Among supporters of both candidates, it's generally agreed upon that Senator Clinton has little chance of catching Senator Obama in terms of pledged delegates, but maintains a pathway to the nomination that relies upon winning enough of the remaining states to convince the requisite amount of superdelegates to overturn the pledged delegate result. While it will be difficult to get the superdelegates to overturn the results of the actual voting, it may be considerably less so if there is a case that they won't be going against "the will of the people" at the same time. If she can't come up with a lead in pledged delegates, it seems that Clinton will need a lead in the popular vote to have a legitimate claim to contest the nomination at the convention in Denver.
I have therefore decided to spend too much of my free afternoon in an attempt to project the popular vote totals through the next round of states to determine the likelihood that Clinton can accomplish this feat. This is a long diary, so in case you don't want to read through all of the details, here are my basic conclusions:
- By including estimates from the currently unreported results from Iowa, Nevada, Washington, and Maine in the national popular vote total, Obama's popular vote lead increases from 592,682 votes to 705,691 votes, an all-important increase of 113,009 votes.
- If upcoming states vote in similar numbers to states with similarly sized pledged delegations to the national convention, there will be about 6,156,919 votes cast in the Democratic Primary between the Mississpi contest on Tuesday and the concluding South Dakota and Montana contests on June 3rd. Senator Clinton will need to win 55.73% of these votes in order to catch Obama for the popular vote lead. In other words, Clinton will have to perform slightly better the rest of the way through than she did in Ohio last Tuesday.
- If current polling in Mississippi, Pennsylvania, and North Carolina accurately predicts the actual voting results, Clinton will need 61.4% of the votes in Guam, Indiana, West Virginia, Oregon, Kentucky, Puerto Rico, Montana, and South Dakota to capture the popular vote lead. This will be very difficult because Arkansas is so far the only state where she has broken 60% of the popular vote.
- Including Florida, whether by seating their current delegation or by using the expected results of a revote, cuts the level of support Clinton must obtain across the remaining states to 53.3%. If MS, PA, and NC go down as indicated by polling, she will need 55.85% of the popular vote in the other remaining states to catch Obama. However, at least three of these other states should favor Obama, making it extremely difficult to reach this necessary level.
- Given the above findings, Clinton must cut into or eliminate Obama's leads in Mississippi and North Carolina, win huge in Pennsylvania, and fundamentally alter the present voting dynamics in order to defeat for the popular vote title. She will also likely have to rely on seating the Florida delegation and maintaining her present lead if there is a revote. If there is a revote in Michigan, it looks like it will be a tie and therefore not a factor in the popular vote totals.
The current Popular Vote count and associated problems
Currently, according to the popular vote totals at Real Clear Politics, Obama leads Clinton in total votes cast 13,007,968 to 12,415,286 (a margin of 592,682). This total does not include the results from Florida or Michigan. However, it also does not include results from Iowa, Nevada, Washington, and Maine, which have not released popular vote totals. Since it is important to include all states if we want an accurate representation of the overall popular vote, I have taken the total Democratic turnout as reported by state media and used the percentage of state delegates won by each candidate to determine popular votes received in states not reporting actual numbers. While this is certainly an imperfect method, it's probably the best way to determine an overall popular vote margin given the information we have. Since Michigan and Floria remain in dispute, and since they may yet hold new contests, I will hold off including their current totals in the count or discussing their implications until the end of this post.
Here are the additional numbers:
Iowa 1/3 (Caucus turnout: 227,000)
Estimated Obama vote: 86,260
Estimated Clinton vote: 65,830
(it should be noted that these results are after realignment, which likely benefited Obama over Clinton, though supporters of other candidates were also realigned in subsequent elections by virtue of their first-choice candidates dropping out)
Nevada 1/19 (Caucus turnout: 107,000)
Estimated Clinton votes: 54,570
Estimated Obama votes: 48,150
Washington 2/9 (Caucus turnout: 244,458)
Estimated Obama vote: 166,231
Estimated Clinton vote: 75,782
Maine 2/10 (Caucus turnout: 45,000)
Estimated Obama vote: 26,550
Estimated Clinton vote: 18,000
<hr>
Current Total Popular Vote (including estimates from non-reporting caucus states):
Obama: 13,335,159
Clinton: 12,629,468
(margin: 705,691 votes)
When popular vote estimates from states that have not released official totals are included, Obama's popular vote lead increases by 113,009 votes.
