Clinton Hangs On, Obama Drops in Polls, With Negative Media Attention

Aka Obama's Sour Apples to Apples, Part Four

Between late February and mid-April, voters in nine states that should/could be "Democratic" in the 2008 Presidential Election were exposed to a considerable amount of negative informative concerning both Hillary Clinton and Barack Obama. The negative information had little impact on how Hillary Clinton was regarded when compared to John McCain.  But it has had a major impact on their perception of Barack Obama, and on the perception of the relative merits of Obama and McCain.

Overall, the worst that can be said about Clinton is the negative press attention has resulted in more previously undecided voters in certain demographic categories expressing a preference for McCain rather than for Clinton.   But Obama is not merely losing "undecided" voters in most demographic categories because of negative media coverage, a significant percentage of voters who had supported Obama over McCain have switched their preferences.

crossposted at Correntewire

The impact of negative media stories can be seen in the changes in the support of White voters for both candidates in hypothetical match-ups against McCain, and the margins among White voters in those match-ups.  Clinton gained support among White voters overall, but her margins against McCain declined in that demographic as more previously undecided voters expressed a preference for McCain.   Obama, on the other hand, saw a decrease in White voter support, and his margins against McCain among White voters dropped precipitously.

These conclusions are drawn from an "apples to apples" comparison of Survey USA polling done April 11-13 in nine states (California, Iowa, Massachusetts, Minnesota, Missouri, New Mexico, Ohio, Oregon, and Wisconsin)  with similar polling done in late February as part the SUSA 50 state poll (conducted Feb. 26-28).  

This is the third of a multipart series examining the polling data in these states.  Part One provided an overview of the polling results, and showed how Clinton was doing better than Obama in 8 of those 9 states.  Part TWO provided an overview of key demographic categories, and took a close look at male and female voters.  Part Three examined "Party Identification" and "Ideological" demographic breakdowns, with an emphasis on the "Independent" and "Moderate" subcategories.  Part Four will look at the media context in which these changes took place, and examine the impact of "new" information on Clinton and Obama in the White voter demographic.


In order to understand the differences that appear in the late February and mid-April polling of the Clinton v McCain and Obama v McCain match-ups, it is necessary to consider the media environment prior to the February poll, and what that environment looked like between the two polls.

In late February, Obama was at a high point; having won 11 straight post Super Tuesday primaries and caucuses.  Texas polls (and the media) were suggesting that Obama would win in Texas, and perhaps even in Ohio.   Up until that point, Obama had received very little in the way of negative media attention and scrutiny, while Clinton was a constant target of attacks in the media.

But between the two polls, a number of significant events for the campaigns took place.  On the whole, negative media attention and scrutiny was focused on Obama during the first three weeks after the February poll was taken, but that focus shifted to Clinton in the weeks prior to the mid-April poll.

    ·    Hillary Clinton released her "3 AM" ad on February 29th.

    ·    The Rezko trial began on March 6th and increased scrutiny of Obama's relationship with Rezko was not far behind.

    ·    The "Jeremiah Wright" controversy erupted --ABC was the first broadcast network to feature the story, on March 13th

    ·    Clinton's "appointment calendar" from her years as First Lady was released on March 19th

    ·    The "phantom Bosnian snipers" story received a great deal of media attention, beginning around March 21st.

    ·    Clinton released her tax returns on April 4th.

All of this information was being digested by voters in the time between the two polls.  In addition, stories about Obama's "bitter/clinging" remarks at a San Francisco fundraiser were gaining currency just as the April poll was being taken - the "bitter/clinging" story first appeared on April 11th, the first of the three days of April polling.

In other words, between the two polls, the media was full of stories raising questions about the character and experience of both Clinton and Obama.   Clinton has weathered these storms - despite all the negative press that was concentrated close to the mid-April polls, Clinton's average margins against McCain held firm.  But the questions raised in early March about Obama's character and experience had a strong negative impact on his margins against McCain.


    Late Feb  mid-April    Late Feb  mid-April  

CA    10%       13%            11%        7%

IA    -5%        -6%             9%         7%

MA    18%       15%             7%        2%

MN     4%         1%             7%         6%

MO    -4%         1%            -6%         -8%

NM     0%        -3%             7%         -6%

OH    10%        11%            10%         -2%

OR    -5%         1%             8%         9%

WI     4%         0%            11%         5%

AVG    3.6%        3.7%           7.1%        2.2%

    ·    Clinton was behind McCain in 3 states in February, and in 2 in April.  Obama was trailing McCain in 1 state in February, and 3 in April.

