Jordan Carr
Political Science 323S: Analysis of Political Campaigns
Iyengar
December 4, 2009
Superdelegates in the 2008 Democratic Presidential Primaries (Table 2, Table 3, Table 4)
The 2008 Democratic presidential primaries were far and away the most exciting and competitive since the Democratic Party revamped their system four decades ago. Two major candidates quickly emerged, as Illinois Senator Barack Obama was able to challenge and eventually defeat New York Senator and former First Lady Hillary Clinton, who was regarded as the establishment candidate preferred by party elites. The problem I am trying to solve is at its most basic level determining how much that establishment support meant to Sen. Clinton, and how steep a mountain then-Sen. Obama had to climb to win the nomination. I will do this first and foremost by tracking performance against superdelegate totals. This causes a few difficulties.
Superdelegates are incredibly difficult to track. There are 852 superdelegates to keep track of and about a hundred of the superdelegates did not throw their support one way or another before Mrs. Clinton suspended her campaign in early June. The sheer magnitude of their numbers makes them difficult to wrangle. Adding to that, any endorsements are unofficial, so it is not always clear what counts as supporting a candidate. Does somebody saying Barack Obama would be a transformative president count? Does somebody saying Hillary Clinton’s qualifications are unparalleled count? It is not always clear. There are basically no two independent sources that have the same totals for superdelegates. I made my best effort to try to compile an independent list of when each superdelegate first endorsed their candidate, however that task quickly proved to be an exercise in futility. Nevertheless, the good people at DemConWatch (who tracked this information throughout the process) sent me along their numbers which are as comprehensive a listing of public endorsements as there is. I say public endorsements because it appears some media outlets went ahead and directly contacted the superdelegates to find out their opinions, but did not publish who said what, only the numbers.
Iowa
According to ABC News, Mrs. Clinton entered the Iowa caucuses with a 158-89 superdelegate lead on Mr. Obama,1 holding a 64%-36% advantage over him (I am counting delegates pledged to irrelevant candidates equally to those that are unpledged). This total is much higher than the number at DemConWatch, which was tracking public endorsements, and extrapolating backward gave Mrs. Clinton a 131-47 advantage (74%-26%).2 Both these figures are telling in a way, but ultimately the DemConWatch one is more useful because it is more accurate in terms of the number of people willing to directly tie their names to the Clinton and Obama campaigns early in the process. When viewed together, if one assumes that the superdelegates are simultaneously trying to shape who the nominee is, that they prefer Clinton as a group, and that the individual superdelegates stand to gain from backing the winning candidate (or the one who wins their geographic region), then it makes sense that the numbers who support the favorite in a more private manner (as judged by the ABC numbers) are smaller than the numbers of her public supporters. Nobody wants to be left supporting a loser, but in the event that Mr. Obama did win the nomination, there presumably would be disproportionate benefit to his early supporters. For my tables, graphs, and all figures from here on out I will use the DemConWatch superdelegate numbers because public endorsements are 1) easier to track 2) more reliable, and 3) much more difficult to back off of.
What is also important is that by any measure, Mrs. Clinton had a significant advantage on Mr. Obama heading into the Iowa caucuses. She had won the invisible primary quite handily, so if that theory were to hold here, she would likely score a few early victories then roll to the nomination with very little contestation along the way. Beyond that, while she faced her handicaps while trying to become the first female president, Mr. Obama was trying to become the first African-American president, the novelty of which cannot be overstated—there had never been a remotely viable African-American who ran for national office to this point, whereas the Democrats had nominated a woman for Vice President in 1984, and the Republicans would in 2008.3 Given that the first caucuses were to take place in Iowa, a state whose demography would seem to favor Clinton due to its overwhelming white population (96% as of the 2000 census)4, her edge seemed to be as large, if not larger than her superdelegate margin would indicate. Mrs. Clinton also held a 3-2 advantage among the Iowan superdelegates who had declared before the state’s caucuses.
