey, Democratic respondents were asked the following question:

"If Obama/Clinton wins the Democratic nomination, do you think the Democratic Party will unite solidly behind her or do you think that differences and disagreements within the party will keep many Democrats from supporting Obama/Clinton?"

Now, this is not exactly the same question Gallup was asking, but it is related and can help provide some additional context to those Gallup results. Regardless of whether Obama or Clinton's name was included in the question, approximately one-fourth of Democrats said that the differences and disagreements would keep many Democrats from supporting the eventual nominee. Fortunately, Pew asked the same question of voters in 2004 and in 1992 after Kerry and Clinton had clinched the party's nomination. In 2004, only 15% of Democrats said that the differences and disagreements would keep many from supporting Kerry. In 1992, however, a much greater 38% of Democrats said that the differences and disagreements would keep many Democrats from supporting Clinton. Ironically, Kerry lost the general election while Clinton won (though Kerry won a higher share of the popular vote than Clinton did). The 1992 case may be an even better comparison than the Republican contest in 2000, since that contest dragged well into the summer.

But here is the real interesting part of the story. In the survey Pew just released, they also asked Republicans the same question about John McCain. 22% of Republicans said that the differences and disagreements within the Republican Party would keep many Republicans from supporting McCain, nearly the same levels as those reported by Democrats for Obama and Clinton. What is also notable is that this figure for McCain was at 32% in a February Pew survey and has already dropped to 22% by the end of March. Thus, the problems for Obama and Clinton among Democrats do not seem that different from those that McCain has with Republicans now. Furthermore, the evidence for McCain seems to indicate that these issues melt away fairly quickly (going from 32% to 22% in just one month). These figures suggest, once again, that the news media (and many Democrats) are putting far too much weight on the Gallup numbers without the proper amount of context.

By the way, this is why I always look forward to Pew's survey reports. They generally provide excellent analysis with plenty of historical context. Kudos to them for including the figures from earlier campaigns when presenting this information.

NOTE: According to the same Pew survey, about three out of every ten supporters of Obama or Clinton say that they will vote for McCain if their favored candidate does not win the Democratic nomination. As I noted yesterday, this is smaller than the 51% of McCain supporters who said that they were going to vote for Bush in 2000 after McCain lost the race for the nomination.

Wednesday, March 26, 2008

Will Democrats Really Defect if Their Candidate Loses the Nomination?

One of the stories attracting attention today is a Gallup finding that a significant percentage of Democrats would vote for McCain if their candidate was not the nominee. But how much do these numbers really reflect what might happen in the general election campaign?

The last somewhat divisive primary campaign was in 2000, between Bush and McCain. In March of that year, the Pew Center for the People & the Press released a report titled "Bush Pays Price for Primary Victory." Following Bush's victory in the 2000 primaries and McCain's exit from the race, the Pew survey found that 51% of those who backed McCain during the primary campaign would vote for Gore in the general election. Only 44% of his supporters said that they would be casting their votes for Bush. Furthermore, a significant share of Bradley supporters also said that they would be supporting Bush in the general election, including 39% of his independent backers. (The figure below comes from the Pew report).

What is notable is not that Gallup finds that some Clinton and Obama supporters currently say that they may vote for McCain if their candidate loses, but that the number is so low compared to what it was for McCain and Bradley supporters in 2000. Only 28% of Clinton supporters (and 19% of Obama supporters) say they'd defect if their candidate lost, whereas half of McCain supporters were saying the same thing after he lost his bid for the 2000 Republican nomination.

Eventually, many of those McCain backers likely returned to vote for Bush and most of the Bradley backers likely returned to vote for Gore. The hard feelings that existed shortly after the end of the primary eventually subsided as the party unified for the general election. It is likely the case that Obama and Clinton supporters would eventually return to the fold and support the Democratic nominee in the Fall as well. However, the key difference between 2000 and 2008 will be the timing. When McCain lost the nomination, Bush had between 7-8 months to court McCain's old supporters. The Democratic nominee will have less time to do the courting this year. The critical question is how much time will he or she have?

UPDATE: I've posted some additional context on this question here.
UPDATE 2: I've now also looked at survey data from fall 2000 to see how common defectors were once the general election came around.

Tuesday, March 25, 2008

New North Carolina Poll Suggests Obama's Speech Helped Him With Whites

There is a new poll out of North Carolina today from Public Policy Polling (PPP). PPP released a poll last week that was based on interviews conducted on Monday, March 17th. This was after the Wright scandal had broken but before Obama's speech on race. In that poll, Obama and Clinton were in a dead heat (Obama 44%, Clinton 43%). However, in the poll conducted on March 24th, Obama's lead was 55-34%, a major shift.

Perhaps even more significant is the movement among whites polled in North Carolina. Last week, Obama trailed Clinton by 26% among whites (30-56%). However, that gap narrowed to just 7% in the most recent survey (40-47%).

This provides some initial evidence that Obama's speech was successful in helping him win back some support among whites who may have been driven away by the Wright scandal. Will there be a similar shift in Pennsylvania? We should know relatively soon.

UPDATE (2:15pm): As if on cue, Clinton has raised the Wright controversy in a recent interview.

UPDATE 2 (2:25pm): On looking over the crosstabs a bit more, I noticed something else interesting in NC. Not only did Obama increase his support among whites since the speech, but he also gained some support among African Americans. Last week, blacks in NC favored Obama by a 72-19% margin. However, in the most recent poll, that margin increased to 80-14%.

UPDATE 3 (6:07pm): Avi Zenilman at notes that PPP changed the way it is defining its sample in its newest survey. You can find PPP's comment on this here. They believe that some of the increase in Obama's numbers can be attributed to this change in the sampling frame.

New Unpledged Democratic Superdelegate Predictions

For the last several weeks, I have been using statistical models to generate estimates for how likely each unpledged superdelegate was to vote for Hillary Clinton or Barack Obama. At last check (March 3rd), the models had successfully predicted the endorsements of 71% of the previously unpledged superdelegates. Thus, this model provides a reasonable way of getting a sense of which candidate the remaining superdelegates will favor.

Before I get to the new predictions, I want to update the information on how accurate this model has been. Since my last estimates, 17 superdelegates have announced their candidate endorsements and we predicted 11 of those correctly (65%). (All information on superdelegate endorsements comes from Democratic Convention Watch). Overall, since we began generating these predictions, 82 superdelegates have announced their endorsements, and we have been correct on 57 of these. Thus, overall, the models have been correct about 70% of the time. Below I have listed the most recent 17 endorsements.

I have updated the predictions for which candidate unpledged Democratic superdelegates are likely to support. As before, I use information about the superdelegates who have committed to a candidate to generate predictions for the remaining unpledged superdelegates. I exclude superdelegates from DC and the territories because we lack complete data from those areas, and from IL, NY, and AR because superdelegates in those states have nearly unanimously cast their support for their native son/daughter. As always, information on the superdelegates is provided by the Democratic Convention Watch site. You can find more about they methodology I use here. Check out the distribution of predicted support among unpledged superdelegates below.

Superdelegates who are between 40% and 60% likely to vote for Clinton/Obama are labeled as "unclear." There are 78 superdelegates in this range. There are 171 unpledged superdelegates who are at least 60% likely to vote for Obama; just 19 unpledged superdelegates are at least 60% likely to vote for Clinton. These predictions suggest that Obama will be able to cut into and even overtake Clinton's superdelegate lead in the coming weeks and months. Unless something significant changes, there seems to be little hope for the Clinton campaign in hoping that the superdelegates will help her erase Obama's lead.

