Obama is not losing white voters as contests go on

On the Front Page right now, Jerome is suggesting, that Obama is systematically losing the white vote.  The problem is there is no data to back this up.  Let's take a look at the CNN exits...

I've compiled a list of all states from Super Tuesday on which were primaries.  I decided to take out Illinois and Arkansas, as both home states saw an obvious swing in a certain direction.  I left New York in primarily because white voters in New York behaved very similarly to other states.  

Alabama - 47% Clinton lead
Arizona - 15% Clinton lead
California - 1% Clinton lead
Connecticut - 1% Clinton lead
Delaware - 16% Clinton lead
Georgia - 10% Clinton lead
Massachusetts - 18% Clinton lead
Missouri - 18% Clinton lead
New Jersey - 24% Clinton lead
New York - 24% Clinton lead
Oklahoma - 14% Clinton lead (note, Edwards got 10% here)
Tennessee - 44% Clinton lead
Utah - 15% Obama lead

Louisiana - 28% Clinton lead
Maryland - 10% Clinton lead
Virginia - 5% Obama lead
Wisconsin - 9% Obama lead

Ohio - 30% Clinton lead
Rhode Island - 26% Clinton lead
Texas - 11% Clinton lead
Vermont - 22% Obama lead
Mississippi - 44% Clinton lead

Pennsylvania - 26% Clinton lead
North Carolina - 24% Clinton Lead
Indiana - 20% Clinton lead

Notice, Obama just won the white vote in primaries in four states, ultra-white Utah and Vermont, and Virginia and Wisconsin, when he was riding high in the nationwide zeitgeist.  

Overall, the raw numbers show no shift except during his best period in late February. On super-Tuesday, he lost the white vote by 17.1% in the average contest.  In late February, he lost it by 6%.  In March, he lost it by 17.8%, and across PA, IN, and NC, the average has been 17.5%.  Please note these numbers are just averages across contests, not weighted by population.    

Anyway, grouping the states by type shows further that there has been no effective change


AL - 47% Clinton
GA - 10% Clinton
OK - 14% Clinton (24% if Edwards is included)
TN - 41% Clinton
LA - 28% Clinton
MD - 10% Clinton
VA - 5% Obama
TX - 11% Clinton
MS - 44% Clinton
NC - 24% Clinton


CT - 1% Clinton
AZ - 15% Clinton
CA - 1% Clinton
DE - 16% Clinton
MA - 18% Clinton
MO - 18% Clinton
NJ - 24% Clinton
NY - 24% Clinton
UT - 15% Obama
WI - 9% Obama
OH - 30% Clinton
RI - 26% Clinton
VT - 22% Obama
PA - 26% Clinton
IN - 20% Clinton

I'm not going to sit down at this moment and do a regression analysis of each state, but there are only three clear trends: Obama tends to win the white vote in states with virtually no blacks (Arizona was the only exception), Obama did better than expected in is post super-Tuesday run-up, and Obama performs better with white voters in states which are wealthier and more educated.

That's all.  His performance from Ohio on just isn't that different from his performance on Super Tuesday (at worst it maybe dropped by 4%-6%).  No systematic collapse to be seen.  

Tags: black, clinton, obama, Primary (all tags)



Re: Obama is not losing white voters as contests g

Thank you!

by Mandoliniment 2008-05-08 08:01AM | 0 recs
Re: Obama is not losing white voters as contests g

Clinton's average margin of victory over Obama with white voters: 16.8%.

by hornplayer 2008-05-08 08:04AM | 0 recs
Re: Obama

You might want to edit your diary. I asked the question, not made the assertion.

Thanks for doing the analysis. The best way to categorize it, to relate to the other graph, is by southern and non-southern, and as a percentage of the vote that Obama got in each state.

by Jerome Armstrong 2008-05-08 08:04AM | 0 recs
Re: Obama

I altered it to say "suggest" as that's more accurate.  

by telephasic 2008-05-08 08:11AM | 0 recs
Re: Obama

"What?  There are people point it out everywhere. I just did. Obama has done less and less well with white people."

