A closer look at ‘popular’ and ‘overall’ votes
Greg Sargent at TPM has referenced a finding in the ABC News/Washington Post poll released today that merits discussion.
First, let me say I think the folks at TPM do a great job and more importantly provide a valuable service. I don’t want anyone to infer I am bashing Greg or TPM. The Democratic primary has generated a lot of harsh remarks between progressive blogs that historically have had relatively common themes and fundamental purposes. I applaud Josh Marshall for not taking aim at other progressive blogs when he has had more than enough justification to do so. My point here is to offer a different perspective and to illustrate how the interpretation of polls can be misleading, including my own. Candidates frequently and intentionally reference polling data inaccurately to accommodate whatever spin they’re trying to push.
The title of Greg’s post is "Poll: Plurality Thinks Super-Delegates Should Back Winner Of Popular Vote" and he focused on Question No. 18 in the poll (pdf -page 16). Following are the question asked and the results
18. (ASKED OF LEANED DEMOCRATS) The Democratic nomination may be decided by so-called "super delegates" who can pick any candidate they choose. Do you think the super delegates should support the candidate who won the most (delegates) in primaries and caucuses; the candidate who won the most (overall votes); or the candidate they think is best, regardless of either delegate or vote totals?
| Delegates | Overall Votes | Candidate they think is best | No opinion |
| 13 | 46 | 37 | 4 |
This is Greg’s comment (emphasis added):
The Hillary campaign is clinging to the hope that she’ll somehow be able to eek out a long shot popular vote win, and support here appears stronger for the idea that the popular vote is the metric super-dels should use.
Obama, of course, is on track to win both the delegate count and the popular vote, and the total support voiced here for supporting either of those is 59% — far higher than the 37% saying super-dels should exercise their own judgment.
I know I am getting way down in the weeds on this issue, but as Greg rightfully noted, this is the very thing Hillary Clinton will attempt to exploit and exaggerate ad nauseam. Therefore, I believe it is important voters can discern between what they may be told and reality.
First, the poll does not define "overall votes." Although, "popular vote" is strongly implied in the question, and Greg seems to have reasonably inferred it that way given his comment. It is important to remember "popular vote" in the Democratic nomination process does not in any way include or is influenced by the number of people that participated in caucuses. It only reflects the number of votes in states that had actual primary elections, and there were many states that only held caucuses.
There is no evidence in the poll’s documentation that suggests the pollster clarified "popular overall vote" to the respondents. Therefore, one has to question how many people were cognizant of that fact and took it into consideration when they answered the question. I can’t prove it, but I instinctively believe many respondents, if not the majority, would not have known what "overall votes" really means. Without further evidence to the contrary, one can only reasonably conclude there is a certain amount of latitude in the "overall votes" response.
There may be an exception or fault in my logic. The poll’s margin of error may account for any misinterpretation of "overall votes." But the poll’s documentation does not give any indication how margin of error was determined. Therefore, in my opinion, it is better to err on the conservative side and assume no provision was made for misinterpretation of the specific term.
Second, although Greg accurately states "plurality" in the title of his post, his comment, in my opinion, is more indicative of a majority than a plurality. Greg writes,"support here appears stronger for the idea that the popular vote is the metric super-dels should use." From my perspective a majority is out of the question and plurality is reasonably debatable.
The definition of plurality, as applied here, is:
In a contest of more than two choices, the number of votes cast for the winning choice if this number is not more than one half of the total votes cast.
When considering the margin of error (+/- 3), the four percent that had "no opinion," and the possibility of respondents misinterpreting overall or popular votes, it is not reasonable to conclude overall votes, as determined by this poll, is valid for for any substantive argument. But as Greg appropriately predicted, Hillary Clinton will likely reference this or other similar data to make an incredulous argument that she must be crowned the party’s nominee.
Question - what about the 257,000 “illegal” Republican votes that were sent to vote for Hillary in Texas, Ohio and Mississippi.
Texas - 119,000
Ohio - 100,000
Miss - 38,000
I thought it was an FEC violation to switch parties and vote in order to influence the result? It is well documented all over the web that Rush Limbaugh et al mobilized Repub loyalists in these 3 states to “bloody up Obama”, “keep the uncivil war” going and try “prevent McCain from running against Obama.
