The Fallacy of the Popuar Vote
Recently, Nancy Pelosi commented that the winner of the pledged delegate count should win the democratic nomination and disregarded the popular vote totals. Others have commented on how flawed the popular vote total is because it has no way to include those who vote in caucus states. The other significant problem is that it doesn't accurately reflect the will of Americans, specifically democrats in who they would like as their candidate.
The Boston Globe yesterday ran this article on Republicans voting in Democratic primaries:
http://www.boston.com/news/nation/articles/2008/03/17/many_voting_for_clinton_to_boost_gop/
Things to note from this article:
1. In Texas, exit polls showed that 119,000 Republicans voted for Clinton but fully intend to vote for McCain in November. Clinton won the Texas primary by almost exactly 100,000 votes. Remove the Limbaugh effect and Obama actually wins the Texas primary.
2. In Ohio, Clinton received 100,000 votes by Republican loyalists. Now Clinton would have still won Ohio, but it would have been by a much narrower margin.
3. In Mississippi, Clinton received 38,000 Republican loyalist votes, which made the blowout look smaller than it was. As most of these were white voters, it also skewed the vote to look more racially divided than it really was.
So, in total there have been 257,000 votes for Clinton by Republican loyalists which will be votes against the Democratic candidate in November. This makes the argument that the popular vote should count in determining the way the superdelegates should vote an absurd one. Take these votes out of the "total" and the popular vote is no longer even close.
The take home point is one that hopefully the superdelegates have noticed. Hillary Clinton's new best friends are John McCain and the Republican loyalists. When the these are the best friends of a Democratic candidate, it is time for that candidate to be done.
The Boston Globe yesterday ran this article on Republicans voting in Democratic primaries:
http://www.boston.com/news/nation/articles/2008/03/17/many_voting_for_clinton_to_boost_gop/
Things to note from this article:
1. In Texas, exit polls showed that 119,000 Republicans voted for Clinton but fully intend to vote for McCain in November. Clinton won the Texas primary by almost exactly 100,000 votes. Remove the Limbaugh effect and Obama actually wins the Texas primary.
2. In Ohio, Clinton received 100,000 votes by Republican loyalists. Now Clinton would have still won Ohio, but it would have been by a much narrower margin.
3. In Mississippi, Clinton received 38,000 Republican loyalist votes, which made the blowout look smaller than it was. As most of these were white voters, it also skewed the vote to look more racially divided than it really was.
So, in total there have been 257,000 votes for Clinton by Republican loyalists which will be votes against the Democratic candidate in November. This makes the argument that the popular vote should count in determining the way the superdelegates should vote an absurd one. Take these votes out of the "total" and the popular vote is no longer even close.
The take home point is one that hopefully the superdelegates have noticed. Hillary Clinton's new best friends are John McCain and the Republican loyalists. When the these are the best friends of a Democratic candidate, it is time for that candidate to be done.
Advertisement
B.Dub,
I love the way this analysis makes conclusions that that actually support Hillary's strength.
1. "In Texas, exit polls showed that 119,000 Republicans voted for Clinton but fully intend to vote for McCain in November."
Shows that Obama has no chance in the general election in Texas against McCain. If Obama can't convince these republicans away from Hillary, he certainly won't be convincing them away from McCain.
2. "In Ohio, Clinton received 100,000 votes by Republican loyalists. Now Clinton would have still won Ohio, but it would have been by a much narrower margin."
Duh? That sounds like win and maybe a win the general.
3. "In Mississippi, Clinton received 38,000 Republican loyalist votes, which made the blowout look smaller than it was. As most of these were white voters, it also skewed the vote to look more racially divided than it really was."
I don't understand the point of this exercise. You seem to be saying that white republicans loyalist should not be classified as being white. A different race all together?
This is comical stuff...
March 17, 2008 10:31 AM | Reply | Permalink
you're either being obtuse, or you didn't read the post.
1. Wrong. This does not show that Obama can't beat McCain in the general. This shows that 100,000 of those TX votes would go to McCain, that's all. You can't assume it's the same subset of voters in the primary, nor that all of Clinton's voters automatically go to McCain or don't show up. Look at the matchup polls for TX, and you'll see that Obama runs pretty well against McCain.
