The myth of Democratic electoral troubles was put in serious jeopardy by 2006, and 2008 should put it decisively to bed. It was always a troubled theory since it largely rested on assuming "elections" meant "Presidential elections" but seeing as two co-equal branches of the US government face regular elections, and the first branch has two co-equal chambers, it would be worth seeing how those elections have been doing.
In my diary last weekend, "Swing State Clusters Tell Story of Potential 'Map-Changing' Obama Landslide", I pointed out how the punditalkcrazy has been utterly oblivious to the actual configuration of battleground states as revealed by state-level polling this year. Regardless of what the national polls say, there just doesn't seem to be much chance that, even at his best, McCain could win more than one or two Kerry states, while Obama could easily pick off half a dozen, even a dozen Bush states. As I argued in my previous diary today, "Electoral Map Typology", it is quite likely the map will change in this election for many elections to come.
Despite Obama's amazingly consistent lead throughout the general election, the talking heads on cable television returned to their incessant bloviating over whether Obama should be leading by more than just five points over McCain. It's really painful to watch these fools who don't bother to pay attention to history to understand how a five-point popular vote victory translates when it comes to the only metric that matters -- the Electoral College. (Hint: it translates to a landslide)
And his chart of popular vote margins to EV margins is pretty straightforward:
Entering tonight's voting, according to my tabulations, Barack Obama held a 357 vote lead in the national popular vote, with a 24,000 vote margin of error.
Add it altogether, and Obama is headed for a victory in the national popular vote. While it remains to be seen if he will move beyond the "margin of error" in my tabulations, I also freely admit that there are numerous other possible tabulations other than my own. For example, Poblano has a popular vote counter with at least 972 possible totals. So, there probably won't ever be a final, consensus total on this matter. Overall, most of the totals favor Obama.
No matter what you think of the popular vote totals, and no matter which popular vote total you subscribe to, I think we can all agree that the nomination process requires significant reform. If you had absolute power over the nomination process, what changes would you make?
Early this morning, I wrote that Clinton currently leads in the popular vote by 19,899 votes. This figure is based upon the bottom line in the current Real Clear Politics count, minus 64,504 uncommited votes in Michigan that, according to exit polls, came from people who indicated they would have supported either John Edwards or Bill Richardson, had those candidates been on the ballot.
However, I now realize that those totals were incorrect. This is because, in Michigan, 27,694 votes were not counted because they wrote in a candidate. When, in accordance with exit polls, 72.9167% of those votes are allocated to Obama, that puts another 20,194 votes in his column. According to the broadest possible definition of one-person, one-vote, this gives Obama an almost comically narrow lead of 357 votes heading into tomorrow's primaries.
Of course, since we are dealing with estimates on the Iowa, Maine, Nevada and Washington vote totals, and since we are also dealing with estimates on the Michigan uncommitted and discarded vote totals, there is a margin of error in these estimates. Specifically, there is about a 3% margin of error in either direction among the estimated votes, since we are dealing with exit polls and the vagaries of the delegate selection process in the four caucus states. A 3% margin of error on the estimated 750,000 votes from these states gives a margin of error range of 22,500 votes in either direction. So, in order for there to be no doubt as to who won the popular vote, it will be necessary for Obama to win tomorrow's primaries by 22,143 votes.
For the sake of rounding, let's say 25,000 is the ultimate, "no doubt" popular victory threshold. This means that Obama needs to win tomorrow's primaries by 24,643 votes in order to definitively declare that he is the popular vote winner. Tomorrow night, while live-blogging returns, I'll provide updates on whether or not Obama will reach that threshold. No matter what happens, there is no definitive way to prove that Clinton won the popular vote. Also, as I indicated last night, the difficulty in determining the popular vote winner speaks to a lack of democracy in the process that needs to be reformed in advance of 2012 and other upcoming nomination campaigns.
What sort of reforms do we need? I say we go with the California Plan, abolish caucuses, and increase the number of delegates to about 6,000. Altogether, these reforms would be the most democratic system possible that still maintains a staggered primary calendar and a delegate-based convention. I'll have more on reforming the process soon.
There are innumerable caveats to any popular vote total in the nomination campaign. Some states held primaries, while other held caucuses. Some primaries were open to all registered voters, others to only Democrats and Independents, and still others to only Democrats. The staggered primary calendar is another major issue, which resulted in many states having different candidates on the ballot, and voters with varying knowledge of results. Some states did not even keep popular vote totals. Michigan and Florida are also obviously major caveats. No campaigning took place in those states before the voting began, many voters stayed home because they were told the elections wouldn't count, and Obama's name wasn't even on the ballot in Michigan. Further, a nomination campaign is not about the popular vote, and there wasn't a single campaign that used the popular vote as a metric before the voting caucusing began. So, the popular vote is a contentious metric in the nomination campaign, to say the least.
