| This morning Newsweek released the results of a new poll of 1215 registered Iowa voters, including subgroups of both likely Democratic and likely Republican caucus goers.
Obama leads just within the margin of error among likely Democratic caucus-goers, although the margin of error is pretty huge (+/-7):
Obama 28, Clinton 24, Edwards 22, Richardson 10
All other candidates are in the single digits. So it's still a close race, and overall it's probably best to think of this as a three-way tie. Still, it's hard not to see this as bad news for Edwards, if only because his strategy depends so heavily on winning in Iowa. I don't think he could survive a third place finish there. Relative to the Pollster averages, this poll has Edwards performing the same as he normally he does (22%), Clinton and Richardson are performing similarly (26% to 24% and 10% to 13%), but Obama is performing significantly better (28% versus 20%). Obviously, it's also good news for Obama.
Voters who indicated they would either definitely or probably attend the caucus (42% of those surveyed) were included as likely caucus goers - 64% of those identified as likely caucus goers attended a previous caucus.
This poll also provides some indirect support for Chris's inflated Clinton poll theory - Clinton leads among registered Democratic voters (MOE +/- 5):
RDV: Clinton 31, Obama 25, Edwards 21, Richardson 6
Obama and Richardson both do better among likely caucus goers than registered Democrats, while Clinton does worse and Edwards does basically the same. Still, Clinton has the highest fraction of supporters who say they are strongly supporting her - 55%, compared to 41% for Obama and 37% for Edwards. Another interesting result from this poll - 80% of those surveyed thought it would be a plus to have Bill Clinton "back in the White House."
Here are the results for the Republican field:
LCG: Romney 24, Thompson 16, Giuliani 13, Huckabee 12
RRV: Romney 25, Thompson 16, Giuliani 15, Huckabee 6
I have to say I'm not excited to see Huckabee break into the double digits, but I am pleased to see Romney continuing to do so well. McCain is in the high single digits - 9 among likely caucus goers and 7 among registered voters.
Last, this seems as good a time as any to mention the Pollster disclosure project. Pollster is one of my absolute favorite blogs - I can't tell you how much I rely on them for understanding trends and using polling data to predict election results. They are now making a formal attempt to encourage ALL pollsters to release more information about likely voter screens, sampling frames, and sample compositions. Having this information available would make a huge difference in my ability to tell you what might really be happening when polls produce contradictory results or when I want to write one of the longer in-depth posts about the race that I love so much. |