2. Wrong again. You can't translate primary results to a general election. Again, it's the polls describing specific matchups that are a more useful barometer.
3. Here's where you seem to be intentionally missing the point, so I'll spell it out for you. The media is making a big deal about the "racial divide" inside the Democratic Party. If those 38,000 Republicans that voted for Clinton had stayed home, Obama would have won a greater share of the white vote, thus insulating him against the ridiculous notion that he is just "the black candidate"
So see, once you divorce yourself from the notion that the primary and general election results are somehow connected, it's really pretty simple.
March 17, 2008 11:15 AM | Reply | Permalink
obtuse? The entire premise of this post is obtuse.
1. Its like saying take some of the black votes away from Obama in Missouri, and he loses.
2. A useful barometer for who? Clinton or Obama? Take a few percentage points away from Clinton in Ohio if you wish. So what! Its still a convincing win.
3. In some population centers with a high percentage of the black vote, there indeed is a racial divide. If you can't see that in the numbers, then you are denial. Don't conveniently sub categorize "white republicans" as being some kind of a sub species in order to prove your point.
Yes, it is pretty simple, your analysis is a farce.
March 17, 2008 11:54 AM | Reply | Permalink
1. No, it's like saying "take away some of the black voters who would otherwise vote for McCain" in Missouri. That's another level of complexity I don't think you can handle, but try.
2. Maybe I should've used smaller words...primary results are not a useful measure for predicting outcomes of general elections. For example, even in a state that hasn't voted yet, Clinton leads Obama for the nomination, but Obama does better than does Clinton against McCain.
3. I have no earthly idea what you mean by "Don't conveniently sub categorize "white republicans" as being some kind of a sub species in order to prove your point."
I don't even really know what you're accusing me of there, so I'll start over.
Fact A: Obama won the black vote 91-9 or something like that
Fact B: Clinton won the white vote 75-25, fueling speculation that there was a racial divide within the Democratic party.
Fact C: It turns out that a significant number of those white voters who went 75-25 for Clinton actually plan on voting for McCain.
Fact D: Without those "strategic voters," Clinton probably doesn't win the white vote 75-25, probably more like 65-35, or even 60-40.
For the purposes of the media, the "racial divide" only really pertains to white voters. That's because it's not unusual for African-Americans to vote in very large numbers for the same candidate, whether in a general election or in a primary. For example, in the general election in 1996, Clinton beat Dole in Mississippi (or maybe it was Alabama, I read this stat like two weeks ago) among African Americans 94-6.
So if the media is only looking at the "white voters" side of the equation, 65-35 is not a racial divide, it's a statistical blip.
March 17, 2008 2:43 PM | Reply | Permalink
this is spiraling into nonsense. to take it back to the original post -
1. Globe article does not say that 119,000 GOP voters who voted for Hillary in Texas would vote for McCain. The article only says 119,000 GOP voters voted for Hillary. Thats all it says.
2. I guess the poster is saying that all or some of 100,000 GOP votes for Hillary in Ohio are Limbaugh votes? Where is that number? I don't see any polls that say that.
3. Whats the point of that exercise. The media is just saying that the votes went along racial lines. OK, a certain percentage of the white vote came from the GOP. But so what. If the gauge is measured based on race, it doesn't matter which party. Its still a racial divide. If you (dkmoorhead) want to isolate the numbers to the democratic party only, a 65-35 divide is not a statistical blip. Thats still a pretty large divide.
You guys are all over the place. You are making assumptions about what the numbers represent and stating it as a fact.
March 18, 2008 3:04 AM | Reply | Permalink
Wow, just wow.
Airwon, do you really believe that, or are you just trolling?
March 17, 2008 12:27 PM | Reply | Permalink
Yes, just keep up the wowing until the smoke clears. And you will be in awe of what is revealed before you. And then you may become a believer in the absurdity of this silly post.
Thought I would try to appeal to your spiritual side as it seems to be very effective on Obama supporters. Unfortunately I need to work on my spiritual lingo...wow, no really, wow, no just wow...
March 17, 2008 1:04 PM | Reply | Permalink
You are clearly looking to elevate the discourse here.