This is a quote from a comment you made in the referenced diary.

by soccerandpolitics 2008-05-08 08:13AM | 0 recs
Re: Obama

Yea, I pointed it out, and asked, has he?  It's certainly what's been argued, but I'm willing to take a look and see if thats actually the case.

by Jerome Armstrong 2008-05-08 08:18AM | 0 recs
Stay classy, Jerome!

by clad 2008-05-08 09:33AM | 0 recs
Re: Obama

With all due respect, how about if you add this information to your diary, Jerome? It deserves to be there along with the chart you've posted on black voters or put up as a front-page diary.

by politicsmatters 2008-05-08 08:31AM | 0 recs
Thank you.

I am re-posting a comment from Jerome's 'analysis' - which your diary supports.

Jerome's front page (surprise) diary http://www.mydd.com/story/2008/5/8/10371 9/1420 is revealing the very disturbing and patently false spin that the MSM has attempted. Why leave a rhetorical question at the end regarding Obama among various white demographics? Why not post the exit polls from OH to PA to IN? Or better - why not post favorability ratings among white and african-american democrats since Jan 2008? Why? Because these will show that Senator Obama has in fact continnued to build a coalition AND is viewed more favorably among democrats now then he was 2/3 months ago. Clinton....not so much.

All those who are blindly letting Jerome and others create a narrative for you, should read the post linked here.


Clinton's comments (whether her own words or putting the AP in her own words) were calculated and I think we will see potentially destructive to her campaign. If you open your eyes and look around - you will see the Hillary Clinton has decided to feed the MSM obsession and also to further divide the democratic party.

Finally - why would Jerome find it necessary to defend Hillary's comment? Finally, Finally, why even push this narrative? Is it helpful for electing democrats?

by Newcomer 2008-05-08 08:05AM | 0 recs

We still want to do way better with white voters in the general, of course, but I'll be interested in seeing what, say, Oregon will look like now that he's the presumptive nominee.

by Dracomicron 2008-05-08 08:05AM | 0 recs
Re: Interesting

Sorry to disapoint, but Oregon won't tell you anything. Oregon is vote by mail so we've already started voting and sending ballots back and because we're vote by mail you won't get any exit polls.

by Obama Independent 2008-05-08 08:24AM | 0 recs
Re: Interesting

Are there any votes on election day in Oregon or is it completely vote by mail?

by politicsmatters 2008-05-08 08:29AM | 0 recs
Re: Interesting

Oregon only uses mail in ballots - no polling places.

by CA Pol Junkie 2008-05-08 08:45AM | 0 recs

Thanks for the info.

by Dracomicron 2008-05-08 08:30AM | 0 recs
Jerome's summary is clear

and things won't get better from here on out with McCain getting more airtime.

by observer5 2008-05-08 08:08AM | 0 recs
If anyone with power here is listening.

This diary deserves to rest next to Jerome's on the front page.

by Newcomer 2008-05-08 08:11AM | 0 recs
'Rust Belt Votes'

To me, three figures from the analysis above shatter this bullshit narrative that the media and now HRC (and implicitly Jerome) is trying to push about Obama and the white working class vote:

OH - 30% Clinton
PA - 26% Clinton
IN - 20% Clinton

Hillary apparently wants to talk about a pattern emerging...I see one, do you?

by Newcomer 2008-05-08 08:15AM | 0 recs
Too small of a sample

and atrributable to reasons other than a time pattern.  The margins in PA and OH are fairly indistinguishable.  What I think distinguishes Indiana from PA and OH are the following:

1) Indiana is a much more Republican state in terms of voting behavior than PA or OH.  When that is the case, the Democratic primary electorate tends to be a little more homogenously liberal;

2) Indiana is less Catholic than PA or OH and Clinton has achieved much larger margins among white Catholics than white Protestants during the primary season.  Indiana also had a relatively large percentage of primary voters indicating that they were neither Protestant nor Catholic.

by lombard 2008-05-08 08:59AM | 0 recs
Re: 'Rust Belt Votes'

"Obama has proven - smearing and falsely accusing opponents of racism works!
Perhaps other AA candidates will use this winning strategy."

go scr*w yourself  try that for strategy and why don't you just try veiwing him as a candidate rather than just a AA candidate  racsist prick

by wellinformed 2008-05-08 09:26AM | 0 recs
Demographic analysis

Poblano proved with his accurate predictions of North Carolina and Indiana that demographics predict everything: not just race, but age, gender, income, and eduction.