The estimates for each unpledged superdelegate are listed here. These estimates show that among Obama's most likely endorsers are Governor Dave Freudenthal (WY), Rep. Dennis Moore (KS), and Rep. Tom Allen (ME). Clinton's most likely endorsers include Reps. Jerry McNerney, Susan Davis, and Lois Capps (all from CA).

Monday, March 24, 2008

Testing Clinton's Argument: Does Winning a State's Primary Translate to Winning the State in the General Election?

Earlier, I jokingly posted a map that was meant to make the point that there is little support for the notion that winning a state's primary/caucus has anything to do with what will happen in the general election. To put a finer point on the question, I went back and looked at the last three contested Democratic nomination campaigns (2004, 2000, and 1992). In each case, I looked at every state that was contested before the eventual nominee had effectively won the nomination (it doesn't tell us much if a candidate wins a state once his opponents have exited the race). I'll address two big questions with this data:

1) Does losing a state's primary mean that a candidate will lose the state in the general election?

The eventual Democratic nominee lost 12 states in these three campaigns: Clinton lost 9 states, Gore did not lose any states, and Kerry lost 3. Of these 12 states that the eventual nominee lost during the nomination campaign, he lost just three of those in the general election. Kerry lost Oklahoma and South Carolina in 2004 and Clinton lost South Dakota in 1992. That means that since 1992, a Democratic nominee has won 75% of the states that he lost during the nomination campaign.

2) Does winning a state's primary mean that a candidate will win that state in the general election?

Since 1992, the eventual Democratic nominee has won 50 states while the race for the nomination was still being contested. The Democratic nominee went on to lose 19 of those 50 states (38%) in the general election. Kerry lost 8 states he carried during the primaries (IA, AZ, MO, TN, VA, UT, GA, OH), Gore lost 4 (NH, GA, MO, OH), and Clinton lost 7 (ID, SC, FL, MS, OK, TX, KS).

Kerry and Gore's campaigns are particularly instructive. In 2000, NH, GA, MO, and OH could all be considered potential "swing states." Gore lost each of these states in the general election despite winning their primaries. In 2004, IA, AZ, MO, TN, VA, and OH could all have been considered potential "swing states." Kerry lost each of these despite winning their primaries.

The bottom line is that there is little support for a claim made by either candidate (Obama or Clinton) that winning a state's primary is closely related to the candidate's chances in that state during the general election. In fact, the Democratic nominees since 1992 have fared better in states that they lost during the nomination campaign (winning 75% of those states in the general election) than they have in states that they won (winning 62% of those states).

The 2004 Electoral Map...Under the New Clinton Argument

The Clinton campaign is now trying to make a case for why she should win the nomination. The argument is that she has won states with more electoral votes than Obama.

Kerry surely would sign on to this logic...check out the landslide victory he would have won if only winning a state in a primary translated into winning that state in the general election:

UPDATE: See my more in depth analysis of how well primary victories/losses predict general election outcomes.

How Would an Obama/Richardson Ticket Fare?

With Richardson's endorsement of Obama on Friday, some have raised the possibility of an Obama/Richardson ticket for the White House. The thought seems to be that Richardson would add some foreign policy heft to the ticket while also helping Obama with the Hispanic community. But how formidable would an Obama/Richardson team actually be in the general election? Even if Richardson could bring a large majority of Hispanics to the Democratic column, would it be enough?

A good place to start this exercise is by looking at 2004 exit polls:

Group % of Electorate % Kerry
White Men 36% 37%
White Women 41% 44%
Black 11% 88%
Hispanic 8% 53%
Other 4% 55%

In 2004, African Americans comprised 11% of the electorate while Hispanics made up just 8%. While 88% of African Americans voted for Kerry, only 53% of Hispanics did so (though this figure has been disputed to some extent). The 53% figure was rather low and rebounded in 2006, when nearly 70% of Hispanics voted for Democratic congressional candidates.

The first step is to figure out what the composition of the electorate would be in 2008. Let's begin by assuming that turnout among African Americans and Hispanics would be higher since there would be a black and Hispanic man on the Democratic ticket. The Pew Hispanic Center estimates that Hispanics will make up 9% of the eligible electorate in 2008. Let's assume that with Richardson on the ticket, their turnout is comparable to their share of the eligible electorate. Let's also assume that with Obama running for president, that African Americans comprise 12% of the electorate in 2008. So, under these assumptions, the electorate would look something like this:

Obama/Richardson % of Electorate
White Men 35%
White Women 40%
Black 12%
Hispanic 9%
Other 4%

Based on this hypothetical electorate, let's produce two scenarios. First, let's assume that an Obama/Richardson ticket would win the same level of support among whites that Kerry won in 2004. However, let's also assume that they would win 90% of the black vote and 80% of the Hispanic vote (10% better than Democratic congressional candidates did in 2006). Under those assumptions, Obama and Richardson would carry 51% of the popular vote.

Second, let's take the worst case scenario for the Democrats and assume that this ticket would lose some support among whites. Let's assume that Obama and Richardson would do 5% worse among white men and 4% worse among white women. If that were the case, then they would capture just 47% of the vote, which would obviously fall short of victory. In other words, increasing Hispanic support from 55% to 80% would be off-set if support among whites dropped by just 4-5%.

But which of these scenarios would be more likely? In many ways, it is hard to imagine that support for a Democratic ticket could decrease among any demographic group in a year as favorable for Democrats as 2008. This suggests that scenario 1 may be more likely. Then again, it is hard to know for sure how this would play out. More importantly, there are a lot of high profile names out there for the position, so we likely will never know how an Obama/Richardson ticket would have fared.

Saturday, March 22, 2008

New York Times Piece on Internet Voting

Even though it seems unlikely at this point, the Michigan Democratic Party could still try to sponsor its own re-vote using mail and Internet ballots (something I've noted before). The New York Times has a story about the potential benefits and pitfalls of voting online; a study conducted by myself and Alicia Prevost is referenced in the article. The debate over Internet voting is complicated and the article does a nice job of addressing difference perspectives. It is worth a read.

Friday, March 21, 2008

Richardson Endorsement: We Predicted It.

Bill Richardson's endorsement of Obama today is news...except to us. Our superdelegate prediction model showed that Richardson was likely to endorse Obama. In fact, our model indicated that it was 77% likely that he would endorse Obama and there was just a 23% chance that he would endorse Clinton.

By the way, in case you were wondering, we also got the Murtha endorsement correct.

Obama Recovery? Daily Tracking Numbers Turning After Speech on Race

I mentioned several days ago that one way to judge the political impact of Obama's speech would be to see what effect (if any) that it had on his numbers in the Gallup Daily Tracking Poll. The day after his speech, Obama's numbers had dropped to 42%, down 8% from 5 days earlier. In the meantime, Clinton support was at 49%, up 5% from the previous week. So, what effect did the speech have? Well, today's Gallup numbers are the first based exclusively on post-speech interviews. The numbers are presented below. The red line indicates the point at which the Wright story broke while the blue line indicates the day of Obama's speech.

While it is still early, the tracking poll appears to show that Obama's speech had the intended political effect of stopping his slide in the polls and even reversing the effect to some extent. Obama has gained 3% in two days while Clinton has lost 2%, and the gap between them is now back within the margin of error. It will be interesting to see if these numbers continue to move in the coming days. In the meantime, it is back to March Madness for me...

UPDATE: Quickly, after the incredible shot Western Kentucky just hit to beat Drake, this poll from CBS also indicates that Obama's speech was a success with voters.