Or should I say
Let your chakras be free of the negative vibes your horse race is generating.
March 17, 2008 6:02 PM | Reply | Permalink
Obama's gotten thousands of votes from outside the party too. Given that independents have a love affair with McCain, Obama can't really count on those votes in the general.
You're acting as if people outside the party have skewed the popular vote towards Hillary. Truth is, both candidates have benefitted from outsider votes. That's the problem with open primaries.
March 17, 2008 1:26 PM | Reply | Permalink
Both candidates did indeed benefit from votes outside the party.
Here's the difference.
Earlier in the race when the strategic dynamics of the Democratic race weren't being studied as heavily and the Republican race was still contested, Independent and Republican voters voted in the Democratic Primaries and mostly voted for Obama. Obviously I can't prove this, but this indicates to me that these voters actually *like* Obama. These folks had the option of casting votes that matter in a Republican Primary which presumably contained candidates that Republicans like more. They specifically chose to cross over and vote in the Democratic Primary and they generally did so to vote for Obama.
Since earlier March, when the Republican race was long settled and so votes in the Republican Primary don't have a chance at affecting the outcome, and when we saw the media dissecting the strategic dynamics of the Democratic race, and when we saw Rush Limbaugh and numerous influential Republican talk show hosts encouraging Republicans to cross over and vote in the Democratic Primary to specifically vote for Hillary Clinton, we have now seen a far greater number of Republicans voting for Hillary. This suggests to me that they are voting strategically, either to prevent Obama from winning the nomination because they view him as the tougher candidate or simply to prolong the race.
This can't be proved, and I don't see a point in trying to prove this. What it does say, however, is that the "popular vote" in a primary election is far different from a popular vote in the general election. A general election popular vote can generally be seen as a pretty good indicator of which candidate is favored by more people. A primary election popular vote is certainly a rough guideline but contains many more dynamics such as minimizing the caucus state votes and containing some amount of cross-party strategy voting.
In other words, the "popular vote" is a rough guideline but isn't the gospel of the primary that tells us, without a doubt, which candidate is favored by more people.
March 17, 2008 1:49 PM | Reply | Permalink
I don't know why but the incredulity and lack of skepticism by liberals/dems never fails to amaze me.
Here are the numbers -
In Iowa a 3% total of voters republican, they broke 44% for Obama, 10% for Clinton
In S. Carolina with 4% they broke for Obama 37% to 20% for Clinton
Alabama with 5% the break was 49% Obama 52% Clinton
Illinois with 6% 60% broke for Obama 36% for Clinton
Missouri with 6% 75% broke for Obama 21% for Clinton
Louisiana with %% 53% broke for Obama 17% for Clinton
Virginia with 7% 72% broke for Obama 23% for Clinton
Wisconsin with 9% 72% broke for Obama 28% for Clinton
Ohio with 9% 49% broke for Obama 49% for Clinton
Texas with 9% 53% broke for Obama and 46% went to Clinton
Mississippi with 12% they broke for Obama at 25% and Clinton at 75%.
It should be obvious to any sentient human being that the republicans are breaking for Obama big time and far less for Clinton. In Ohio and Texas both with 9% of voters republican the break was even in Ohio and Obama had the larger margin in Texas. This is from the GLOBE's own statistics. Does it occur to you that you're being had by republicans and the Globe?
March 17, 2008 2:25 PM | Reply | Permalink
So sorry to hear you're not a sentient human being, then, BevD, because what is obvious to me is that, as this post asserts, since Rush Limbaugh started urging Republicans to vote Clinton, 257,000 votes have been cast for Clinton that will not be cast for Obama in November.
If you can produce a poll that shows, among those crossover voters in the other states you list, how many of those people who voted for Obama in the primary will vote McCain, then it's almost an apples-to-apples discussion, since (for one thing)the GOP nomination had yet to be decided when a lot of these primaries took place.
Until then, you've got nothing that shows me that the Obama votes in Missouri, Wisconsin, et. al. aren't genuine new Democratic voters, while I (actually, the original poster of this) have numbers showing you that Clinton's support is weaker than advertised.