    ·    Clinton maintained the double digit leads she held in three states (CA, MA, OH).  In the three states that Obama had double digits leads in February, those leads declined to the single digits in 2 states (CA, MI), and in one state Obama is now trailing McCain (OH)

    CLINTON       OBAMA    

CA      3%        -4%

IA     -1%        -2%

MA     -3%        -5%

MN     -3%        -1%

MO      5%        -2%

NM     -3%       -13%

OH      1%       -12%

OR      6%         1%

WI     -4%        -6%

AVG    0.1%      -4.9%

    ·    Clinton gained ground against McCain in 4 states by +1% (Ohio) +3% (California), +5%, and +6% (Oregon).  Obama gained ground against McCain in only one state (Oregon), and only by +1%.  

    ·    Clinton lost ground to McCain in 5 states by -1% (Iowa), -3% (Massachusetts, Minnesota, New Mexico), and -4% (Wisconsin).   Obama lost ground to McCain in 8 of 9 states, by -1% (Minnesota), -2% (Iowa, Missouri), -4% (California), -5% (Massachusetts), -6% (Wisconsin), -12% (Ohio), and -13% (New Mexico).

There is only one conclusion that can be reached from this data - Hillary Clinton is far less vulnerable to negative media messages than Barack Obama.  Voters seem to treat negative media attention about Clinton as ambient, albeit unpleasant, noise - it's just the same people saying the same things about Clinton that they've been saying for years.   While negative information about Clinton may be reinforcing impressions about her, it does not appear to be changing many people's minds.  Absent some "blockbuster" revelation about Clinton, her support appears firm, and because her negative public image stands in stark contrast to what people see when they focus on the "real" Hillary Clinton, she has strong potential to increase her support and improve her margins against McCain.

But Obama's support is "soft", and he is extremely vulnerable to negative media attention.  Voters do not really know who Barack Obama is, and what he really stands for and believes, and all the information about Obama is "new" to most voters.  Information that raises questions about his readiness for the Presidency, his character, and his political philosophy has had a major impact on people's views of Obama, especially when they are comparing Obama to McCain.   Negative information about Obama is likely to lead to continued deterioration of Obama's overall support and his margins against McCain.


While the purpose of this study is to compare shifts in support between late February and mid-April for Clinton and Obama when matched against McCain, it is important to keep in mind that, while Survey USA's methodology was the same, the population samples are not identical.    Differences in Party Identification and Ideology should have the most impact on the overall support and margins for Clinton and Obama when they are matched against McCain, rather than in changes in Clinton's and Obama's relative support and margins.   Nevertheless, changes in the sample may have some impact on how well Clinton and Obama do relative to each other among White voters when each is matched against McCain. (e.g. given the GOP "Southern Strategy", changes in the percentages of Republican voters could have an impact on the percentage of White voters who will not vote for Obama based solely on his race.)

Chart D-0 shows the changes in the relative percentages of Democrats and Republicans,  and Liberals and Conservatives,  in each of the 9 states (and on average.)   It should be noted that the chart describes the relationship between the parties/ and deologies, and that "More Democratic" can mean merely that fewer Republicans and more "Independents" were part of the April sample.

    ·    California - Less Republican/Conservative (more Independents and Moderates)

    ·    Iowa - More Republican and more Liberal (fewer Democrats and Independents, more Liberals)

    ·    Massachusetts - Less Republican, More Liberal (more Democrats, Independents, and Liberals)

    ·    Minnesota - More Democratic, less Conservative (fewer Republicans and Conservatives, more Democrats and Moderates)

    ·    Missouri - Less Republican, more Liberal (more Independents, fewer Moderates)

    ·    New Mexico - More Conservative (fewer Moderates)

    ·    Ohio - More Republican and Conservative (fewer Moderates)

    ·    Oregon - More Democratic and Liberal (fewer Moderates)

    ·    Wisconsin - Less Republican, more Liberal (more Independents and Moderates)

    ·    NINE STATE AVERAGES - Republicans -0.8%, Democrats +0.1%, Independents +1.4%, Conservative +0.6%, Liberals +3.0%, Moderates -1.3%


Overall Clinton outperformed Obama among White voters between late February and mid-April in the nine "Democratic should/can win" states surveyed.  Clinton gained support among White voters, while Obama lost support.  But both lost ground among White voters to McCain - with Obama losing far more ground than did Clinton.  