New Democratic Delegate Estimates for Pennsylvania, Indiana and North Carolina

Before I get to the estimates for the upcoming primaries, I wanted to note that my last prediction for Mississippi was a 19-14 delegate split in favor of Obama. This was exactly how the delegates ended up being divided in last week's contest, which means the polls helped us predict the delegate split exactly.

Of course, the next primary, Pennsylvania, is not for about a month, which means it is really too early to get a good sense of how the delegates will be divided in that state, much less those that come later. But it is still useful to get a sense of where things stand now and what it would mean for the overall delegate race. This is particularly salient given the number of news stories that have appeared in the past few days suggesting that without a Michigan and Florida re-vote, Clinton has almost no chance in catching up in the delegate race. Is this really the case?

Well, as things stand now, it definitely seems very unlikely that Clinton will make a significant dent in Obama's delegate lead. The polls in Pennsylvania look very good for Clinton, putting her up by 13% in that state, on average. With such a large margin of victory, Clinton would net almost 25 delegates in the state, a significant pick-up. But she would still be over 100 delegates behind at that point and Indiana and North Carolina don't look nearly as favorable toward her candidacy. In North Carolina, the poll-based estimates are that Obama would pick up 11 delegates, almost half of what Clinton would have gained in Pennsylvania. In Indiana, the only poll we have is one taken in mid-February showing 40% for Obama, 25% for Clinton, and a lot of undecided voters. If this lead held up, Obama would net 16 delegates in the state, meaning that he would have erased all the delegate gains Clinton made in Pennsylvania. And after these three primaries, there are very few big delegate contests left.

The bottom line is that it seems nearly impossible to see how Clinton could catch up to Obama's delegate total. Thus, her only real hope at this point is to close the gap enough with Obama so that she can make an argument that not seating the Florida and Michigan delegations are keeping her from being in the delegate lead. If the can do that, it may be an argument that resonates with superdelegates or the credentials committee at the Democratic Convention. For Obama, the goal is to keep enough of a delegate lead that the Florida and Michigan delegations are basically irrelevant (he would win even if they were seated). If he can do this, then he can probably make sure that there will not be a fight at the convention, something he desperately needs to do to give himself the best chance of winning the general election. Of course, August is a long way off...heck, even Pennsylvania's primary seems a long way off at this point. So, I'll keep updating these delegate estimates as new polling data becomes available over the next several weeks.

Thursday, March 20, 2008

Appearance in Good Morning America Piece

For those interested, I appeared briefly in a piece on the debate over a Michigan re-vote and the delegate battle that aired on Good Morning America today:

The Obama Campaign's Primary Predictions Reveal Clinton's Questionable Campaign Strategy After Super Tuesday

You may recall that in the wake of Super Tuesday, the Obama campaign accidentally sent a reporter for Bloomberg a spreadsheet that included the campaign's predictions on how the vote would turn out in each of the remaining primaries and caucuses. The memo was unique in its ability to give us a window into the Obama campaign's expectations about how they viewed the upcoming contests playing out. This was not the campaign's spin designed to set expectations for the news media; these were their actual expectations.

Now that we've had several contests since the memo was released, it is even more clear how costly the month of February was for the Clinton campaign, which struggled to compete at a level that even the Obama people expected. The figure below plots the share of the vote that the Obama expected to get in each state (based on their memo released just after Super Tuesday) against the share of the vote Obama actually received. States falling along the line would be those in which Obama did just as his campaign expected him to do. States falling above the line are those in which Obama's actual voted total exceeded his campaign's expectations, and those below are where he failed to meet those expectations.

The striking point one can take from this figure is that much of Obama's lead was generated by his ability to not just win a lot of states in February, but to win them by larger margins than his campaign expected him to. In Virginia, DC, Washington, Nebraska, and Maryland, Obama performed significantly better than his campaign had expected. His campaign expected to lose Maine narrowly, but he won the state easily.

What happened? Why was Obama able to perform so much better than his own campaign thought he would? One reason for this is the fact that the Clinton campaign appeared to approach the Democratic primaries as if they were winner-take-all. Essentially, she chose not to campaign heavily in states where she didn't think she could win. Yet, in a race where delegates are allocated proportionally, failing to compete may have cost her more delegates in a number of states that she was going to lose. For example, she all but ignored the Potomac Primary states (VA, MD, and DC), and Obama took advantage by not just winning those states, but winning them by large margins that netted him significantly more delegates than his campaign expected. In fact, this happened throughout the month of February until Clinton's "Alamo moment" on March 4th.

Clinton competed heavily in the March 4th primaries, understanding that this was her last opportunity to stay in the race. And what happened as a result? Well, the actual vote totals received by Obama fell back into line with what his campaign had originally expected. In fact, the predictions generated by Obama's strategists a month earlier were exactly correct for Texas, and just barely off for Ohio, Vermont, and Rhode Island. The bottom line: when Clinton competed in a state, the predictions largely matched the actual outcome. Where Clinton got into trouble was failing to compete in several states between Super Tuesday and March 4th, a period during which the Obama campaign exceeded their own internal expectations and built the delegate lead that Clinton is now trying desperately to overcome.

Wednesday, March 19, 2008

Initial Verdict: No Gain for Obama from Speech on Race?

As I noted last night, the Gallup daily tracking polls may give us the first clue into whether Obama's speech has been effective in halting the slide in his support since the Jeremiah Wright scandal first broke last week. Well, today's poll is out and so far, it does not look like Obama's speech helped him. Today's figures show Clinton with a 49-42% lead over Obama, which is a 4 point larger gap than existed in yesterday's poll. Support for Obama has dropped from 50% to 42% in the last 5 days while Clinton's support has increased from 44% to 49%.

The Gallup daily report also provides some insight by noting: "Obama's campaign has been plagued by controversial remarks made by his former pastor, Rev. Jeremiah Wright. Obama delivered a major speech on race Tuesday to try to move beyond the controversy. The initial indications are that the speech has not halted Clinton's gaining momentum, as she led by a similar margin in Tuesday night's polling as compared to Monday night's polling."

What will the political impact of the Obama speech be?

By most accounts (including, admittedly, my own), Obama's speech on race today was significant for the graceful and forthright way that it addressed the issue of race in America. But the impetus for the speech was political, and it is less clear what the political effects of the speech will be. Note from the Gallup tracking poll (pictured below) that Obama's standing in the Democratic nomination race has declined steadily since the videos of Jeremiah Wright's sermons started making the rounds on television news programs last week (I've denoted the timing with the red horizontal line).

Gallup's tracking poll uses a three day rolling sample. Thus, the last day shown (March 17th) included interviews conducted on Saturday, Sunday, and Monday. Tomorrow's release will be the first to include interviews taken after Obama's speech on race (those interviews were conducted this evening), though they will only comprise one-third of the sample. By the end of the week, we'll have survey points with interviews taken entirely after the speech. It will be interesting to see what happens in tomorrow's survey (does the decline level off, or even tick back up a bit?), but especially the following two days. Stay tuned...

Tuesday, March 18, 2008

Florida and Michigan Delegations: Where Obama and Clinton Could Find Common Ground?

The Florida re-vote is likely dead. The Michigan re-vote is clearly in jeopardy. The Obama campaign is trying to run out the clock on this since no action probably advantages him. But does it really? Two important points to consider:

1) The earlier DNC vote to sanction Michigan and Florida is not necessarily binding. Anything could shift at the convention, even, potentially, the seating of the full MI and FL delegations.

2) If Obama is viewed as being a roadblock to having the Michigan and Florida voters having a say in this election, it could hurt him in a general election.

