Lake Research Partners, a Democratic firm, and American Viewpoint, Inc., a Republican firm, conducted voter surveys in seven states where five female and two male governors were running for re-election. The research measured voter attitudes and perceptions of the governors focusing on their leadership skills, issue identification and job performance. The measurements for the women governors were then compared to those for the men to assess the impact of gender on voters’ views.
The survey was conducted by phone using professional interviewers. The survey reached 2,734 likely registered voters ages 18 or older in seven states, including 408 in Arizona, 400 in Connecticut, 404 in Hawaii, 404 in Kansas, 400 in Michigan, 365 in Rhode Island and 353 in Wisconsin. The survey was conducted between October 30th and November 2nd, 2006.
Telephone numbers for the survey were drawn from voter files. The sample was stratified geographically in each state based on the proportion of voters in each region. Data was weighted slightly by gender, age, education, party identification, union membership, employment status and race to reflect the attributes of the actual population of registered likely voters.
In interpreting survey results, all surveys are subject to possible sampling error, among other kinds of unmeasurable response error; that is, the results of a survey may differ from those which would be obtained if the entire population were interviewed. The size of the sampling error depends upon both the total number of respondents to a given question and the percentage distribution of responses to that question. The margin of error for the total sample is +/-1.9%. For Arizona, Connecticut, Hawaii, Kansas and Michigan the margin of error is +/-4.9%; for Rhode Island and Wisconsin it is +/-5.2%. The margin of error on split sampled questions and among voter subgroups is higher because the sample sizes of these groups are smaller.
Evaluation of 2006 Exit Poll Data
Lake Research Partners performed an analysis of the 2006 state exit polls developed and conducted by Edison Media Research and Mitofsky International on behalf of the National Election Pool. NEP is comprised of ABC News, Associated Press, CBS News, CNN, Fox News and NBC News. The exit poll results are based on interviews with a probability sample of voters exiting polling places in each state on election day. The error due to sampling for most state exit poll questions is approximately +/-4%. The New Mexico dataset was weighted slightly to reflect the gubernatorial vote results in that state.1
In 2006, 36 states held elections for governor; exit poll data was available in 24 of these states. Our analysis focused on these 24 states. Nine of these states had at least one woman candidate for governor (Arizona, Connecticut, Hawaii, Illinois, Maine, Massachusetts, Michigan, Nevada and Texas). The other 15 states had male gubernatorial candidates only (California, Florida, Georgia, Maryland, Minnesota, Nebraska, New Mexico, New York, Ohio, Pennsylvania, Rhode Island, Tennessee, Wisconsin, Wyoming and Vermont).
Campaign Tracking and Interviews
Beginning in the spring of 2006, Staton Hughes led a bipartisan team of researchers who tracked the 11 gubernatorial contests in which women were nominees of the major political parties or competitive Independent candidates. Researchers followed print and Internet coverage, and collected campaign commercials and debate programs. Following the November election, these researchers interviewed 55 participants in those campaigns, including candidates, campaign managers, finance directors, press secretaries, consultants and reporters who covered the races.
The Democratic candidates for governor included: Lucy Baxley (AL), Jennifer Granholm (MI), Janet Napolitano (AZ), Kathleen Sebelius (KS) and Dina Titus (NV). The Republican candidates for governor included: Kerry Healey (MA), Linda Lingle (HI), Sarah Palin (AK), M. Jodi Rell (CT) and Judy Baar Topinka (IL).
1For more information on the 2006 exit poll methodology, please refer to: The Roper Center for Public Opinion Research