March 17, 2008 2:51 PM | Reply | Permalink
You don't have numbers that support anything. You have anecdotal evidence consisting of a few stories by REPUBLICANS of which you're naive enough to take at their word. There was NO difference in the number of republicans who broke for democrats in Ohio the numbers were EXACTLY the same for Clinton and Obama. In Texas the opposite is true, the majority of republican voters broke for Obama, NOT Clinton.
If 3,800,000 voters went to the democratic primary in Texas, and 9% were republicans voting for democrats, that means that 343,000 voted for a dem, if Clinton had 119,000 of those votes, that means that 226,000 voted for Obama. There's no effect here, in fact the numbers say exactly the OPPOSITE to what you're claiming. With the two states mentioned, Ohio and Texas, Ohio broke evenly for both candidates and in Texas, Obama received the majority of republican votes. The MATH contradicts the claim. There is NO poll that states that 257,000 voters will vote for McCain who would not vote for Obama - that doesn't even make sense. There is NO poll that surveyed the 300,000 republican crossovers in either Ohio or Texas where Limbaugh claims success that would even indicate an anomoly much less a trend.
Now here's a little clue for you - if Limbaugh can't influence enough republican voters to vote for Huckabee, what makes you think he can influence enough republican voters to break for Clinton? These are anecdotes, not evidence. Why would anyone in his right mind take the word of Limbaugh for ANYTHING?
March 17, 2008 5:22 PM | Reply | Permalink
While you and Desidero pat each other on the back, perhaps you could tell me when Rush made his plea relative to the numbers you posted.
The 2008 Mississippi Democratic primary took place on March 11, 2008
The Texas Presidential Primary was held on March 4.
Based on the numbers BevD posted, that would imply that Rush had some potential to influence the primaries following his request, as opposed to the commonly expected influence reflected backwards in time.
I guess us folkses have herd times with dates an all. must be that koolaid r somthin
March 17, 2008 6:01 PM | Reply | Permalink
And don't forget that Bill Clinton was on Rush Limbaugh's show (with Mark Davis filling in for Rush) the day before the TX primary promoting Clinton.
March 17, 2008 6:47 PM | Reply | Permalink
Look at the damned numbers. If there was a "Limbaugh effect" in Texas and Ohio, then the numbers would favour Clinton NOT Obama. In Ohio there was a 1/1 break, for every one voter that broke for Clinton, one broke for Obama. In Texas, for every one voter that broke for Obama, .086 broke for Clinton. It has nothing to do with "cults" or "partisanship" or anything else, the NUMBERS speak for themselves.
It doesn't matter what Rush says, statistically the numbers say the exact opposite of what he is claiming. If there was a Limbaugh effect the numbers would be greater for Clinton AFTER he announced his "strategy" for Ohio and Texas - there was NO jump in republican voters in either state for Clinton. The Wisconsin primary that followed Ohio and Texas broke for Obama 72% to 28% for Clinton.
Now I hate to destroy your faith in Limbaugh, it's obvious that you respect the man and his strategic thinking, but he is wrong.
March 17, 2008 9:19 PM | Reply | Permalink
Bev,
Look at the numbers yourself. Notice anything after awry? Like the huge jump in support amongst republicans in the last three contests for Hillary while Obama's remained relatively stable?
Now, what could explain this change? Could it be all of the evidence cited above with prominent republicans pushing for cross over votes to Hillary? We can't be sure, but that is the most logical explanation.
Your argument that Clinton and Obama's support amongst republicans is equal misses the point.
Consider this: there is no push from any republican sources to vote for Obama, yet he still gets a number of their votes. There HAS been a considerable push from many Republicans (Limbaugh, Coulter, etc,) to crossover for Clinton. This could be because they know Clinton energizes their base or because they think Obama is a stronger candidate, because they think Clinton is a better alternative to Obama (the lesser of two evils), or maybe because they genuinely prefer Clinton to McCain. Given the history of the Clinton's I find the last scenario unlikely.
So thanks for providing the numbers, they clearly prove the existence of the 'Limbaugh democrats'.