At this point, the Clinton campaign is desperate. It recognizes that without re-votes, Michigan delegates almost certainly won't be allowed to vote on the nominee (how can you justify seating a delegation when Obama was not even on the ballot?) and Florida likely will not be given voting rights either (if you can't seat MI, how do you seat FL?). Because of this position, the Clinton campaign would likely agree to almost any compromise. So, what should the Obama campaign offer?

If you are Obama, why not offer the following?

1) Superdelegates from both states should be disqualified. Many of the superdelegates are directly to blame for this mess and most from Michigan and Florida are at least complicit in what happened. The sanction should remain in place for them, something the Obama campaign would favor since Clinton currently holds an edge in endorsements among the Michigan and Florida super-delegates.

2) Push for a party-run, internet/mail re-vote in Michigan. The Michigan Democratic Party was already planning for such an election before the state legislature moved the state's primary date up (see the plans for this primary here), so it should not be too difficult to dust off those plans and put them back into action. As I've noted in an earlier post, Obama and Clinton would likely split the delegates in Michigan if there was a re-vote. An internet primary may even help him perform better, since it would encourage more participation from younger voters who have been more supportive of his candidacy. Bottom line: pushing for an Internet/mail re-vote in Michigan allows Obama to make the claim that he is trying to make sure the voices of Michigan voters are heard and would not likely put his delegate lead in danger.

3) Push for half of Florida's pledged delegates to be seated. This would be the toughest decision for the Obama campaign. They would essentially be giving Clinton a net pickup of 19 delegates in Florida. However, a re-vote might have produced a Clinton edge that would have been much greater. For 19 delegates, Obama would be buying some goodwill from the voters of Florida as well as some insurance against other options that could cost him more delegates.

All in all, this plan would likely not cost Obama any more than 20 or so delegates, not nearly enough for Clinton to seriously threaten his advantage among pledged delegates. In addition, the move would win him some goodwill among voters in Michigan and Florida, something that he might be very glad to have if he does end up as the party's nominee. So, why not offer this compromise? Two reasons. First, it is hard to give up even 20 delegates in this tight race, particularly when it looks as if you won't have to give up any. Second, putting Michigan back in play at the end of the calendar likely means that Clinton would have all the more reason to stay in the race until the end. Obama may be hoping for a knock out punch before June, and putting Michigan back on the calendar may end any chance of that happening.

What do you think...would this compromise work? Should it work?

What Happens to Florida's Delegates Now?

The news today is that there will not be a re-vote in Florida. This news is obviously not good for Clinton in any way. First, let's start with what might have happened in a re-vote. According to a Rasmussen Reports survey conducted a couple of weeks ago, Clinton held a 55%-39% advantage over Obama if a re-vote were held. If her lead held up, it would have given Clinton a net gain of about 30 delegates.

However, if we assume a re-vote is not going to happen, then what are the other options available? Here are several possibilities, and what they would mean for the delegate count:

1) Not seat the delegation. The DNC has already said that the delegates elected in Florida will not count. If this decision stands, then neither candidate would receive any delegates from Florida, a scenario that clearly favors Obama.

2) Seat the full delegation that was elected in January. This seems unlikely to happen. However, if it did, Clinton would receive 105 delegates, Obama 67, and Edwards 13. That would give Clinton a net gain of 38 delegates. A net gain like this would help Clinton significantly cut into Obama's lead; of course, this is precisely why this solution is so unlikely. However, you never know what might happen at the convention if these delegates would make the difference.

3) The Nelson plan. Florida Senator Bill Nelson has suggested that the DNC do what the RNC did with Florida: only seat half their delegates. If this were to happen, each candidate's delegate total from the January primary would be cut in half. That would mean that Clinton would receive 52.5 delegates, Obama 33.5, and Edwards 6.5.With this proposal, Clinton would take a net gain of 19 delegates.

4) FL congressional delegation plan. This plan is a bit unclear at the moment. Evidently it would take into account both the Florida vote as well as how the vote has gone in other states since Florida voted. It is obviously difficult to see how this would translate into delegates, but it seems likely that it would result in a net advantage for Clinton that would be roughly similar to that in the Nelson plan.

At this point, the Obama campaign seems to be happy to allow the Florida situation to languish. After all, the longer those votes don't count, the longer Obama maintains a significant edge over Clinton. But wouldn't it make sense for Obama to come out in support of a modified Nelson plan? Specifically, his campaign could argue that the pledged delegates from Florida be cut in half and that all superdelegates from Florida should not be given a vote (something I argued earlier here). Even though Obama's lead is relatively slim, he could probably afford to give up 19 delegates to Clinton. And by banning Florida's superdelegates, he'd know that he wouldn't lose any more ground to her in that state. Essentially, Obama would be buying an insurance policy by endorsing the plan. For the price of 19 delegates, he can make sure that he doesn't take a bigger hit in the state if something unexpected happens at the convention (leading to the full delegation being seated). There is also the point that the good will and good press that such a move would earn him would probably more than make up for the delegates he'd give up to Clinton in Florida by accepting a reduced delegation. Will Obama do this? Perhaps not. Should he? What do you think?

Sunday, March 16, 2008

Is Nancy Pelosi a Closet Obama Supporter?

As a high profile party leader, Nancy Pelosi has refused to support either Obama or Clinton for the Democratic nomination. But twice last week she seemed to have been subtly helping the Obama campaign.

Earlier last week, Pelosi essentially helped to put an end to the "dream ticket" story by saying that she simply did not see it happening. The Clinton campaign had been pushing this story as a way to signal to voters that they could get her experience without losing out on the energy that Obama offers. Of course, the "dream ticket" talk was not helpful to Obama and his campaign had been doing its best to kill the story throughout the week. But nothing his campaign did was quite as effective as Pelosi's definitive statement that it was not going to happen.

Then, on Sunday, Pelosi reiterated the point that superdelegates should not overturn what the voters decide. This is, of course, the very point being made by the Obama campaign since Obama is unlikely to relinquish the pledged delegate lead before all is said and done. Pelosi's comments are essentially the kind of thing you would expect to hear from an Obama surrogate.

Is all of this coincidence? Probably. But the Clinton campaign cannot be happy with the way that Pelosi has been acting like a pseudo-surrogate for Obama during the past week.

Evidence of the Uncertainty of Delegate Math: Obama Picks Up Delegates in California and Iowa

One of the interesting stories of the week has been the fact that Obama has picked up more delegates this week from contests that happened a month or more ago than either candidate has picked up in contests held in March. First, there was this story from Carl Bialik noting that while all of the major news organizations had estimated that Clinton would win 207 delegates in California, the total will actually be 203. That means that 4 delegates were subtracted from Clinton's tally and added to Obama's, for a net Obama gain of 8 delegates.

(One other note about California: these adjustments mean that my delegate predictions were only off by 2 delegates, rather than off by 7 as it originally appeared.)

Then, on Saturday, the next step in the Iowa caucuses produced a dramatic change in the delegate count from that state. The networks essentially use the initial delegate counts from the precinct caucuses on January 3rd to estimate how many delegates each candidate would eventually receive from the state. However, partly because of some changes due to aggregation and partly due to the realignment of a significant number of Edwards delegates, the delegate count tilted heavily in Obama's favor after the county conventions on Saturday. The original estimate from Iowa was that Obama would win 16 of the state's delegates, Clinton would win 15, and Edwards 14. However, Obama was the big victor in the county conventions on Saturday and the new estimates are that he would win 25 of the state's delegates, Clinton would take just 14, and Edwards 6. Because Clinton actually lost a delegate and Obama picked up 9, that amounts to a 10 delegate net pick-up for Obama in Iowa over the weekend.