<hr>
Moving Forward
While Senator Clinton will clearly need to score some big victories to cut into Senator Obama's lead in the popular vote it will probably be easier to erase a popular vote deficit of 700,000 than a pledged delegate deficit of 155. Here I will attempt to project the popular vote totals of each candidate in upcoming states by comparing turnout in comparable states. Since the DNC appropriates pledged delegates between the states based on both population and size of the Democratic electorate, I will use turnout in states that have already voted (figures from the CNN Election Center) with similarly sized pledged delegations to the states that have yet to vote in order to project future turnout. This is certainly imperfect in that it does not account for open vs. closed primaries (I will not compare caucus states to primary states), the dynamics of the race at a given time, or either candidate's field operations in a certain state. Where available, I will use polling to predict an outcome and assign the popular votes accordingly. Here we go:
Mississippi 3/11 (33 delegates. States with similarly sized delegations: Arkansas (35), where turnout was 307,318)
Polling average: Obama 52.7, Clinton 37.7 (I will project this margin of 15 points to cover nearly 100% of the votes cast rather than just 91%, making it Obama 56.7, Clinton 41.7)
Estimated Obama vote: 174,249
Estimated Clinton vote: 128,152
Pennsylvania 4/22 (158 pledged delegates. Similar states: Ohio (141 pledged delegates, turnout of 2,186,831) and Illinois (153 pledged delegates, turnout of 2,003,800). Average similar state turnout = 2,095,316)
Polling Average: Clinton 47.5, Obama 35.8 (Adjusted for same margin on total ballot: Clinton 55.5, Obama 43.8)
Estimated Clinton vote: 1,162,900
Estimated Obama vote: 917,748
Guam 5/3 (4 pledged delegates. Similar contest: Virgin Islands (3 pledged delegates, turnout of 1921)
North Carolina 5/6 (115 pledged delegates. Similar state: New Jersey 107 pledged delegates, turnout of 1,109,369)
Polling Average: Obama 47.3, Clinton 38.5 (Adjusted: Obama 54.3, Clinton 45.5)
Estimated Obama vote: 602,387
Estimated Clinton vote: 504,763
Indiana 5/6 (72 pledged delegates. Similar State: Missouri (72 pledged delegates, turnout of 820,453)
West Virginia 5/13 (28 pledged delegates. Similar state: New Mexico ("caucus in name only," 26 pledged delegates, turnout of 148,404)
Oregon 5/20 (52 pledged delegates. Similar states: Alabama (52 pledged delegates, turnout of 539,925), Arizona (56 pledged delegates, turnout of 443,426). Average turnout = 491,676)
Kentucky 5/20 (51 pledged delegates. Similar states: Alabama (52 pledged delegates, turnout of 539,925), Arizona (56 pledged delegates, turnout of 443,426). Average turnout = 491,676)
Puerto Rico 6/1 (55 pledged delegates. Similar states: Alabama (52 pledged delegates, turnout of 539,925), Arizona (56 pledged delegates, turnout of 443,426). Average turnout = 491,676)
Montana 6/3 (16 pledged delegates. Similar States: Vermont (15 pledged delegates, turnout of 151,380), Delaware (15 pledged delegates, turnout of 95,979), Average turnout = 123,680)
South Dakota 6/3 (15 pledged delegates. Similar States: Vermont (15 pledged delegates, turnout of 151,380), Delaware (15 pledged delegates, turnout of 95,979), Average turnout = 123,680)
<hr>
According to these admittedly very rough estimates, there will be 6,156,919 votes cast through the remainder of the Democratic Primary campaign. Again, these numbers were obtained simply by extrapolating from similar states, and do not take into account exactly how the DNC pledged delegate allotment has corresponded to turnout or other factors such as the day of the primary, the population density of a state, open vs. closed primaries, the present dynamics of the race, field operations, the fact that there is now no active challenger on the Republican side, etc. However, it does give us a rough estimate of the number of votes left to be counted, and thereby the size of the wins Senator Clinton needs to catch Senator Obama in the popular vote count. Using the 6,156,919 figure and the present popular vote margin estimate of 705,691, Clinton will need to win 3,431,305 of the remaining votes, or 55.73% of those votes remaining in order to tie Obama in the popular vote. While this will be made more difficult if Obama scores victories in Mississippi and North Carolina as indicated by polling, it still equates to simply winning the remaining states with a slightly higher margin than she did in Ohio.