Nothing else matters: not Wright, not Ayers, not "bitter", not Tuzla, and not the gas tax pander.  It is both reassuring and troubling at the same time: trivial stuff truly doesn't sway the electorate, but the outcome of the primaries has been set in stone since the beginning.

by CA Pol Junkie 2008-05-08 08:16AM | 0 recs
Re: Demographic analysis

Well, since the beginning, there has been a shift among the percentage of the black vote that Obama gets, I'll look into the white vote for him and post it later.

by Jerome Armstrong 2008-05-08 08:19AM | 0 recs
Re: Demographic analysis

There's no single "white vote".  White voters in the Research Triangle don't vote the same way as white voters in the Blue Ridge Mountains.

by CA Pol Junkie 2008-05-08 08:29AM | 0 recs
Re: Demographic analysis

And they don't vote the same as white voters in Maine and in Montana.

by politicsmatters 2008-05-08 08:30AM | 0 recs
Re: Demographic analysis

and older white voters are not voting the same way as the younger generation..

by soros 2008-05-08 08:54AM | 0 recs
Jerome you are wrong

NY Times did an article a couple of Sundays ago that showed that Obama has increased his support among white voters while Hillary has seen a 36% point decrease among blacks.  You need to do your research before you post.

What Hillary said is offensive to this black and many others. She has made very clear that we do not count. Should she somehow get the nomination she will lose as Rasmussen did a poll that showed that she barely got 50% of the black vote.  You can't win with that number.

by sweet potato pie 2008-05-08 08:49AM | 0 recs
Obama's white vote problem is a myth

If you break down the white vote into male and female there is even less of a "white voter problem" then the talking heads claim. Hillary Clinton has a strong hold on the white woman vote and because white women make up the majority of the white vote this causes the numbers to look much worse for Obama then they actually are. I went through the exit polls available and pulled them out.

Alabama M 70 27 F 73 23
Arizona M 46 45 F 58 34
Arkansas M 71 24 F 85 11
California M 35 55 F 56 36
Connecticut M 40 57 F 56 42
Delaware M 47 47 F 61 35
Florida M 45 27 F 59 21
Georgia M 46 48 F 57 40
Illinois M 37 59 F 43 56
Indiana M 59 41 F 61 39
Louisiana M 51 31 F 63 29
Maryland M 45 48 F 56 38
Massachusetts M 48 49 F 65 34
Mississippi M 68 30 F 71 23
Missouri M 55 41 F 59 38
Nevada M 46 40 F 55 31
New Hampshire M 30 38 F 46 33
New Jersey M 58 39 F 72 27
New Mexico M 38 59 F 46 51
New York M 52 43 F 65 33
North Carolina M 55 42 F 65 33
Ohio M 58 39 F 67 31
Oklahoma M 55 32 F 56 26
Pennsylvania M 57 43 F 68 32
Rhode Island M 52 47 F 71 29
South Carolina M 28 27 F 42 22
Tennessee M 58 32 F 73 21
Texas M 49 49 F 59 40
Utah M 29 64 F 49 49
Vermont M 35 64 F 41 56
Virginia M 40 58 F 53 47
Wisconsin M 34 63 F 52 47

by Obama Independent 2008-05-08 08:50AM | 0 recs
Re: Obama's white vote problem is a myth

Forgot to explain, the first number is Clinton and the second is for Obama.

by Obama Independent 2008-05-08 08:51AM | 0 recs
Re: Obama's white vote problem is a myth

Traditionally, women are more likely to vote for the Democratic candidate in a general election.  Is this is a hidden advantage for Obama, should he win the nomination?  Many female voters - especially older female voters - are undoubtedly disappointed that a woman will probably be denied the nomination, so they may be more likely than men to say they won't vote for Obama in the GE now AND be less likely to follow through in the fall.  If not for HRC, they might have been in Obama's corner all along.