Altogether, Obama added 13 delegates to his total in the past few days while Clinton lost 5 delegates, an overall net gain of 18 delegates for Obama. To put that total in perspective, Clinton only picked up 9 delegates in Ohio on March 4th.

So, should we expect to see this kind of movement in other caucus states as they move from the precinct to the county level and beyond? Probably not. Most of Obama's additional delegates came from former Edwards supporters and there were many of these in Iowa since he actually finished second in the state. The only other state to hold a caucus before Edwards dropped out was Nevada, where he won only 4% of the vote. In the remaining caucus states, you may see an adjustment of one or two delegates in either direction because of how votes are aggregated as the process moves forward, but nothing like the 10 delegate swing you saw in Iowa this weekend.

Friday, March 14, 2008

What About Those Superdelegate Add-ons?

One of the interesting quirks about the Democratic superdelegates is that many of them do not even exist yet. This is because each state is allowed to name anywhere from 1 to 5 extra superdelegates as superdelegate add-ons. There are 76 of these add-on superdelegates, and according to the Democratic Convention Watch site, only three of these have been named.

Twenty-four states determine the add-on superdelegates at their state conventions: ND, AZ, SC, IL, OH, CO, KS, NV, AK, WY, ME, MS, KY, MN, TX, MT, ID, IA, VA, WA, NC, OR, PR, NE.

Now, if the add-ons are named at the state conventions, then presumably the Clinton and Obama campaigns could be mounting efforts to make sure that the individuals who are chosen to be superdelegates at these state conventions support them and not the other candidate. It is anyone's guess which candidate's supporters control the party conventions in most of the states, but where it does become a little more clear is in caucus states. In caucus states, the precinct caucuses are the first step in selecting individuals who will eventually make up the attendees to the state convention. Thus, if a candidate won a particular state's caucuses, then it is likely that the candidate's supporters will also control the convention. And if that is the case, then that candidate's supporters could be sure to select a superdelegate add-on that will support their favored candidate.

So, why is this important? Well, keep in mind that Obama has done very well in caucuses. He won Iowa (though not a majority), Colorado, Idaho, Kansas, Minnesota, North Dakota, Alaska, Nebraska, Maine, Wyoming, Texas, and Washington. In each of these states, the state convention selects who the add-on superdelegates will be. Texas has 3 add-ons, Washington has 2, and each of these remaining states has one apiece. Thus, assuming Obama's supporters control the state conventions in each of these states, he would pick up 15 superdelegates by virtue of his supporters choosing add-ons that they know will support him.

An additional five states allow the delegates elected at the district-level in the primaries to determine the superdelegate add-on: CA, NH, UT, VT and IN. Clinton won in CA and NH, which have a combined 6 add-on superdelegates. On the other hand, Obama won UT and VT which each have 1 superdelegate add-on. (Indiana has not yet voted).

Thus, based on this information, we can expect that Obama would pick up at least 17 superdelegate add-ons while Clinton would pick up 6. To my knowledge, these superdelegate add-ons are not presently included in the running totals provided by news outlets, so there is reason to think that Clinton's superdelegate advantage is at least 10 fewer than is commonly being reported.

Many of the remaining add-ons are chosen by state party executive committees, where it is more difficult to figure out which candidate is favored. It will be interesting to see how these executive committees select these superdelegate add-ons, since these decisions could have an important influence on the state of the delegate race.

Wednesday, March 12, 2008

Solution for a Michigan and Florida Democratic Re-vote: Bench the Superdelegates?

The debate over what to do with Michigan and Florida continues on with no clear solution in sight. The Obama campaign rightly argues that you cannot seat delegates elected in contests that were not supposed to count and in states where the candidates did not even campaign (and when he wasn't on the ballot in the case of Michigan). The Clinton campaign rightly argues that it is problematic to have the citizens of Michigan and Florida lose their voice in the process because of the actions of their state parties and their elected officials. In the meantime, some oppose a re-vote because it is an expensive option and may reward the actions of the state officials by allowing their states to have a key role in determining the ultimate nominee despite the fact that they broke the rules. Thus, the problem for the DNC is how to protect the citizens of Florida and Michigan without rewarding the actions of those in the state who broke the national party rules in the first place.

One solution may be to allow a re-vote in both states and seat the pledged delegates elected in those contests, but maintain the penalty against Michigan and Florida's superdelegates. The superdelegates, after all, are the party leaders and elected officials from these states that pushed these primary dates up in the first place (or at least supported the move). Stripping the Michigan and Florida superdelegates of their votes would punish those most responsible for the fact that the states broke the DNC's rules while allowing the states' voters to still have a say in the process. Those superdelegates could still attend the convention, but would not be allowed to cast a vote for the presidential nominee.

Politically, this solution may win the support of both campaigns. The Clinton campaign would likely support this plan since it would assure that Michigan and Florida were in play and Clinton's support in Florida would presumably allow her to pick up several delegates. The Obama campaign may also go for this solution, since it assures that Clinton will not get the support of the Florida and Michigan superdelegates when those states are re-instated into the process. According to the Democratic Convention Watch website, Florida has 26 superdelegates and Michigan has 28, for a total of 54 superdelegates between the two states. Fifteen of these superdelegates presently support Clinton while only 5 support Obama. The Obama campaign will be far more agreeable to a plan that doesn't immediately shave 10 delegates off of their lead before anyone in either state casts a (second) vote.

Besides the political benefits of such a solution, this plan would also make a lot of sense. The citizens of Michigan and Florida should not be punished for the actions of their elected officials, but this plan punishes those elected officials while still allowing the citizens to participate in the nomination process.

What do you think? Would this be a workable solution to MI/FL problem?

Tuesday, March 11, 2008

Accuracy of March 4th Democratic Delegates Predictions (And Updated Estimates)

With most of the delegate totals from the March 4th contests now known, it is time to once again check up on how close my predictions were to predicting the actual delegate count from the four primaries held on that day. In the table below, I present the predictions I made (based on survey averages) compared to the actual delegate totals that the candidates took from each state. Keep in mind that we still do not know how 9 delegates from the Texas caucuses will be allocated.

As you can see, in Ohio and Vermont, the pre-election estimates I generated got the delegate totals exactly correct. In Rhode Island, Clinton's victory was bigger than the polls predicted, which meant my estimates were off by two delegates (she took 13 delegates from the state, rather than the 11 I predicted). Finally, in Texas, the estimates were quite close to the final tally. Even if one candidate gets all of the remaining 9 delegates, the estimates will have only been off by 7 delegates; if those delegates are allocated roughly evenly, then the estimates will have been off by 1 or even exactly correct. The bottom line: the survey-based estimates I generated predicted that Clinton would pick up 10 delegates from the March 4th primaries and those estimates will be off by no more than 5 (Clinton will finish anywhere between a net loss of 4 delegates and a net gain of 5 delegates, depending on how those final 9 delegates are allocated).

One thing that is notable about these estimates is that they were so close in Texas, even though I and others suggested that the polls would not be a very accurate guide to understanding how delegates would be allocated in that state. Some districts in Texas had a lot more delegates available than others and the caucuses threw another hiccup into the estimates, but at the end of it all, the polls did a reasonably good job, even in a state with such an odd system.

So, what do the surveys say about the primaries coming up in the next two months? Currently, there have been no polls in Indiana, but we do have at least three surveys to work from in each of the other three states. Based on these surveys, Obama will pick up 5 delegates in Mississippi, which, combined with his win in Wyoming over the weekend, will mostly cancel what Clinton picked up on March 4th. Clinton is still favored to win Pennsylvania by a wide margin, which would give her a net gain of about 20 delegates in that state, but North Carolina favors Obama by a fairly significant margin, which would allow him to make up about half of the delegates Clinton would pick up in Pennsylvania. Bottom line, as the polls currently stand, it is going to be difficult for Clinton to make up significant ground in the delegate count in the coming months.