March 18, 2008 11:25 AM | Reply | Permalink
Well the only thing I can say then is that you're either dumb, you don't understand statistics, you don't understand the difference between anecdotal and empirical evidence or you just refuse to believe that Rush Limbaugh is a big, fat, lying idiot. I'm thinking that it's probably the latter.
March 18, 2008 1:54 PM | Reply | Permalink
Don't confuse 'em with numbers, Bev - numbers ain't words. You know, "figures lie and liars figure" and all that. Just say, "When Obama asks Republicans to cross over and vote for him, that's good. When Republicans do it on their own, that's good. When Republicans vote for Hillary, that's bad. When Rush asks them to do it, that's evil."
Simple arguments for simple folkses. After all, math is tough - any Barbie can tell you that.
March 17, 2008 5:00 PM | Reply | Permalink
Some of these people are so blinded by partisanship that they would rather believe some lie from Limbaugh than the numbers - what the hell is the matter with these people? Limbaugh of all people!
March 17, 2008 9:24 PM | Reply | Permalink
BevD,
As usual I totally agree with you on this one. I justed wanted to point out that if we are talking about the same Globe article, it concludes that while possible, it is unlikely that enough strategic voters turned out to make the "Limbaugh effect" effective. The Globe is actually with us on this one.
March 17, 2008 11:31 PM | Reply | Permalink
The state most in question right now, i.e., the only state to have a primary after McCain has earned the GOP nomination, is MS. And, in MS, Clinton received an astonishing 75% of the crossover vote.
Some of that might have to do with it being a southern state (look at the AL crossover numbers, for example). But MS is an outlier compared to all of the other states that have voted.
We really can't conclude from that one primary that crossover voters are listening to Limbaugh. There is anecdotal evidence to support it, and some numbers to support it, but we will have to see what happens in the remaining primaries.
Keep in mind that Bill Clinton did go on Rush's show to promote Hillary the day before the primary. And Hillary did essentially agree that Limbaugh's supporters did cross over to vote for her:
If, in the remaining open primaries, the GOP vote for Clinton continues to be so high after having favored Obama during the early states, it will demonstrate more strongly that the GOP is trying to throw a wrench in the Democratic nomination process by voting for Hillary.
March 17, 2008 5:44 PM | Reply | Permalink
Shhh. You sound too logical. That is against cult policies.
March 17, 2008 6:04 PM | Reply | Permalink
There are NO numbers that would support Limbaugh's claim. None. There are some anecdotes, but even those numbers if indeed the stories are true, are too small to make any kind of demographic, even a liars demographic.
Hillary Clinton didn't "essentially agree" that Limbaugh's supporters voted for her, she was making fun of Limbaugh.
Out of 12 primaries the republican vote favoured Clinton in two primaries, Alabama and Missippi and there is NO evidence even anecdotal that the break out vote had anything to do with Limbaugh. There is no "astonishing" crossover vote - Virginia went 72% to Obama, Missouri went 75% to Obama, Wisconsin went 72% to Obama with the breakout vote. If you're going to extrapolate meaning, you could say that this "astonishing" breakout vote went to Obama because they wanted to make sure Clinton lost the nomination. It would be just as ridiculous as the opposite claim made now. The two primaries Limbaugh claims credit for are exactly the opposite of his "effect" - 49/49 Ohio and 53/46 Texas with Obama receiving the majority of the breakout vote.
March 17, 2008 10:01 PM | Reply | Permalink
Bdub the original poster asserts,
"In Texas, exit polls showed that 119,000 Republicans voted for Clinton but fully intend to vote for McCain in November."
"In Ohio, Clinton received 100,000 votes by Republican loyalists. Now Clinton would have still won Ohio, but it would have been by a much narrower margin."
How many of those 119,000 in Texas and 100,000 in Ohio represent the crossover Limbaugh effect vote?
You don't have the numbers that would prove your point. You are just making assumptions that the 119,000 in Texas reflect the Limbaugh effect.
This isn't to say that conservatives have not tried to influence the Democratic primary. They may have. We just don't know to what effect. Check out this post from James Wolcott in Vanity Fair.
http://www.vanityfair.com/politics/blogs/wolcott/2008/03/ive-never-bough.html
March 17, 2008 11:52 PM | Reply | Permalink