Just pondering...

by KTinOhio 2008-05-08 09:08AM | 0 recs
Re: Obama's white vote problem is a myth

There is no question that some will be disappointed should Senator Clinton fail to win the nomination. And there is little doubt that there will be a small segment that will blame Senator Obama for defeating her. But elections are not decided in a vacuum. You can't say that those who will choose Clinton over Obama in a primary will automatically go with McCain in a general election. It's said that you vote with your heart in the primary and your head in the general election. When you look at the disaster the Republicans have presided over during the Bush administration, when you look at the fact that a new Republican administration will turn the Supreme Court hard conservative for decades to come it's easy in my mind to go with the Democratic nominee, whomever they end up being. In 2000 most of John McCain's supporters swore they would never support Bush and that never came to pass.

by Obama Independent 2008-05-08 09:30AM | 0 recs
I wonder...

HRC's largest margins come from the oldest voters.

African-Americans, especially African-American men, have lower life expectancies than whites, so it stands to reason that they are underrepresented among the elderly.

Could this explain - at least in part - HRC's large margins among the elderly?  And, if the oldest cohort (60+ or 65+, depending on the analysis) is excluded, how much closer are the shares of the total vote and the white voe won by the two candidates?

by KTinOhio 2008-05-08 09:01AM | 0 recs
Re: I wonder...

If I had to guess, I'd say that there are some pretty unflattering reasons why older white voters don't vote for Obama.  This is just a guess.

by the mollusk 2008-05-08 09:17AM | 0 recs
Maybe so...

...and I won't deny that those old enough to have grown up before the civil rights era might be a tough nut to crack.  My point was that the dead don't vote in large numbers (in most places, anyway), and all else equal, it makes a difference if a cohort is 11% black instead of 17% black.

by KTinOhio 2008-05-08 09:26AM | 0 recs
Re: Maybe so...

This would be a good time to look at voting patterns by different demographic variables taken together.  We might know each candidate's share of the white vote, or the female vote, or the age 18-29 vote, or the born-again Protestant vote; we don't know how well each candidate did among white born-again females aged 18-29 -- or how numerous such people are in the population.  That would tell us something.

by KTinOhio 2008-05-08 09:30AM | 0 recs
You'd need exits...

covering 10,000 people in order to segregate groups according to three variables and have non-trivial margin of error.  

by telephasic 2008-05-08 09:55AM | 0 recs
Re: You'd need exits...

Very true.  We'd have to be very selective in choosing our variables and strata within variables.  But even a detailed breakdown of whatever numbers we get now would be better than nothing.

by KTinOhio 2008-05-08 10:30AM | 0 recs
It doesn't matter.

I doesn't matter if Obama loses the white vote, as long as he wins enough Blacks and Latinos to make up for his loss among Blacks.  Look at this differently.  If you insist that Obama (or any Democratic presidential candidate) must be able to win the white vote in November then, after adding in the Black vote as well, you are saying that they candidate should win by a 13% margin.  When has that ever happened to a Democrat?

In fact, EVERY Democratic presidential candidate loses the white vote, but makes the loss up among Blacks.  That's not news.  It's simply reality.

by Manic Lawyer 2008-05-08 12:31PM | 0 recs
Re: It doesn't matter.

That's not reality. The Democratic candidate doesn't always make up for the lose of the white vote with black votes. More often than not, the formula of 90% of blacks plus a minority of white does't work for the Democrats.

by Mayor McCheese 2008-05-08 12:52PM | 0 recs


Advertise Blogads