Saturday, March 8, 2008

Delegate Estimates for Upcoming Democratic Primaries

It looks as though Obama is on his way to another caucus state victory in Wyoming today. Based on the current vote totals from the state (with 78% of the precincts reporting), Obama would likely take 7 delegates from today's caucuses compared to 5 for Clinton.

Over the next two months, there will be just four primaries, beginning with Mississippi on Tuesday, Pennsylvania on April 22nd, and Indiana and North Carolina on May 6th. But why wait to start trying to get a sense of how many delegates each candidate will pile up between now and mid-May? Using averages of the surveys conducted in the upcoming primary states, I've allocated the delegates as the polls would suggest they will be divided. This method has been fairly reliable in most of the earlier primaries (see here and here). Currently, there have been no polls in Indiana, but we do have at least three surveys to work from in each of the other three states.

As you can see, Obama is expected to win Mississippi and the polls suggest that he would have a net gain of 5 delegates from that state. Clinton is presently running strong in Pennsylvania and if the survey averages are correct (and stay where they are), she would pick up 22 delegates in that state. North Carolina is another Obama state, and he is currently expected to pick up 11 delegates there. While we have no survey data on Indiana yet, there is some reason to believe that Obama and Clinton would probably run pretty evenly there. If they did, then by mid-May, Clinton would have only cut into Obama's lead by 5-10 delegates.

What does this mean? Well, unless something drastically changes in the race, it means that Clinton desperately needs Florida to have a re-vote if she has any chance of significantly cutting into Obama's lead in pledged delegates. Rasmussen Reports recently polled voters in Florida and Michigan to see which candidates voters would prefer if those states held another primary. In Michigan, the candidates were tied, which means Obama and Clinton would split the delegates from that state relatively evenly. Florida, however, offers a big pick-up opportunity for Clinton. According to the poll in that state, Clinton leads Obama 55-39%, which would allow her to pick up around 30 delegates from that state. That still would not over-take Obama's pledged delegate lead, but it would allow her to get closer and the closer it gets, the more important those superdelegates become. Ironically, the news reports over the past week have suggested that Michigan is more likely to hold a re-vote than Florida, but we will see.

I'll continue to update these estimates over the coming weeks (or months)...

Thursday, March 6, 2008

Who is Winning the Swing States?

Senator Clinton's campaign regularly notes that Senator Clinton is winning the big (re: important) states while Senator Obama is winning in places like Nebraska and Idaho: places not at all likely to turn blue on any television network's election night maps. What about this claim? The real argument should not be about state size, but about who is winning the swing states. Indeed, the argument also ought to consider whether the candidate is winning swing states that went the other way in 2004. When we do that, Sen. Clinton's argument evaporates to a point...but Senator Obama does not improve enough to answer the question, "but who is winning the swing states?" with a definitive pronouncement either.

First, 12 states fit into the 5% rule I have arbitrarily developed based on the Bush-Kerry vote difference in each state in 2004. The Democrats have six to hold and six to flip into their column. Leaving Florida and Michigan aside (at least until June!), Senator Obama has won primaries in swing states with 36 electoral votes (16 of which were in GOP hands in 2004). Senator Clinton has won primaries in swing states with 34 electoral votes (30 of which were in GOP hands in 2004). So, Senator Obama has won swing states with a larger number of electoral votes than has Senator Clinton, but Senator Clinton has more more swing states with electoral votes that can flip to the Democrats' side. As the whole race has been, it is pretty much even steven. This makes Pennsylvania, all the more important for each candidate. The same is true for potential do overs in Florida and Michigan.

Obama fans may point to my leaving of Missouri off of this list, which would give Obama 11 more electoral votes in the "Flip" column. Obamaniacs can also point to a new survey in the Lincoln Journal Star, which suggests that Obama would only narrowly lose Nebraska to John McCain while also winning two of Nebraska's electoral votes. Nebraska awards two of its five votes for statewide election victory and each of the remaining three for victory in one of Nebraska's three congressional districts. Obama is currently winning two of those districts, according to the poll. Interesting stuff.

What Would Happen in a Michigan Democratic Primary Re-Vote?

With the Democratic nomination race extending indefinitely, the political pundits are buzzing about the possibility of a re-vote in Florida and Michigan. For what its worth, the possibility seems greater that Michigan will pull this off. After all, the Michigan Democratic Party was already prepared to host their own party-run primary (complete with voting by mail and Internet) before the state legislature got involved and changed the plans. They also have a more difficult case to make with the credentials committee since Obama's name did not even appear on the ballot.

So, what would happen if Michigan pursued such a primary? Well, a good place to start is with what happened when they held their first primary in January. Remember, Hillary Clinton was the only major candidate on the ballot in Michigan (Kucinich, Dodd and Gravel's names also appeared). However, Clinton took just 55.3% of the vote statewide, compared to 40% for "uncommitted." Uncommitted actually won in Washtenaw (Ann Arbor...home of the University of Michigan) and Emmet Counties. This primary also was held following Clinton's victory in New Hampshire (and before Obama's win in South Carolina). Given that Clinton took just 55.3% of the vote in a contest where neither Obama nor Edwards appeared on the ballot, it seems unlikely that she would receive a higher percentage in a re-vote. If we assume that Clinton would receive 55% to 45% for Obama in Michigan, then Clinton would gain about a 12 delegate edge in the state.

Now, we can get an even better sense of how things might go in Michigan from the exit polls that were taken in January. One of the questions asked on the exit poll was who the respondent would have voted for if all the candidates had been on the ballot. We can look at these figures in two ways. The simple way is just to see how these voters broke down in the exit poll. 46% said that they would still vote for Clinton, 35% would have voted for Obama, and 12% would have voted for Edwards. If you split the Edwards vote evenly between Obama and Clinton, then the vote would be 52-41% in favor of Clinton. This would yield a 16 delegate edge for Clinton in the state.

Another way to look at the exit polls is to use them to get a sense of how many Clinton voters would have voted for Obama and assume that anyone who didn't vote for Clinton in January (when she was the only candidate on the ballot) would not do so now. Based on the exit polls, 18% of those who would have voted for Obama if his name had been on the ballot ended up voting for Clinton. So, this suggests that about 6% of those who voted for Clinton in January actually wanted to vote for Obama. If we subtract that 6% from Clinton's total and give the remaining vote to Obama, the split in Michigan would be about 50-50. In this case, Clinton may not pick up any delegates.

Of course, a lot has changed since January and a party-run primary with early voting available on the Internet or by mail could change things significantly. In addition, turnout among Obama supporters may have been suppressed given that they knew that his name would not be on the ballot. Nevertheless, based on what happened last time, it appears as though the best case scenario for Clinton in Michigan would be a 16-20 delegate edge and it is likely that she would get less than that or even no edge at all. It all depends on how you crunch the numbers.

UPDATE: Just to elaborate a bit on what may happen with an event that is run by the Michigan Democratic Party, let me refer you to a page that was set up by the Michigan Democratic Party when they were originally set to sponsor their own primary on February 9th (before the MI state legislature got involved). In short, the MDP had already arranged for a party-run primary (technically called a caucus, but not really a caucus). Thus, it stands to reason that they could simply implement the plan they originally had but for a later date. The key parts of the document are as follows:

"Voting centers will be open between 10 A.M. and 4 P.M. EST. You can vote at any voting center in the county where you live anytime during those hours. ... Voters must complete a ballot, including a public declaration that they are a Democrat and are or will be a registered voter before the November election."

