Pollster.com

Articles and Analysis

 

POLL: SurveyUSA Missouri


SurveyUSA

Missouri
McCain 50, Obama 43

 

Comments
sherman:

First good news for McCain in a while. He needs Missouri if he wants any kind of chance at winning.

____________________

Undecided:

Senator McCaskill from MISSOURI as VP does nothing to help Obama, even against Fiorina.

The previous SurveyUSA poll (5/16-18/2008) was McCain-48 and Obama-45 with 7 undecided. So this is a 2 point gain for McCain and 2 point loss for Obama.

____________________

Andrew_in_California:

I'm very interested to know how responsive people are to VP picks in polls historically but otherwise good news for McCain.

____________________

brambster:

I can't claim to know what numbers are the most accurate between different surveys from the same pollster one month apart, but there are suttle changes in this survey from the SurveyUSA one on May 21st.

http://www.surveyusa.com/client/PollReport.aspx?g=56a201bd-db14-4567-bba2-dffea45ee60e

The most striking is the party ID, which saw a 3 point gain for R's, a 1 point loss for D's, and a 2 point loss for I's. That's enough to swing 3 point difference into a 7 point difference, in fact it explains the difference perfectly.

This wasn't just a party ID movement among voters, it was either a different sample, or different weighting. You can check both polls for their stance on abortion, and this swung from a 3 point lead for pro-choice, to dead even with just 3% and 2% undecided/refused. This isn't an issue that will change much in one month without something newsworthy.

I have looked for regional differences between the two and it pretty much looks like Kansas City. On May 21st, Obama was winning that region by 14 points, but the one today shows him losing that same region by 5 points. A 19 point swing in one region that is also typically favorable to Democrats. Maybe there's something local in KC that would influence just them, but the rest of the state was fairly static.

I also looked back in older polls that also had Clinton in them, and the pro-choice/pro-life margin can swing almost 20 points. That doesn't make much sense to me at all unless your samples are too small. Of course none of us trust a single poll anymore with the invent of trends and averages of polls.

Lastly I compared the breakdowns to 2004's exit polls and while there is a very slight shift in party ID that is favorable to Obama since then, I would guess that this state is much closer to even right now. Unlike 2004, the governors race looks to heavily favor the Democrat, and there is no Senate race so this gives additional advantage over 2004 where there was a 1 point margin on the gubernatorial race and a 13 point win by an incumbent Republican Senator.

____________________

1magine:

Never saw Obama winning this state. Maybe he does. But Looking in June/ July just seems too early.

____________________

brambster:

@1maginie

I have this on my list as the 7th "Bush state" to fall to Obama (that is if it falls). I see Iowa, Colorado, New Medico, Virginia, Ohio and Nevada falling before Missouri does. Obama would need a 3% to 5% national vote lead before this happens. If Obama gets a 3% margin nationally, his election is certain regardless of what happens in Missouri.

I think that Virginia is much more favorable to Obama because it's not as starkly demographically polarized as Missouri is, and therefore probably less politically polarized or possesses less political inertia than a state like Missouri that has been fought over in every election since 2000.

____________________

Andrew_in_California:

I would say that SurveyUSA's black population projection is smaller than the actual Democratic Primary which was around 18%. If a 6% difference means anything.

____________________

brambster:

@Andrew

Don't mistake the breakdowns in a Democratic primary from those of the general election. 2004's exit polls showed 8% AA overall.

http://www.cnn.com/ELECTION/2004/pages/results/states/MO/P/00/epolls.0.html

____________________

Andrew_in_California:

That's a good point, but after this election is all said and done it will be interesting to see how much larger the African American population number projections change from 2004. SurveyUSA's 4% increase from 2004 should at least allow us to question its appropriateness.

____________________

Undecided:

U.S. Census Bureau Quickfacts for Missouri:
Total population 5,842,713
Population change 2000-2006 4.4%
Black persons 11.5%

More info from Fair Vote 2020:
No. African Americans registered to vote 399,000
No. African American citizens 18+years 455,000
Percentage of African Americans registered to vote 87.7%

Missouri Secretary of State:
Total registered voters 4,007,174

Calculated AA registered voters = 10%
Calculated AA citizens 18+ years = 11.4%

Survey USA included 12% AA voters, which is in line with the U.S. census population growth info. In fact, it is generous considering that the total number of eligible AA voters of 11.4%.


