A CBS/New York Times poll released yesterday indicates Newt Gingrich has gained a substantial lead among likely Republican Iowa Caucus voters. Thirty-one percent of respondents stated they would support Gingrich, putting him at a 14 point lead ahead of Mitt Romney, who is only one point ahead of Ron Paul. Part of the reason for Gingrich’s newly emerging lead may be attributed to increasing support from Republicans who identify with the Tea Party. Of likely Republican Caucus go-ers polled who stated they support the Tea Party, 41 percent selected Gingrich as their preferred candidate.
Tepid Tea Party Support?
Republicans who associate with the Tea Party originally mostly supported Michele Bachmann, founder of the Tea Party Caucus in the U.S House of Representatives, propelling her to an Iowa straw poll win in August. Concerns about Bachmann’s electability caused Tea Party support to shift to Rick Perry, but his poll numbers suffered a significant drop after a series of gaffes during televised debates. Herman Cain quickly became the new Tea Party favorite, and he surged in the polls until several allegations of sexual harassment and extra-marital affairs caused many Republicans to withdraw their support. At the time the CBS/NYT poll was conducted, Cain only had seven percent support among likely Iowa Caucus participants. He suspended his campaign on December 3, about mid-way through the polling period. While a number of his former supporters in Iowa apparently switched their support in favor of Bachmann, causing her standing in the CBS/NYT to increase to nine percent from four percent, this poll indicates that Iowa Tea Party supporters currently favor Gingrich by a 30 point margin over Bachmann. Gingrich is experiencing similar favorability among Tea Party supporters nationwide, as a Gallup poll states 47 percent of Republican voters who say they support the Tea Party now support Gingrich, a number that is up from 18 percent last month.
CBS/NYT Poll Methodology
The poll was conducted between November 30 and December 5 via directory-listed landline contacts and random dialing of cell phones. The sample consisted of 642 registered Iowa Republicans who stated they planned to participate in the Caucuses. Weighting was used to account for selection probability and demographics. The margin of error was +/-4 points.
Possible CBS/NYT Poll Flaws
Methodology released by CBS news was somewhat vague, and it is not known whether respondents were contacted by live or automated interviewers. Since the survey itself was long many respondents may have hung up if contacted by a recording. However, more in-depth analysis reported on previous polls conducted by the New York Times indicates that live interviews were conducted, so it is probable that their methodology remains consistent.
CBS/NYT defined “likely Republican Caucus participant” as a respondent who was both a registered voter and indicated he or she “definitely” or “probably” would attend the Caucuses. The poll questions reveal that this sample includes only 31 percent of those initially contacted, and only 49 percent of the 642 Republicans who expressed their desire to vote on January 3 stated they would definitely participate. Since actual turnout is almost always less than polled expectations, some pollsters use special formulas designed to reduce the sample in a way that may imitate realistic Caucus attendance. However, these formulas are often based off exit polls from previous elections, which may not accurately forecast turnout in a new election being conducted during a different political climate.
Additionally, the CBS/NYT poll respondents who indicated they will attend the Caucuses included 34 percent who had never previously attended the Caucuses. This represents a significant portion of the small sample without a past history of participation, who may be less likely to actually turn out.
Who is the Tea Party?
Respondents were asked whether they supported the Tea Party in order to gather a separate sample of Republican Caucus go-ers who were Tea Party supporters. The polling data does not reveal the size of this sample. Party affiliation is generally considered an unreliable sampling variable, since it is fluid. The fact that this poll was conducted immediately before, on, and after December 3, the day Herman Cain suspended his campaign, may have caused a shakeup among Republicans, making their stated support or non-support for the Tea Party less reliable.
Despite the possible flaws, the results for the CBS/NYT Iowa poll follow the same trend as other Iowa polls, when the sample of likely Republican Caucus-goers is considered. The RealClearPolitics average has Gingrich with a 12.4 percent lead over Romney, while the CBS/NYT poll has Gingrich ahead of Romney by 13 points.
Is the Shift Significant?
According the the CBS/NYT poll results, Tea Party supporters are not only looking for a candidate who represents their values, electability is also important. Twenty-four percent of Tea Party supporting Caucus-goers selected Michele Bachmann as the candidate most aligned with their values, and 12 percent selected Gingrich. However, 45 percent of the same sample selected Gingrich as the candidate they believed would be most likely to beat President Obama in the general election. Many who will vote in the Iowa Caucuses have not made their final decision, and 66 percent of respondents surveyed by CBS/NYT stated their minds may change before January 3.
CBS News. CBS/NYT Poll on Iowa: 12/6/2011. Accessed December 9, 2011.
CBS News. The Iowa Caucuses: Key Voter Groups and Issues. Accessed December 9, 2011.
Jan, Tracy. Tea Party Members Get Behind Gingrich. Boston Globe. Accessed December 9, 2011.
RealClearPolitics. Iowa Republican Caucus. Accessed December 9, 2011.
Decoding Science. One article at a time.