If the party followed this plan, voters would have to state that they were a Democrat. Keep in mind that Michigan does not have party registration.

"Yes, you can vote by mail if you qualify. ... Unlike traditional absentee voting in Michigan you do not need a reason such as age or disability to vote by mail. Applications must be received by February 2, 2008 at 6 P.M. (or earlier if the Caucus date is moved).A ballot will be mailed via the U.S Postal Service to those who apply and are registered to vote."

And this...

"Yes, you can vote over the Internet if you qualify. Follow the same application procedures as voting by mail."

The Michigan Democratic Party held an event just like this in 2004 so it would probably be relatively easy for them to print ballots and set up internet voting in short order (they had already been planning on this as late as last summer). Setting up in-person voting sites may be a bit more trouble, but they could likely pull that off as well. Simply put, because they had already planned to hold an event like this, the Michigan Democratic Party should be in a position to pull off a primary-like event, complete with early voting (by mail or Internet).

UPDATE 2: A Rasmussen poll finds that Obama and Clinton are tied in Michigan if that state were to hold a re-vote. This fits with the above analysis that took 6% off of Clinton's total from the first primary since those supporters actually preferred Obama.

Who will decide if there are Michigan and Florida do-overs?

Last night, members of the Michigan and Florida congressional delegations met to strategize about potential do-over primaries and caucuses in their states. But DNC members and state party leaders - not members of Congress - will ultimately decide if, how, and when Michigan and Florida will hold new delegate selection contests. The governors of both states have already indicated that if there are new primaries or caucuses, they will be funded privately by the state and/or national parties.

If the state parties decide to hold new contests, they would need to get approval from the DNC. According to the 2008 Delegate Selection Rules, the DNC Rules and Bylaws Committee has jurisdiction over all state delegate selection plans.
Rule 19 E states: "The DNC Rules and Bylaws Committee shall retain jurisdiction over the approval of amendments to state Delegate Selection Plans and state delegation compliance with equal division requirements, even after the Convention Credentials Committee assumes jurisdiction over challenges to the credentials of delegates."

Since leaders in Michigan and Florida are talking about staging new delegate selection events (probably caucuses), these could be viewed as "amendments" to the original Michigan and Florida delegate selection plans, and therefore under the jurisdiction of the RBC. Right now, no one seems to be talking about challenging the credentials of the Michigan and Florida delegates, since the DNC does not currently recognize any delegates from Michigan or Florida. A challenge to the state delegations would fall under the jurisdiction of the convention credentials committee. But staging new caucuses in Florida and Michigan seems to pretty clearly fall under the jurisdiction of the RBC. Coincidentally (or maybe not), the co-chairs of the Rules and Bylaws Committee are James Roosevelt Jr. and Alexis Herman, who are also two of the three chairs of the Credentials Committee.

The members of the Rules and Bylaws Committee are all DNC members, and therefore all unpledged "super" delegates. Harold Ickes, a top Clinton advisor, is a member of the Committee.

Since we are starting to hear more about the Credentials Committee, I'll say a few words about the convention committees in general. The
Democratic National Convention is governed by three committees: Credentials, Platform, and Rules. Each committee has 186 members, 161 of whom are chosen by each state's delegates. The remaining 25 members of each committee are reserved for "Party Leaders and Elected Officials" (called "PLEOs"). These PLEO members were appointed by DNC Chairman Howard Dean in January 2008, along with three chairs for each committee.

The DNC has put together a list of names and short bios of chairs and PLEO members of the standing committees.

For links to the DNC rules and a list of RBC members, please visit the CCPS Party Conventions Research Page.

Updated Predictions on Unpledged Democratic Superdelegates

One thing that the March 4th primaries guaranteed is that the debate about Democratic superdelegates will continue for a while longer. While Clinton presently leads Obama among super delegates who have pledged their support, about half of the superdelegates remain uncommitted. So who will these unpledged superdelegates support?

I have now updated the predictions for which candidate unpledged Democratic superdelegates are likely to support. As before, I use information about the superdelegates who have committed to a candidate to generate predictions for 291 unpledged superdelegates. I exclude superdelegates from DC and the territories because we lack complete data from those areas, and from IL, NY, and AR because superdelegates in those states have nearly unanimously cast their support for their native son/daughter. As always, information on the superdelegates is provided by the Democratic Convention Watch site. You can find more about they methodology I use here. Check out the distribution of predicted support among unpledged superdelegates below.

Superdelegates who are between 40% and 60% likely to vote for Clinton/Obama are labeled as "unclear." There are a lot of superdelegates in this range, 109 to be exact. There are 103 unpledged superdelegates who are at least 60% likely to vote for Obama; 84 unpledged superdelegates are at least 60% likely to vote for Clinton. What these numbers suggest is that neither candidate is likely to build a large margin among superdelegates.

If these estimates are even remotely accurate (so far, the model has performed reasonably well, predicting over 70% of the Superdelegates who have endorsed in the last several weeks), then it is unlikely that Clinton would be able to take a net advantage of more than a 100 superdelegates once all is said and done. This means that if Obama is able to build more than a 100 delegate lead among pledged delegates, it is unlikely that Clinton could make up that advantage with superdelegates. You can see the estimates for each unpledged superdelegate here. The most likely Obama supporters include Rep. Dennis Moore (KS) and Gov. Dave Freudenthal (WY). The most likely Clinton supporters include Sen. Barbara Boxer (CA) and Reps. Capps, Davis, and McNerney (CA).

UPDATE: Just hours after I posted these predictions, the Democratic Convention Watch Site added Barbara Boxer to the Clinton column. Their decision to do so is based on this article. As noted above, the model predicted her support for Clinton.

Wednesday, March 5, 2008

March 4th Primary Wrap Up

So, where do we stand now?

Well, we know one thing we already knew...John McCain is the Republican nominee. One thing we didn't know...he gets at least a seven week free pass while the Democrats battle it out.

And for the Democrats...

As I noted earlier, Rhode Island and Vermont end up as a push...Obama picks up 3 delegates in Vermont and Clinton picks up 3 in Rhode Island. Clinton should pick up around 15 delegates from Ohio. That leaves Texas. Currently, Clinton is projected to pick up 1 delegate in the Texas primary (this just shifted). It is still possible that Obama could pick up slightly more delegates based on his strength in certain districts that get more delegates, so we'll have to wait till all the votes are in before we really know.

The Texas caucuses are even more of a mystery, but it seems that Obama will not fare as well in these caucuses as he has in other caucuses (i.e. he won't win by a 2-to-1 margin). However, he is presently up 55-45% with 25% reporting. If that margin holds, then Obama would gain 5 or 7 delegates there. Thus, Obama will likely pick up a delegate or two in Texas. Bottom line, Clinton should pick up about 15 delegates tonight, probably not too many more than that.

Clinton finished strong...the exit polls showed her with a big advantage among those who decided late, and the early vote in Texas went to Obama while the election day vote was won by Clinton. So, how will this affect the way the Obama campaign approaches the coming weeks? Will they go on the offensive more? And can Clinton stay on the offensive for 7 weeks without suffering any backlash?

What is the story going forward? Obama should win in Wyoming and Mississippi which may help him re-establish some momentum, but after Mississippi we have 6 weeks until Pennsylvania. We will need something to talk about during all of that time, and the two main stories will be superdelegates and Florida and Michigan.