CHART D1 shows that:

    ·    Clinton gains White voter support in 5 states, but in only 2 (MO, +3%, OR +6%) is the increase in support at least +3%.  

    ·    Obama gains while voter support in 3 states, but only in Oregon (+5%) does he gain 3 points or more among White voters.

    ·    Clinton loses White voter support in 2 states, but only in New Mexico (-5%) is that loss more than -2%.

    ·    Obama loses White voter support in six states, including 4 where the loss is -3% or more (CA -9%, NM -12%, OH -8%, WI -4%)

CHART D-2 shows that:

    ·    In terms of gaining addition support from White voters during the six week period, Obama outperformed Clinton in only one of 9 states - Massachusetts, and the difference there was less than 3%.

    ·    Clinton outperformed Obama by 3% or more in terms of White voter support in 5 states: California (10 points better, Clinton +1%, Obama -9%), New Mexico (7 points better, Clinton -5%, Obama -12%), Ohio (10 points better, Clinton +2%, Obama -8%), and Wisconsin (4 points better, Clinton unchanged, Obama -4%).

    ·    Clinton had a nine state average gain in White voter support of +0.8%, while Obama had an average net loss of White voter support of -3.2%.

    ·    In late February, Obama had more support than Clinton among White voters in these states by an average of 1.9% (Clinton 45.2%, Obama 47.1%).  By mid-April, Clinton had 2.1% more support among White voters (Clinton 46.0%, Obama 43.9%).


Although Clinton gained and Obama lost White voter support between late February and mid-April, both lost ground with White voters when matched against McCain as previously undecided White voters made up their minds.  But Clinton's margin losses were far less severe than Obama's.

CHART D-3shows that:

    ·    Clinton lost ground against McCain among White voters in 7 of 9 states, and in 4 state the loss was -3% or more:  California (-5 points, from +3% to -2%), Massachusetts (-6 points, from +15% to +9%), Minnesota (-3 points, from +2% to -1%), and New Mexico (-12 points, from -14% to -26%).

    ·    Clinton gained ground among White voters in 2 states, Missouri (+4 points, from -7% to -3%) and Oregon (+9 points, from -4% to +5%).

CHART D-4 shows how:

    ·    Obama's margins among White voters improved in only one state, Oregon, by 6 points (from +3% to +9%).

    ·    Obama margin's among White voters declined by 3% or more in 6 of the 9 states: California (-21 points, from +10% to -11%),  Iowa (-4 points, from +10% to +6%), Minnesota (-3 points, from +5% to +2%), New Mexico (-27 points, from a tie to -27%), Ohio (-18 points, from +5% to -13%), and Wisconsin (-10 points, from +11% to +1%).

Clinton's 9 state average margins against McCain among White voters declined by -2.1 points (from -0.44% to -2.55%.

Obana's average margins decline much more severely, by -8.8 points (from +3.7% to -5.1%); Obama had been leading McCain overall among White voters in February in these 9 states, but was losing to McCain by mid-April.

    ·    Clinton's average margin loss (-2.1%) was less than the average reduction in undecided White voters(-3.2%) in Clinton v McCain.  

    ·    Obama's average margin loss (-8.8%) was more than three times the average reduction in undecided White voters (-2.3%) in Obama v McCain.

    ·    In late February, McCain lead Obama among White voters in only 1 of 9 states.  In mid-April, McCain beat Obama among White voters in 4 states.

    ·    In February, Obama's average margin among White voters was higher than Clinton's (Obama +3.7%, Clinton -0.4%).  By mid-April, Clinton was doing better than Obama among White voters (Obama -5.1%, Clinton -2.6%).


In the period between February 26-28 and April 11-13, the myth of Barack Obama hit the wall of reality.  The significant decline in White voter support is mirrored in his significant loss of overall support (see Part One), loss of support among both men and women (see Part Two), loss of support among Liberals and Moderates, and the extraordinary level of decline in the support of "Independents" (see Part Three).