However, whatever the difficulties of applying the value to the specific "election" that is the 2008 Democratic Party presidential nomination campaign, there is also an obvious value to the principle that the individual with the most support of the electorate should win any given election. Governing power should always derive from the popular will, and we should always work to make our system of government more democratic. The lack of a clear, consistent definition of the popular vote in the Democratic presidential nomination campaign speaks of the serious flaws in the process itself. For all of the reasons listed in the first paragraph, not only is there no universally accepted definition of the popular vote, but as a party we are also a long way from instituting a democratic form of intra-party governance. Major changes need to be made in advance of the 2012 nomination contest, and all future nomination contests, so that our election process better adheres to democratic principles.
As I have argued in the past, within the context of the 2008 Democratic nomination contest, any attempt to determine who won the "popular vote" should adhere to democratic principles itself, as best as can be done. This is because the "popular vote" is not a legal argument, and not specific to any campaign, but instead a moral one based on abstract principles of democracy. As such, popular vote totals should do the following:
Include the will of all those who participated in delegate selection contests for the Democratic National Convention.
Allocate only one vote to each participant in those contests.
In cases where participants did not have their preferences recorded, do everything possible to estimate those preferences.
Failure to do this is to engage in the "popular vote" argument in bad faith, since it turns a moral argument about democratic principles into a selective, partisan argument about power. And yes, one side is more guilty of the other on this front. However, that does not lessen the principles involved--it lessens those who twist those values.
According to the above principles, with South Dakota and Montana left to vote, Hillary Clinton currently holds an extremely narrow 19,899-vote lead over Barack Obama in the popular vote. Here are the current totals:
Clinton: 17,916,763
Obama: 17,896,864
These totals include Iowa, Maine, Nevada and Washington, even though no official popular vote numbers were kept. They also include Florida, even though there was only minimal campaigning in the state before the primary took place and even though many people thought it wouldn't count. These totals also include Michigan, even though Obama's name was not on the ballot. They do, however, also allocate 72.91% of the "uncommitted" vote to Obama, which is the amount of the uncommitted vote exit polls indicate he would have received in the state had his name been on the ballot. In short, these numbers are the final line from the Real Clear Politics popular vote count, minus 64,504 votes in Michigan that came from people who indicated they would have supported either John Edwards or Bill Richardson, had they been on the ballot.
Now, with about 275,000 votes left to go in South Dakota and Montana, and with Obama holding double-digit leads in both states, it would be pretty surprising if Obama did not end up as the winner of the popular vote. Of course, since these are estimates, there is also a small margin of error in this count that might throw the outcome of the "popular vote" into question. Undoubtedly, supporters of both sides will also continue to push different totals, for all of the reasons listed above. However, this count is really the only popular vote total worth making, because it is the one that most closely adheres to the democratic principle of one person, one vote. It is, course, still imperfect.
Unless something surprising happens on Tuesday, Barack Obama will narrowly win the popular vote. Despite all of the imperfections in the system, that should still matter to anyone who holds democratic principles and intra-party democracy as valuable. Just as importantly, it should also make plain the need reform the process in determining our nominee, so that a disaster like this never happens again. While it is unlikely anyone reading his will ever live to see another nomination campaign this close, it isn't only close elections where adhering to democratic principles matter. We need to do everything we can to make sure that the system is as fair as possible to all of the people participating in the process, and about upholding our own values in the process, not just about electing the cult of personality of the month. That is a perspective that I think has been largely forgotten in this nomination campaign on both sides, and needs to be regained as quickly as possible.
Bill Clinton is once again pushing the "popular vote" angle on his wife's race against Barack Obama. As most know, the popular vote claim for Hillary Clinton is predicated on counting Michigan and Florida where the candidates didn't campaign or run ads (and in MI Obama's name wasn't on the ballots). Without those two states Obama is ahead in total votes and will finish June 3 ahead in total votes.
But I'd like to offer something new about this popular vote argument because it is clear we will be hearing more about it in the days to come. It has to do with a point made by the Obama campaign about the validity of the popular vote metric. Their point is that, given that the whole contest is predicated on winning delegates, their strategy was focused on winning delegates and not on maximizing total votes.