With regard to Florida and Michigan, I made the point way back on January 29th that Florida and Michigan could very well schedule new events to essentially re-vote. It appears as though this will be a major storyline in the coming days since it does not look like this campaign will be ending any time soon. I suspect that Dean may get the campaigns together to work something out on this front in the coming weeks. It will be interesting to see what that solution is, particularly since the Clinton campaign will want to argue for primaries while the Obama campaign will want caucuses.

With regard to superdelegates, I have been using statistical models to predict which candidates the unpledged superdelegates will support. So far, the models have been right over 70% of the time. These models currently show that there are more unpledged superdelegates out there who would likely support Obama than there are supers who would likely support Clinton. But we will be updating the model and generating new predictions some time tomorrow, so stay tuned.

Finally, we had well over 700 visits on the blog today, our biggest day since Super Tuesday. Thanks for everyone who visited (and come back!).

Tuesday, March 4, 2008

March 4th Primaries Live Blog: Late Late Edition

Will Clinton's lead continue to widen in TX, or will it narrow. I'm looking at the counties where the most precincts haven't come in yet:

Williamson County (1 of 100 precincts in)...Obama currently ahead 25k-19k there.
Tarrant County (356 of 639 precincts in)...Obama ahead 95k-81k.
Travis County (177 of 211 precincts in)...Obama ahead 106k-62k.
Jefferson County (1 of 117 precincts in)...Obama ahead 14k-9k.
Harris County (641 of 875 precincts in)...Obama ahead 195k-145k.
Denton County (69 of 123 precincts in)...Obama ahead 24k-18k.
Dallas County (248 of 695 precincts in)...Obama ahead 114k-65k.

Bexar County (468 of 623 precincts in)...Clinton ahead 101k-75k.
Cameron County (40 of 99 precincts in)...Clinton ahead 22k-10k.
El Paso County (93 of 171 precincts in)...Clinton ahead 58k-25k
Jim Wells County (1 of 22 precincts in)...Clinton ahead 3,000-1,000.
Upshur County (1 of 21 precincts in)...Clinton ahead 3,500-1,900.

Bottom line: more of the precincts that have not yet reported appear to come from Obama areas, so don't be surprised if you wake up and the margin is a little slimmer than it was when you went to bed.

Just to finish the delegate math ahead of us. The last two polls in PA put Clinton up by 4-6%. If that margin stays where it is, Clinton would only pick up somewhere around 10 delegates in Pennsylvania.

The last several polls in North Carolina put Obama up by 10% or more. If that margin held, Obama would pick up 10-15 delegates in North Carolina, thereby off-setting anything Clinton picked up in Pennsylvania.

Of course, PA is 7 weeks away and NC isn't until May. Given how much we saw the polls move just in the last few days in TX and OH, a lot could happen in those states before voters head to the polls.

By the way, there are no polls (that I know of) in either Wyoming or Mississippi. If we assume that Obama wins Mississippi with a similar share of the vote as he captured in neighboring Alabama, he would pick up 3 or 4 delegates. Let's assume that he wins Wyoming 2-to-1, as he has in the other western caucuses. He would net another 3 or 4 delegates from that contest. That means he would have gained back 6 to 8 delegates by this time next week. That may only be about 1/3 to 1/2 of what Clinton has gained tonight. Of course, it is important to remember, Obama has a lead of about 150 pledged delegates, so Clinton can't catch up by gaining 20 one week, but giving back half the next.

The TX Sec. of State site is presently estimating that Clinton will take 65 delegates to 61 for Obama. A net gain of 4. We'll see if she expands on that any more. As noted earlier, VT and RI are a wash. It is unclear how the Texas caucuses may break down, but the preliminary evidence from below suggests that they may mirror the primary vote. So, it will come down to the margin in Ohio. She currently is up 55-43%. If that holds, then she would pick up 15-17 delegates from Ohio. That would mean that her total gain from the night would be somewhere between 20 and 25 delegates.

In Texas, Obama won the early vote 50.6-47.7%. The election day vote is almost exactly flipped right now, 50.7-47.3% in favor of Clinton.

MSNBC calls Texas for Clinton. She has officially won the night, and this campaign certainly goes on for a least a month and a half longer.

Looking at the county-by-county vote totals in Texas. Obama won big among the early vote in Harris County, but so far he is only winning the vote that was cast today narrowly. Is this a sign of how the campaign turned in the last few days?

Speaking of which, the exit polls show that among those that decided in the last three days (10% of the electorate in Texas) Clinton won 69-31%. And she won those who decided today (11% of the electorate) 55-45%. That is a big advantage for Clinton among late deciders.

Keith Olberman mentions that Obama called Clinton tonight. MSNBC pundit quips, "I'm surprised he didn't wait till 3am to call."

And what about those caucuses? Well, the Texas Democratic Party is providing some initial vote totals from a handful of senate districts.

SD 8: Obama 62%, Clinton 37% (the primary vote there is presently 59-41% for Obama).
SD 14: Obama 66-33% (primary vote is 63-36% for Obama).
SD 16: Obama 60-39% (primary vote is 53-46% for Obama).
SD 29: Clinton 75-24% (primary vote is 69-29% for Clinton).
SD 30: Clinton 59-40% (primary vote is 61-36% for Clinton).

Bottom line: Obama doesn't appear to be dominating these caucuses the way he has dominated caucus-only states. In these five districts, they appear to be going roughly like the primary vote.

The Secretary of State's website in Texas is providing a running delegate breakdown along with their report of the vote totals. Though Clinton leads by about 50,000 votes, she is presently down by two delegates.

It is increasingly looking as though the VT/RI combo will be a wash for the candidates. If the current leads hold, Obama will net 3 delegates in VT and Clinton will net 3 delegates in RI.

He just started speaking, but it sounds as though Obama may be breaking out a new speech tonight. For many, his New Hampshire speech was the strongest of his campaign. Will he pull another well-received speech out of a rough night?

If you are a Democrat the problematic number for you is 7. That is how many weeks before Pennsylvania votes. Which means , at a minimum, there are seven weeks for McCain to raise money and benefit from the increasingly negative Democratic campaign.

Clinton is out speaking and sounding very much like the upstart underdog. She begins by noting that she was counted out but refused to be knocked out. She then goes on to say, "we're just getting started." Don't forget, McCain's nomination campaign just ended.

Update on the "big four" districts in Texas that hold the large delegate counts:

SD 14 (8 delegates): Obama 64%, Clinton 35%
SD 13 (7 delegates): Obama 77%, Clinton 22%
SD 23 (6 delegates): Obama 77%, Clinton 22%
SD 25 (6 delegates): Obama 67%, Clinton 31%

Just so you don't think I'm forgetting about the caucuses, the Texas Democratic Party website presently has this notice up: "As of 10pm, we are not yet reporting Precinct Convention results."
The early vote totals just came in from Harris County (Houston). Obama got 110k of those votes to 66k for Clinton.

Wither the Goalposts?

Earlier tonight, MSNBC's Keith Olbermann joked that those paying attention to the Democratic primary should not stand near the goalposts as they keep getting moved. In Chris Matthews' interview with Pennsylvania Governor Ed Rendell tonight, the Democratic institution (and Clinton supporter) said more than once that "If Clinton wins Ohio and Texas..." she can go on. He seemed to be clearly signaling that if Obama wins one of the two - and Texas has to be the best bet - that pressure on Clinton to drop out may ratchet up a bit. Bill Richardson has made similar states to Rendell's and, as Brian noted in his blog tonight, there are reports of a group of 50 superdelegates ready to jump aboard the Obama train.

Then again, if Clinton wins both big states tonight -- or if she wins Ohio and claims to have had Texas stolen from her (as some of her surrogates have done in a conference call to reporters) -- the goalposts may move to Pennsylvania.