During the same period, Hillary Clinton showed her staying power.  Her numbers among White voters can best be described as "mixed, but stable overall", reflecting the mixed results with varying demographic groups, and stability of her margins against McCain.

While these nine states cannot tell us what is going on in the other 41, they do represent states that Democrats need to win, or at least be competitive in, for the November election, and they are the kinds of states that Democratic Party super-delegates have to keep their eye on when deciding whom to support for the Democratic nomination.


Chart D-5 shows the "two-survey" average demographic breakdown for "White", "Black", "Hispanic", and "Other" voters.  Unfortunately, most of the nine states surveyed had relatively small percentages of non-White ("Black", "Hispanic", and "Other").  This results in completely unreliable data for non-White demographic groups in most cases (e.g. the survey data shows Obama going from having 73% of the "Black" vote in Iowa to only having 47% of the Black vote in mid-April.  This is highly unlikely, and probably the result of the extremely small sample of Black voters from Iowa.)

But there are sufficient non-White voters in two states in each category to do an "apples to apples" comparison.   But because of the relatively small sample sizes, these numbers cannot be considered authoritative.

CHART D-6 shows how:

    ·    Both Clinton and Obama increased their African American support in Missouri (Clinton +2%, Obama +3%) and Ohio (Clinton +9%, Obama +12%), with Obama showing greater gains in support in both states.

    ·    Both Clinton and Obama showed small losses (-1%) in support among Hispanics in New Mexico.  In California, Clinton shows a slight increase (+3%) in Hispanic support, while Obama saw a significant increase in support (+13%) among Hispanic voters.

    ·    In California, where the "Other" category consists mostly of Asian-Americans, Clinton gained support (+9%) while Obama lost support (-7%)  

    ·    In New Mexico, where the "Other category consists mostly of Indigenous/Native Americans, Clinton's support was unchanged, while Obama lost support (-9%)

CHART D-7 shows how:

    ·    Both Clinton and Obama saw slight decreases in their margins against McCain among African Americans in Missouri.  In Ohio, Clinton's margin increased by +7% against McCain, while Obama's margin improved even more (+22%).

    ·    Among Hispanic voters in California, Clinton saw an increase in her margin against McCain (+5%) while Obama's margin improved substantially (+22%).  In New Mexico, Both Clinton's (+1%) and Obama's (+2%) margin among Hispanics saw a small increase.

    ·    In California, Clinton saw a significant improvement (+29 points) in her "Other" (primarily Asian American) margins, going from -12% to +17%., Obama's advantage over McCain in this category decreased by -5 points (from +16% to +11%).  

    ·    In New Mexico, Clinton's margin in the "Other" (primarily Indigenous/Native Americans) category declined slightly (by -2 points, from -23% to -25%), while Obama's margin deteriorated significantly (by -26 points, from -2% to -28%.)

Data tables used for this post can be found at http://www.glcq.com/election08/apples/apples4_data.htm

< Do you know where your food has been? | Robo-Calls and reasonable basis' for voting decisions >
  • The Online Magazine with Liberal coverage of crime-related political and injustice news

  • Contribute To TalkLeft

  • Display: Sort:
    This is some serious work (5.00 / 2) (#1)
    by andgarden on Tue Apr 29, 2008 at 12:05:43 AM EST
    I'm sorry we aren't allowed to post images here, because your charts are illuminating.

    bright colors (5.00 / 4) (#2)
    by p lukasiak on Tue Apr 29, 2008 at 06:35:40 AM EST
    I'm sorry we aren't allowed to post images here, because your charts are illuminating.

    that's why I used such bright colors ;)

    Sorry about the formatting... every time I try to fix it, it just gets worse (like i tried to put a link in at the beginning to all the charts, and that got screwed up, and totally messed up the tables and line spacing because every time you update anything with unusual formatting, it adds extra paragraph breaks)


    thanks for the analysis (5.00 / 1) (#3)
    by Josey on Fri May 02, 2008 at 02:45:49 PM EST
    i read your article with great interest. (none / 0) (#4)
    by hellothere on Thu May 08, 2008 at 11:50:55 PM EST
    i agree as a layman. it is clear you did a lot of hard work and you should be proud of it.