Mrs. Clinton’s advantage in press coverage was slightly smaller, but still significant. According to Google News’s archive from December 2007, the number of articles that mentioned “Hillary Clinton” but not “Barack Obama”5 dwarfed the number that mentioned “Barack Obama” but not “Hillary Clinton” 2,590-1,830 (59%-41%).6
So, by any stretch of the imagination, Mrs. Clinton was the favorite heading into the Iowa caucuses, but she lost the vote there by a margin of 93,952-73,663 (38%-29%), finishing in third behind former North Carolina Senator John Edwards. In what would become a consistent feature of the campaign, Obama would win a caucus state, and in doing so collect a share of the delegates disproportionate to his vote share. In terms of the campaigns’ strategies, it seemed like the Obama campaign was playing to win a hard fought battle while the Clinton campaign was playing to win an early decisive knockout victory, and found itself caught off guard when it realized the primary season would be extended the full length.
New Hampshire Through Super Tuesday
After Hillary Clinton’s disappointing show in Iowa, it was clear that media coverage began to shift its stories toward Barack Obama; he was the new candidate, there were more new stories to tell about him. In between the Iowa caucuses and New Hampshire primary, there were 554 news stories exclusively about Barack Obama7 and only 370 that were about Hillary Clinton and not Barack Obama (60%-40% in Mr. Obama’s favor).8 Seemingly, all the momentum was with Mr. Obama. Pollster’s composite poll gave Barack Obama a 6 point advantage heading into the New Hampshire primary.9 Mrs. Clinton famously broke down in tears at a campaign event in New Hampshire, which only went to show how far her once seemingly invulnerable campaign had fallen.10 And yet, Mrs. Clinton exceeded expectations in New Hampshire, winning the state narrowly and saving her candidacy from a potentially quick and embarrassing end.11
After New Hampshire came Nevada on January 19, which should probably be regarded as a draw, considering that while Mrs. Clinton won the state’s popular vote by 6 points (51%-45%), Mr. Obama won 14 delegates to Mrs. Clinton’s 11, so in that sense he made more progress toward the nomination.12 A week later came South Carolina on January 26, which Obama won convincingly, 55%-27% while claiming 33 delegates to Mrs. Clinton’s 12.13 South Carolina has a large black population (28.5%), and as such was believed likely to go to Mr. Obama.14 That said, it did not help when former President Bill Clinton compared Mr. Obama’s victory in South Carolina to Jesse Jackson’s, a comment that some believed had racial undertones.15 As such, it was clear that Mr. Obama had scored a major victory on the cusp of Super Tuesday, when 22 states plus American Samoa would vote.
All in all, January 2008 was a tremendous success for Mr. Obama and a mitigated disaster for Mrs. Clinton. Unlike in December 2007 when Mrs. Clinton dominated the media coverage, in January 2008, Mr. Obama had become the major story. Between the day after the Iowa caucuses and Super Tuesday on February 5, there were 4,120 stories that mentioned Barack Obama but not Hillary Clinton,16 and 3,180 that focused exclusively on Mrs. Clinton, a 56%-44% advantage for Mr. Obama.17 Mrs. Clinton entered the primaries with an advantage in superdelegates of 131-47 for a margin of 84 (74%-26).18 By February 3, Mrs. Clintons’ public endorsements totaled 198 and Mr. Obama’s numbered 105, meaning her advantage was 95 and 65%-35%, which is roughly the same percentage as the ABC News figure and a wider advantage than either that or the initial DemConWatch one.
There a couple possibilities for this puzzling resiliency of Clinton support in an otherwise disappointing month. It could be the result of superdelegates who had been supportive of Clinton rushing to endorse her in a moment of need, or (more likely in my opinion) the result of the data finally starting to catch up with the endorsements that had already been made, especially keeping in mind that few in the media or elsewhere paid much attention to the superdelegates before it became a real possibility that they might determine the nominee.