In the proportionally-allotted delegate system used in the Democratic praimaries it is most efficient in terms of advertising dollars and candidate time (UPDATE: and organizing time/money) to "buy" the cheapest delegates. Each delegate vote is worth "1" but the cost of acquiring each delegate vote varies. A simple way to see this, for example, is to divide each state's population by the total delegates it will have. You'll find that each delegate in Vermont represents about 40-thousand people and each delegate in Texas represents more than 100-thousand. If you're a smart campaign and you're looking for one extra delegate, better to spend your time and money in places where, say, reaching a million people translates into 25 potential delegates instead of just 10 potential delegates. Is this what the Obama campaign did? The table below sure makes that case. The 15 "cheapest" states/contests are listed along with the winner and the delegate margin the candidate won. Obama won 12 of these contests and his total pledged delegate margin among these 15 is 42 (so far, WY and SD may pad that).
LOCATION
POP.
DELE- GATES
POP./DEL.
VOTE WINNER (DEL.MARGIN)
Dem.Abroad
100,000 (est.)
7
14,286
Obama +2
Amer.Samoa*
57,291
3
19,097
Clinton +1
D.C.
572,059
15
38,137
Obama +11
Guam*
154,623
4
38,656
Obama +0
Vermont
608,827
15
40,588
Obama +3
Wyoming*
493,782
12
41,148
Obama +2
Alaska*
626,932
13
48,226
Obama +7
N. Dakota*
642,200
13
49,400
Obama +3
Rhode Island
1,048,319
21
49,920
Clinton +5
S. Dakota
754,844
15
50,323
Obama**
Delaware
783,600
15
52,240
Obama +3
Maine*
1,274,923
24
53,122
Obama +6
New Hampsh.
1,235,786
22
56,172
Clinton -3
Montana
902,195
16
56,387
Obama**
Hawaii*
1,211,537
20
60,577
Obama +8
*Caucuses
**Candidate significantly ahead in most recent poll
If you continue down beyond the 15 listed here you'll start finding more Clinton victories. At the very bottom of the list, where reaching the large numbers of people required to achieve an extra is expensive, you'll find that Clinton wins 4 out of 6 states. These victories come at great expense and after doing this math I am not surprised the Clinton campaign is in debt.
Now, the flip side to this argument is that, therefore, each pledged delegate that Clinton won should be "worth more." And that is Bill Clinton's argument, although he bases it on the fact that Obama did well in caucus states (which I've designated above with an asterisk) where each "vote" is, in his opinion, disproportionally powerful. But the table above shows that it's not just caucus states where "cheaper" delegates could be pursued.
The Obama campaign had the best strategy for the system we have in place. In the general election do we want the candidate who will maximize the popular vote or the one who will work to maximize the electoral vote? Remember, Al Gore would have won with any other low population state. He wouldn't have needed Florida. And remember, he won the popular vote, for what that's worth.
Clinton's enormous victory in West Virginia last night will invariably bring arguments over the "popular vote" in the Democratic nomination campaign back into the discussion. Even before West Virginia, which Clinton won by around 137,500 votes, the Clinton campaign was arguing that it was "in striking distance" of the popular vote lead. In fact, according to counts that do not (fixed) include participants in the delegate selection events in Iowa, Maine, Nevada and Washington, but do include the non-delegate selection events in Florida and Michigan, and also do not include the preferences of the "uncommitted" participants in Michigan, the Clinton campaign is indeed ahead. And so, right on cue, the Clinton campaign is now claiming they lead in the popular vote.
The problem with this popular vote total is that it is a moral argument about the will of the participants in the Democratic nomination campaign, not a legal argument over the definition of the winner of the nomination campaign. Legally, the Democratic nominee is determined by delegates, not by popular vote totals. For a moral argument about the popular vote--a.k.a. the will of the nomination campaign electorate--to carry weight, it needs to be as inclusive as possible in its vote totals. Instead, this vote total pretends that the over 550,000 caucus goers in Washington, Nevada, Maine and Iowa, not to mention the quarter of a million uncommitted voters in Michigan, didn't actually have candidate preferences in the nomination campaign just because those candidate preferences weren't recorded. Excluding those 800,000 participants in the nomination campaign from a popular vote toal, especially when exit polls and turnout numbers make close estimates on those preferences quite simple, renders that popular total pointless. Since popular vote totals are statements of moral value, mass exclusions of this sort drains any popular total of all its moral force.
