A new poll came out today showing Hillary Clinton and John Edwards way out in front of everyone else in Iowa.
Here are the results from the American Research Group (May's support in parenthesis)…
Clinton 32% (31%)
Edwards 29% (25%)
Obama 13% (11%)
Richardson 5% (8%)
Dodd 2% (2%)
Biden 2% (3%)
Kucinich 1% (4%)
Gravel 1% (1%)
Clark 1% (1%)
Undecided 14% (14%)
At first, I was surprised Hillary was leading with over 30% when other polls show her around 20%. However, if you look at the trends, Hillary has consistently been in the low 30's in this poll, except for May when she had just 23%. In fact Hillary is basically right where she was back in December before her campaign even started (32% to 31%).
Ezra Klein seems to think John Edwards' support is declining, but when you look at the trends in the American Research Group polls, it shows Edwards gaining support. In the past month he increased by 4%. Overall, he has improved from 20% back in December to 29% in June. 9% is the largest increase from any candidate.
If you look at Obama's trends in the American Research Group polls, he peaks at 23% in February right after he announced. He has been on a downward trend since then. I think Obama's support is stronger than the 13% he is polling because the polls is of likely caucus goers. Obama has made a big splash with younger people who would not be included in this poll.
Another interesting thing about this latest poll is looking at the preference by party. Among those who identified themselves as no party, 39% were in favor of Edwards, 31% in favor of Clinton, 11% in favor of Obama, and 3% in favor of Richardson. Edwards' support here only strengthens his campaigns claim that Edwards is in the best position to win independent voters and in red states.
Here's analysis from past ARG polls in May and March .
Originally posted at Century of the Common Iowan.
2 Comments
I think something is wrong with their screen
They have consistently had Hillary way higher than anyone else. At least two other Iowa polls have shown Hillary below 20 percent, but ARG almost always has her over 30.
I would like to know more details about their likely voter screen. We know that at most 15 to 20 percent of registered Ds will come out on caucus night. How many other registered Ds or registered voters is ARG polling?
There's no way I believe Obama is that far back in Iowa either, by the way.
desmoinesdem Mon 2 Jul 5:10 PM
ARG sucks
Best way to deal with ARG polling is to treat them with a grain of salt. DMD is also right about their screen–there is something seriously wrong with it. The fact that they include “No Party” registrants in the sample is wrong, considering that for Iowa’s caucuses you have to be a registered member of a party to participate. While having “no party” respondents included in a general sample is fine, having it in a caucus poll screws up the screen because the “no party” title makes it less likely for the person to attend the caucus since they have more work to do simply in order to be able to participate.
Oh, and if you need some expert opinion, here are Drew’s thoughts from January about ARG’s polling.
chris-woods Tue 3 Jul 3:28 PM