____________________

kingsbridge77:

I would say that SurveyUSA's black population projection is smaller than the actual Democratic Primary which was around 18%.

Blacks represent a smaller share of GE voters. Primaries are differents. Blacks are a much Let's not compare apples and oranges.

____________________

avatarMM:

survey USA is very innacurate because they poll a very small sample.

Obama losing Kansas City? lol.. not gonna happen.

also like someone above said about the pro-choice thing.. i dont think in 1 month people changed their position on that..

their polls seem to be all over the place.. i actually respect Rasmussen more, although even they under-do dem ID.

____________________

avatarMM:

the % of black vote in 2006 was 13%.. this poll has them at 12%.

you can expect it to be at 16-18% this time around.

also they have McCain WINNING young voters which will not happen in any swing state.

survey USA polls are fun cause they put out polls all the time.. but they just aren't accurate

____________________

Undecided:

I surmise that your 13% black vote was from exit polls for the 2006 Senate Race (McCaskill).

The SurveyUSA was accurate enough for a poll. They used a proportion of black voters roughly equivalent to the breakdown of those old enough to vote. (See my post above of population and voter data.) Remember that poll samples intend to reflect the overall population. In the case of Missouri, blacks make up just under 12% (whether total population or population of those eligible to vote).

The only way to INCREASE the black vote to such larger margins (16-18%) is to get the non-blacks to stay home on election day.

____________________

Undecided:

Regarding the USASurvey pro-life/pro-choice beliefs (49-49)-- again within the range expected for a Missouri polling sample.

Note that in the 2006 Missouri senate race exit poll that 52% thought that abortion should be illegal.


____________________

1magine:

I do not see MO changing hands in this election. I think SUSA is relatively accurate.

Personally I am more concerned about not seeing regular polls in FL & OH. To be as cynical as possible - Obama does not need VA, MO, Alaska, GA or FL - this election I think either comes down to OH - OH- OH or its a 1964 type landslide. But I believe OH is the best bell weather of the country's electoral direction. Big cities, small towns, lots of suburbs & a population (ethnic/religous) breakdown that mirrors the entire nation. Keep an eye on OH polls through Aug. If they stay steady into September, with a 5-7 point Obama lead then its an electoral blowout, if it stays close w/o Obama never losing, then he edges McCain, if he loses OH - McCain is in the driver's seat.

____________________

Andrew_in_California:

First you are comparing 2000 Census data to 2008. Much has changed (I worked on the Census in my own County and found profound changes). I don't know how they changed for Missouri but there is probably decent evidence of many AA were displaced in Katrina are probably residing in Missouri.

Second I think its far better to compare the Democratic AA turnout than to the 2004 GE because there is some representation of how AA will be turning out for Barak Obama when white people couldn't even vote for Kerry. :p

____________________

onelightonvoice:

1magine -

Agree about florida, disagree about ohio. I see a blue line, from Maine to Florida. if the line doesn't continue to florida, then it will be a bit closer, but still an obama victory. mccain will still lose even if he eeks out ohio. If obama should win va, nc, sc, and ga, then ohio is irrelevant. if he even wins one of those states, ohio is a non-issue, due to the fact he is winning others out west (see below). I don't really believe this MO poll. It will be close there, but MO will also be a non-issue. It all begins and ends out east.

***************NEWSFLASH******************


New polls from quinnipiac-

http://firstread.msnbc.msn.com/archive/2008/06/26/1169534.aspx

____________________

Undecided:

If someone was referring to my posts concerning Census Data as being old... Please note that the U.S. Census Bureau updates it numbers using other data to make an estimate, albeit not an official census, the numbers are in the ballpark. This estimate was made for 2006 which uses data AFTER Aug 2005 (Hurricane Katrina).

Next... two other data sources were provided that use the LATEST voter registration numbers. All three sets of data are consistent with one another.


____________________



Post a comment




Please be patient while your comment posts - sometimes it takes a minute or two. To check your comment, please wait 60 seconds and click your browser's refresh button. Note that comments with three or more hyperlinks will be held for approval.

MAP - US, AL, AK, AZ, AR, CA, CO, CT, DE, FL, GA, HI, ID, IL, IN, IA, KS, KY, LA, ME, MD, MA, MI, MN, MS, MO, MT, NE, NV, NH, NJ, NM, NY, NC, ND, OH, OK, OR, PA, RI, SC, SD, TN, TX, UT, VT, VA, WA, WV, WI, WY, PR