Now, that this is public knowledge, isn’t there technical and ethical justification to render these nul and void, since Hillary didn’t “win” them fair & square?
Not only would this take Obama’s popular vote lead up to 1,084,308, once the 257,000 was deducted from Clinton’s total, a margin she could never catch up to, but it would alter the outcome of that much discussed BIG state, Texas.
When 119,000 votes are deducted from Clinton’s 1,459,814, she’s left with 1,340,814 votes. Well, Obama won a total of 1,358,785 votes in the Texas Primary - with the illegal votes anulled, Obama would be the winner of both the Texas Primary & Caucus. Then Hillary would have to stop saying he can’t win the big states, because her win in Texas is a fake.
Right or wrong?
http://www.boston.com/news/nation/articles/2008/03/17/
Annemarie -
First, just an fyi. The link you gave seems to be incorrect. It returns “page not found.”
I don’t know how the FEC rules may (or may not) apply in the situation you articulate, but it is reasonable to assume theoretically it may apply under some law that prohibits “rigging” elections. I know they exist, but without reading the specific statute, I can’t begin to guess what standard or criteria must be met to qualify for violating the law.
The current problem with enforcing any FEC regulations today is that in order for the Commission to take any action or make any ruling the commission must have enough commissioners present to constitute a quorum. Since four of the six seats on the commission are vacant, they can’t call a quorum, consequently, everything sits in limbo. This is the sole reason why McCain is still floating along in spite of violating campaign finance laws that could put him in prison for five years.
Bush has demanded that Congress approve all four of his nominees as a group, rather than individually, which is the way it should be done. Rightfully so, Congress has refused to consider any “wholesale” approval of nominees. One of Bush’s nominee’s is very controversial, but I can’t remember who it is at the moment.
Getting to the specifics of the point you raised, I imagine it would be very hard to prove, beyond a reasonable doubt, that all the purported Dumbbaugh votes were in fact violating the law. I’m certain probable cause would not be met since Dumbaugh did not speak directly to the voter(s). The voters could easily assert they decided to vote for Clinton or whomever, and accordingly they executed their right to choose. Dumbaugh making requests over the radio for people to vote for Clinton is not any different from a legitimate campaign ad for Clinton. Now, it proves the people who did his bidding are about as dumb as brick, but…that’s about it.
I think, if it gets down to literally counting votes, the party is going to be in a world of hurt — more than any vote-count decision will ameliorate. But, if they do, they will have to come up with some method of accounting for people who voted in caucuses, which will not be transparently accurate to voters and will set off another kerfuffle.
I did not put this in my post because I can’t remember the precise specifics of the research and it was going to be way too time-consuming to dig it up. Anyway, I did hear a pundit talking about a very reputable professor (Boston College maybe?) that had actually developed an algorithm for determining how many people voted in caucuses.
Summarizing, based on his research and calculations, he concluded if caucus participants were included in the “overall” or “popular” vote, Obama would have a net lead of about 2 million votes over Clinton. That 2 mill may not be correct, but it is reasonably close.
I do remember them stating, in effect, if the calculated votes were applied, it would push Obama well past any threshold that Clinton could possibly meet, much less surpass. And that included “worst case scenarios” (advantage Clinton) for whatever votes could be garnered in the states where primaries had not been held yet (PA, NC, etc.)
At this point, I really don’t think Clinton is giving popular vote, or anything remotely related, a second thought. I believe her strategy is to try to convince super-delegates and pledged delegates that Obama cannot be elected in the general. She is trying everything possible to destroy him personally — it’s the only option she has left.
All the other options defy the laws of mathematics and probability. So, there is no realistic basis for her campaign to expend time and money on those efforts. Now, that’s not what she or her campaign will say publicly, because they are playing serious mind-games with the citizenry. The key is not allowing Clinton to win her psychological warfare, which will not be easy because the media just sucks it all up. And unfortunately, that influences people, and is a big, but separate problem, which I discussed in this post.