This brings us to Super Tuesday, February 5, 2008. The momentum seems to be with Mr. Obama; he is receiving more media coverage, he has emerged, seemingly from nowhere, to be a legitimate contender to take the Democratic nomination for the presidency from a New York Senator and the First Lady to the only Democratic president in the past 28 years. Still, by most accounts, the race was still Mrs. Clinton’s to lose, as polls indicated she would be favored to win Alabama, Arizona, California, Colorado, Connecticut, Delaware, Massachusetts, Minnesota, Missouri, New Jersey, New Mexico, New York, Oklahoma, Tennessee and Utah while Mr. Obama was the favorite in only Georgia and his home state of Illinois alone.19 By any measure, Mr. Obama’s Super Tuesday performance exceeded expectations. He won both states he was favored in, as well as taking Alabama, Colorado, Connecticut, Delaware, Minnesota, Missouri, and Utah from Mrs. Clinton’s column. In addition, he outperformed the polls by a wider margin than Mrs. Clinton in Georgia, New Mexico, New York, Oklahoma, and Tennessee.20 The only states where Mrs. Clinton could be said to have exceeded expectations were California, Illinois, and Massachusetts, and none of those overachieving performances were remotely like Mr. Obama’s in Alabama, Colorado, Delaware, Georgia, Minnesota and Utah where the actual results were all more than ten points better for Mr. Obama than the poll results.21
Mr. Obama’s performance on Super Tuesday was certainly better than expected, but who actually won? Mr. Obama did, but narrowly by a margin of 847 delegates to Mrs. Clinton’s 834.22 From this point, it was clear that neither candidate would be able to score a decisive blow early, and so while “momentum” would not necessarily be a moot concept from here on out, its value would be extremely limited. Neither side would be able to make a legitimate claim that the other side has no chance until several months later. By February 10, Mrs. Clinton had 224 superdelegates pledging support to Mr. Obama’s 127, a difference of 97, the largest it would be throughout the entire campaign. 23 Mrs. Clinton held 64% of superdelegates to Mr. Obama’s 36%.
February and March
The long-term trends after Super Tuesday through March universally favored Obama. In caucuses and primaries in Louisiana, Nebraska, Washington, the U.S. Virgin Islands, Maine, District of Columbia, Maryland, Virginia, Hawaii, Wisconsin, Ohio, Rhode Island, Texas, Vermont, Wyoming and Mississippi, Mr. Obama’s votes earned him 493 delegates to Mrs. Clinton’s 369.24 This period was noteworthy for spawning the two most salient feeding frenzies of the campaign, one that hurt each side, both of which broke, interestingly enough during the early part of the gap between the March 11 Mississippi primary and the April 22 Pennsylvania primary. It is perhaps coincidental that these two major stories would break during what would otherwise have been the longest lull in the campaign, but I am somewhat inclined to believe otherwise—not necessarily that media outlets were sitting on these stories, but that they no longer were devoting maximum resources to the campaign, or that the campaigns decided it would be a good time to dump their own bad stories at a time where they would have minimal impact. The second theory makes more sense for the Mr. Obama’s controversy as it had already begun leaking over the past few weeks and it was the Friday after the Mississippi primary when the story broke, but this theory seems at least a touch conspiratorial.
The first media firestorm to break was Mr. Obama’s longtime pastor Jeremiah Wright being caught on tape saying inflammatory things about the United States.25 Mr. Obama unsurprisingly distanced himself from the comments and eventually severed ties with Reverend Wright.26 The fallout led to Mr. Obama giving a “speech on race,” a 45-minute long monologue discussing race relations in America and Mr. Obama’s own racial background—an altogether unprecedented action for a presidential candidate.27 Ensuing contests seem to evince that Mr. Obama was not badly damaged by the incident.