The popular vote total in the nomination campaign is a moral argument about fairness, intra-party democracy, and legitimacy. It isn't even a moral argument that many people accept, given, among other issues, the variances in voter eligibility from state to state, the various "legitimacy" of the nomination events used in the totals, and the staggered primary calendar itself. Still, as an argument over moral legitimacy to the Democratic presidential nomination, it should not be used lightly and arbitrarily. Certainly, it should not exclude hundreds of thousands of people who tried to participate in the nomination campaign by means of the only event offered to them. When all of the people who attempted to participate in their state's only delegate selection event (or their state's only potential delegate selection event, as is the case in Florida and Michigan) are included, Barack Obama still leads by just under 260,000 votes even after Clinton's 137,500 vote victory in West Virginia. Barring some pretty shocking results in the remaining five contests, Obama will still hold that lead after all the voting is complete on June 3rd.
Of all the remaining arguments the Clinton campaign will probably make over the next three or four weeks, the popular vote argument will disgust me the most. They will claim the moral high ground by supposedly having the most support from participants in the nomination campaign, when in reality such a claim can only be made by ignoring the 800,000 participants in the nomination campaign whose candidate preferences were not recorded (but whose preferences can be closely estimated). Making a moral argument that involves shutting voters out of the process is very disturbing, and gives me bad memories of the Florida recount. Remember, if there was a review of all statewide ballots in Florida, Gore would have won according to any vote counting standard. Excluding voters from popular vote counts, and then claiming legitimacy based on fraudulent counts, does not have a positive track record in America.
Here's my prediction. Clinton will unfairly claim a popular vote win.
This is based on three assumptions, (by the way, I included the new 26,000 votes recently awarded Obama from the provisional ballots in Ohio)
#1: FL & Michigan are counted
#2: Obama receives all of the "uncommitted votes" out of Michigan, while Hillary gets her votes out of Michigan
#3: The numbers include "estimates" from caucus states.
Now here's how I came up with my numbers. Please feel free to double-check my math if you think it is incorrect and I'll change it. Lastly, I do not agree that Michigan should be counted at all, nor Puerto Rico. However, the Clinton campaign will, so we might as well figure out how they're gonna look at this. And yes, I know the popular vote doesn't matter. But it matters to the Clinton campaign (and some super delegates), so we're gonna be hearing abou it. The margins out of Kentucky and West Virginia are generous to Clinton, but not inconceivable.
While the nomination campaign might very well be effectively over, especially given the vast, pro-Obama superdelegate movement over the past week, it remains possible that at some point between now and June 3rd, the Clinton campaign will argue that it leads in the popular vote. It needs to be pointed out now that in the broadest possible count of the popular vote, the count that includes all people who participated in the nomination process, that such a claim will simply be false. As I argued back on April 23rd:
In keeping with the principle of one person, one vote, the only good metrics to use are the ones with the broadest popular participation. As such, when measuring the popular vote, it is best to throw the widest possible net. This means to include Florida. It also means to include the estimates from caucus states that did not release popular totals, which stand at Obama 334,084--223,862 Clinton. Finally, it means to include Michigan, but also to allocate Obama 72.91%, or 173,368 of the uncommitted vote. This number is derived by dividing Obama's exit poll support in Michigan by the combined exit poll support of Obama, Edwards and Richardson, and then multiplying that number with the total uncommitted vote.
A second look at the official Michigan results indicates that the Obama uncommitted vote estimation should actually be 173,664. My earlier estimation was slightly lower because it did not use the official Michigan results.
Second, the Obama campaign's allies in Michigan organized an effort to get people in Michigan to vote for "uncommitted" in the Democratic primary, helping to bring the uncommitted share of vote to 40 percent. So the Obama camp can't reasonably argue supporters participated in the GOP primary and didn't vote in the Democratic contest.
Either some of those uncommitted votes were for Obama, or none of them were. In this passage, the Clinton campaign states that at least some of the uncommitted votes were for for Obama. To then argue that Obama gets zero votes in a popular vote count that includes Michigan just isn't "reasonable," in the Clinton campasign's own phrasing.
At the same time, it would be inaccurate to re-allocate the entire Michigan vote based on exit polls, since it retroactively takes votes that were in fact cast for Clinton and shifts them to other candidates. The only fair and democratic way is to give Clinton all of her Michigan votes, and then to allocate the uncommitted votes based on exit polls of Obama, Edwards and Richardson support, as those were the three candidates who removed their names from the ballot. So, 173,664 is the total for Obama from Michigan.