If polling in Mississippi, Pennsylvania, and North Carolina bears out, and the turnout is as expected, the popular vote totals will stand as follows:
Obama: 15,029,543
Clinton: 14,425,283
(margin: 604,260 votes)
The remaining states without polling numbers account for just 2,644,916 of the over 6 million remaining votes. If Obama wins Mississippi and North Carolina by the expected margins, and Clinton wins Pennsylvania by the expected margin, Clinton will need to win 1,624,588 of the remaining votes, or 61.4%, in Guam, Indiana, West Virginia, Oregon, Kentucky, Puerto Rico, Montana, and South Dakota. This will be very difficult, considering Clinton has yet to win that level of support in any state outside of Arkansas, and that Obama has performed well in states similar to Montana, South Dakota, and Oregon, and has yet to lose a state neighboring Illinois, as Indiana and Kentucky each do (although these two states would seem to favor Clinton demographically). Clearly, Clinton does not seem poised to win the popular vote unless the campaign moves drastically in her favor.
<hr>
Michigan and Florida
The wrinkle in any statement on the popular vote is how to count the totals from Florida or, less controversially, Michigan. With Florida included in the current popular vote totals, Clinton trails Obama in popular votes by 13,584,182 to 13,286,272, a margin of just 297,910 votes. Adding the estimates from states without officially released popular vote totals (Iowa, Nevada, Washington, and Maine), the margin increases to 410,919 votes. If the current Florida delegation is seated, or if Florida revotes and provides Clinton with a similar margin (which looks like it would be the case according to Rasmussen), she will need just 53.33% of the popular vote in all remaining states. With Florida included, if the polling in Mississippi, Pennsylvania, and North Carolina is accurate, after those states the race will stand as follows:
Obama: 15,605,757
Clinton: 15,296,269
(margin: 309,488)
In this case, Clinton would need to win 55.85% of the vote in Guam, Indiana, West Virginia, Oregon, Kentucky, Puerto Rico, Montana, and South Dakota to pass Obama for the popular vote lead. Again, this seems highly unlikely if, as expected, Obama wins a couple of those states.
Michigan will likely have much less of an effect on the popular vote count than Florida. This is because no one takes their original vote seriously, and if they revoted, Rasmussen says it would be a tie and therefore a wash in terms of the popular vote.
Conclusion
All things considered, it appears Clinton will need to score a huge victory in Pennsylvania, turn things around in Mississippi and North Carolina, and probably also win do-over primaries in Florida and Michigan in order to lay true claim to the popular vote lead heading into the convention. While most news organizations currently ignore the Iowa, Nevada, Washington, and Maine caucuses when calculating popular vote totals, there is no doubt the Obama campaign will raise the issue if the popular vote enters the debate in a serious way at any point in the future. Clinton needs to seriously alter the dynamics of the race to capture the popular vote lead, but is better positioned here than she is in the pledged delegate race.