For Mrs. Clinton, the story was her claim that she had come down under “sniper fire” in Bosnia. Witnesses (including the comedian Sinbad who was on the trip)28 refuted her account and eventually a video showing a young girl greeting Mrs. Clinton upon landing forced her to claim she “misspoke.”29 Where Mr. Obama’s errors were one of association, and perhaps understandable within the context of a church—it is unlikely that many people would agree with the broader assertion that their pastor’s worldview invariably reflects their own—Mrs. Clinton’s was an out and out fabrication transparently designed to make herself look better, especially in comparison to Mr. Obama (keeping in mind the entire purpose of the fabrication was to emphasize her foreign policy experience).
Nevertheless, the Jeremiah Wright story proved to have longer legs. Over the remainder of the campaign there would be 12,400 stories that featured both the names “Barack Obama and Jeremiah Wright,”30 but only 6,560 that featured “Hillary Clinton and Bosnia or sniper or fire.31 This is perhaps a reflection on Mr. Obama’s decision to address the issue head-on, but also on the fact that it just had a lot more angles. Mrs. Clinton lying is a scandal, but not one that there is a lot to say about, particularly after she tacitly admitted it. Mr. Obama’s radical pastor spawned disagreement and debate from all sides of the political spectrum (but mostly from his opponents).
“Bittergate” and Remaining Contests
On April 6, at a private fundraising event in Kentfield, CA, Barack Obama described certain small town Americans as “bitter” people who “cling to their guns and religion.”32 Media outlets would not pick up on the story for a week or so though, but the ensuing “bittergate” would be a theme that the Clinton campaign and Mr. Obama’s detractors would harp on, warranting about 1,300 news articles in the two weeks.33 This signaled a change for the Clinton campaign which to this point had remained largely above the fray (with exceptions, naturally). The obvious reason for the change of course presumably being that she had moved from presumptive nominee to a desperate underdog at this point. Though a ruling on whether or not Florida and Michigan’s delegates would be seated (the two states had moved their primaries up without the permission of the national Democratic Party, resulting initially in the complete stripping of their delegates) would not come down until May 31,34 it was clear at this point that Mrs. Clinton needed something tremendous to happen in order for her to win. Superdelegates had begun leaking to Mr. Obama. Mrs. Clinton’s onetime lead of 97 superdelegate commitments had shrunk down to just a 255-230 lead (53%-47%), a 25 superdelegate margin in just over two months.35 Considering the long odds she faced, it is unsurprising that she would “go negative,” so to speak, more forcefully than she had before.
Interestingly, in the remaining contests, it appears that the desperation of Mrs. Clinton’s situation served her well. From here on out, Mrs. Clinton would win the pledged delegates in the remaining contests by a margin of 304-262 (54%-46%).36 That she would perform her best when her campaign was at its most desperate is most curious. In large part, the demography of the remaining states favored Mrs. Clinton, as they were largely whiter and more rural, factors that had favored her throughout the primaries. Still, Mrs. Clinton’s victories in these contests were not nearly wide enough and she effectively ended the race on June 7 by suspending her campaign.37 Barack Obama would be the Democratic nominee for President.
Superdelegates and the Electoral Process
Much fuss has been made about the role that superdelegates played in the nominating process, with some believing that they would swing the nomination one way or another (they decidedly did not). Some may claim that the superdelegates’ swing toward Barack Obama late in the process as his nomination became more likely reflects their exertion of influence over the electorate, and not vice versa. This is a mistake. Rather than looking at the superdelegates nationally, I decided to break them down by state. Importantly, I measured how many superdelegate commitments had received before the primary contest in that superdelegate’s state (meaning a California’s superdelegate’s endorsement after that state’s primary would not count). Again, with superdelegate counts, there is always a chance of error, but I am confident that my numbers on this front are on the whole representative.
The results were somewhat surprising. All in all, Hillary Clinton received 234 superdelegate endorsements in states before those states’ primaries (from here on I will refer to those as “local superdelegates” for convenience) to Barack Obama’s 174 (57%-43%).38 Despite this overwhelming superdelegate support, Mrs. Clinton did not win the election, falling narrowly to Mr. Obama in the end.