Overall, this means the best one-person, one-vote popular vote total is the bottom most total from Real Clear Politics, plus 173,664 for Obama. This leads to the following current, grand totals:
Obama: 17,087,483
Clinton: 16,690,099
This gives Obama a current margin of 397,384, which Clinton is very unlikely to erase during the final six contests. There is talk of up to two million people voting in the Puerto Rico primary, considering that there was 80% turnout in their last gubernatorial election. However, that was a general election, not a primary. Also, the local gubernatorial election is bound to draw way more voters than a Democratic primary, given the lower level of campaigning that has taken place and the more indirect impact it will have on Puerto Rican affairs.
It is now virtually guaranteed that Obama will win a narrow plurality of the popular vote / participation in the Democratic nomination campaign. Neither candidate will have won a majority, but the "will of the electorate" will still have been served by the outcome. That is a very good thing, since I know that I would not have been the only person extremely uncomfortable with nominating a candidate who did not receive the most popular support in the process. The day we nominate a candidate who did not receive the most popular support / participation will be a dark day for the party.
The popular vote is a fallacy, as I've written before. A brief recap: 1) Candidates would camp out in large urban areas like LA and Brooklyn and never spend so many millions trying to split 200,000 votes in the whole state of NH; 2) No state would rationally hold a caucus, thereby disenfranchising its say in the nomination selection; and 3) unlike the general election concept of one person, one vote, allowing independents or Republicans to vote in some states but not others badly skews the simplistic moral argument underpinning popular vote.
Tuesday night, I am watching two numbers. First, if Obama takes down a combined 98 pledged delegates then pledged delegate checkmate can officially be declared, with the remaining 37 proportional races guaranteed to give Obama at least a minimum 1 vote and thus put him over the top on the minimum viability alone.
Second, I want to see Obama erase Clinton's PA popular vote gain, which would finally drive a stake into that argument.
This diary is a straightforward analysis of what it will take to regain 214,224 votes.
1. Excluding Michigan and Florida, Obama has a popular vote lead of 610,000 votes.
2. Including FL's beauty-contest result and an exit poll based estimate in MI (which improves it from a non-contest where Obama wasn't on the ballot to sort of a FL-type beauty contest where no one could campaign), Obama currently has a 251,000 lead in the popular vote.
Obama now leads in all five counts that I am tracking:
Straight and Narrow Count: Popular votes plus state delegates, no Michigan or Florida
Obama: 9,284,899
Clinton: 8,568,941
Best Possible Obama Count: Popular votes plus estimated popular support from caucus attendees in states that only counted delegates. No Michigan or Florida
Obama: 9,560,675
Clinton: 8,761,747
Best Possible Clinton Count: Popular votes plus state delegates plus Michigan and Florida
Obama: 9,853,940
Clinton: 9,754,300B
Broadest Possible Count: Popular votes plus estimated popular caucus support plus Florida plus estimated Michigan support with Obama on ballot
Obama: 10,349,066
Clinton: 9,885,732
Mixed Broad Count: Popular votes plus Michigan and Florida plus estimated caucus popular support
Obama: 10,129,716
Clinton: 9,947,106
Combined with the pledged delegate counts, numbers like these indicate that if Obama sweeps the next six states through March 4th, then the campaign is over. Clinton needs wins in at least some of the following: Hawaii, Wisconsin, Ohio, Rhode Island, Texas and Vermont. If she fails to win at least some of those states, she can take it to the convention and argue about super delegates, Michigan and Florida all she wants, but it won't matter. At that point, super delegates will defect en masse, and the credentials committee will belong to Obama. In fact, Obama would probably lead even with Michigan, Florida and super delegates factored in. If Clinton does manage to win some of those states, then we are heading into the Pennsylvania Interval.
Update: In the interests of pointing out something more substantive, once Obama takes a lead of 202 in pledged delegates (currently 135 by my count), and is ahead in all of these different counts, then we will have a presumptive nominee.
Here are four different ways to count the current popular vote in the Democratic presidential nomination campaign. All of the data used to compile these numbers can be found in a document created for Open Left here:
1. Straight And Narrow Count Popular votes plus state delegates, FL and MI not included Obama: 8,119,171
Clinton: 7,916,422
2. Broadest Possible Count Popular votes, plus Florida, plus estimated popular support in state delegate states, plus estimated Michigan results with Obama on ballot Clinton: 9,233,213
Obama: 9,183,338
3. Best Obama Count Popular votes plus estimated popular support in state delegate states Obama: 8,394,947
Clinton: 8,109,228
4. Best Clinton Count Popular votes plus state delegates plus Florida plus Michigan with zero for Obama Clinton: 9,101,781
Obama: 8,688,212
I will update these numbers as the process moves forward. Other combination are possible, but these are the four where I will focus.