In 26 contests where Mrs. Clinton entered holding an advantage in local superdelegates, the results are strikingly mediocre for her. In those states where Mrs. Clinton held an advantage in local superdelegates, she had 195 endorsements to Mr. Obama’s 65 (75%-25%). Yet in the voting, the delegates that were up for grabs were split almost perfectly: 1023 to Mr. Obama, 1019 to Mrs. Clinton.39
On the flipside, in the 17 contests where Mr. Obama held the advantage in local superdelegates, he was very successful. Mr. Obama held a 89-23 (79%-21%) advantage in local superdelegates in those states and that translated into a 491-349 (58%-42%) advantage in pledged delegates for Mr. Obama.40
One theory to explain the data may be that Mrs. Clinton was the establishment candidate and therefore it was Obama who required the approval of trusted elites (using local superdelegates as a stand-in) to gain widespread voter approval, and that superdelegate disapproval did not alter opinion of Mrs. Clinton who was already a known commodity. The numbers do not really bear this out though. If local superdelegate support was a viable predictor for Mr. Obama’s success, his pledged delegate share should have kept pace with his local superdelegate share in the states where he had won most of the superdelegates. Instead, he fell well short of that mark, winning 58% of those delegates while winning 79% of local superdelegates.41
Instead, the phenomenon most likely at work was simply that the electorate cared little of what the superdelegates thought, and once they made their preferences clear, the superdelegates were compelled (or chose) to fall in line behind the locally (or nationally) preferred candidate. Considering that Mr. Obama’s local superdelegates supported him in a larger proportion than the electorate at large (in the states where more of them supported him), it does not really make sense that the electorate was following them as they did not go as far as the superdelegates, whereas in the states Mr. Obama trailed in local superdelegates, he outperformed his superdelegate percentage by a wide margin.42 That the results are so disparate suggests a lack of influence by the superdelegates on the proceedings.
For example, Mr. Obama did better in caucus states than he did in the ones where he had received a majority of local superdelegate endorsements, receiving 65% of all caucus delegates (to be fair, he did also have 60% of local superdelegate endorsements in those states).43 What is especially interesting is that if you break the data down a little further, it reveals that Mr. Obama actually performed slightly better in the caucus states where Mrs. Clinton had received either an equal number or more local superdelegate endorsements than in the states where he had received a majority of local superdelegate endorsements.44 The difference there is marginal, but it is clear that he was dominant in the caucuses, and that for whatever reason the voting format provides a better indication of the results than superdelegate endorsements. Nowhere is this more clear than in Texas, which has a combination of primary and caucuses, with Mrs. Clinton winning the primary portion and Mr. Obama winning the caucuses.45 One could even go so far as to make the argument that because he was not the establishment candidate his supporters were naturally more fevered and excited in a way that made him a more viable candidate in caucuses where intensity of preference is of greater value due to the time costs they put on participants, but that is for another paper to resolve.
Conclusion
In terms of momentum, it seemed that if anything after the early stages of the 2008 Democratic primaries that the process was working in reverse. In reality, what was happening was simply that states more demographically disposed to one candidate or another would often appear in alternating fashion—as is bound to be the case when candidates have roughly equal support across the nation. That said, Mr. Obama was able to ride an early wave of momentum after January, which allowed the voting public to consider him as a seriously viable candidate in a way they may not have prior to the Iowa caucuses and other early contests.
The 2008 Democratic primaries demonstrated the limited value of superdelegates in determining the nominee. Candidates who would by no means be considered the party elite’s choice have repeatedly won the Democratic Primary. In addition to Barack Obama, George McGovern and Jimmy Carter come to mind. The superdelegate endorsements seemed to follow the electorate, not lead it, which certainly makes sense as there is often very little incentive to make an early declaration. While it can be helpful to signal one’s preference for the eventual candidate early, among over 800 superdelegates, it does not necessarily entail any future rewards. In short, while the invisible primary may exist in terms of vetting candidates, it did not pick the winner of the 2008 Democratic primaries.
|
