UNH Survey Center

UNH Poll Finds Sununu and Smith in Close Race for U.S. Senate

By Erika Mantz
UNH News Bureau

July 2, 2002

DURHAM, N.H. Republican incumbent Bob Smith has narrowed the gap with Congressman John Sununu in the GOP primary race, and hypothetical races between both candidates and Democratic Governor Jeanne Shaheen are close, according to the latest Granite State Poll conducted by the University of New Hampshire Survey Center.

The Granite State Poll is sponsored by UNH. Six hundred fifty-eight (658) New Hampshire adults were interviewed between June 23 and July 1, 2002 (margin of sampling error, +/-4 percent). Included in this sample were 529 likely voters in the 2002 general election (margin of sampling error, +/-4 percent), and 230 likely GOP primary voters (margin of sampling error, +/-6 percent). For more detailed results and methodology, visit the Survey Center Web site at www.unh.edu/survey-center and click on Press Releases.


As the New Hampshire weather heats up, so, too, does the race for New Hampshire's U.S. Senate seat. Two-term incumbent Republican Bob Smith of Wolfeboro is being challenged by fellow Republican John Sununu of Bedford, a three-term representative from New Hampshire's First District. Sununu continues to enjoy a lead over Smith, although the margin has shrunk over recent months. In the most recent Granite State Poll, 52 percent of likely GOP primary voters said they plan to vote for Sununu, 40 percent plan to vote for Smith, 2 percent favor Scot Stremsky, and 6 percent are undecided. Smith has cut into Sununu's lead since the Spring 2002 Granite State Poll in which 59 percent of likely GOP primary voters said they planned to vote for Sununu and only 30 percent said they planned to vote for Smith. "This race is by no means over," said Andrew Smith, director of the UNH Survey Center. "Smith's television ads appear to have had an impact on Republicans. While Sununu still has a lead, and has also started running television spots himself, Smith has more money on hand for advertising before the primary."

Three-term Governor Jeanne Shaheen of Madbury is unchallenged for the Democratic nomination and will present either Smith or Sununu with a significant challenge in November. Sununu continues to run stronger than Smith against Shaheen. Currently, 51 percent of likely November election voters favor Sununu, 42 percent favor Shaheen, 3 percent favor some other candidate, and 4 percent are undecided. Sununu's lead has stayed relatively unchanged since the Fall 2001 Granite State Poll. In a hypothetical race between Smith and Shaheen in November, Shaheen is leading Smith among likely voters 47 percent to 45 percent, with 3 percent favoring some other candidate, and 5 percent undecided. This race also has remained quite stable since October. "It should be no surprise that the November matchups have remained essentially unchanged," Andrew Smith said. "Shaheen, Smith and Sununu are all well known candidates and there are few voters who are undecided."


Sununu is leading in the race for U.S. Senate, in part, because he has maintained the highest favorability ratings of the three Senate candidates. Among likely November voters, 56 percent have a favorable opinion of Sununu, 27 percent have an unfavorable opinion, 8 percent are neutral, and 8 percent don't know enough about him to say. Sununu's net favorability rating, the percentage having a favorable opinion of him minus those having an unfavorable opinion, is +29 percent. Smith's favorability ratings are considerably lower than Sununu's with 43 percent of likely November voters saying they have a favorable opinion of him, 38 percent having an unfavorable opinion, 11 percent neutral, and 9 percent not knowing enough about him to say. Smith's net favorability rating stands at +5 percent. Shaheen's favorability ratings are approximately the same as Sununu's with 56 percent having a favorable opinion, 31 percent having an unfavorable opinion, 12 percent neutral, and 4 percent don't know enough to say. Shaheen's net favorability rating is + 25 percent.

Further evidence of the importance of general favorability can be seen in the reasons voters give for why they support a candidate. Sununu voters say the most important reasons they plan to vote for him in the primary are that they like him better than the other candidates (41%), that they generally agree with him on the issues (25%), that he is experienced (8%), and that Smith left the GOP in 2000 (6%).

Similarly, Smith supporters say the most important reason they plan to vote for him is that they like him better than the other candidates (32%), that they generally agree with him on issues (23%), that he is experienced (13%), they support his environmental positions (5%), and that he has done a good job as senator (5%).


Democrat Martha Fuller Clark is facing only nominal opposition in the Democratic primary for John Sununu's First Congressional District seat and should easily win the nomination. However, the Republican primary is up in the air with eight candidates vying for the nomination. While she does not face any well known Republicans, Clark will have to work hard to win seat, which is traditionally Republican. When asked who they would vote for Congress, residents of the district who are likely to vote in the November election are evenly split between Clark and a yet-be-named Republican. Clark is favored by 37 percent of likely district voters, 41 percent favor the Republican candidate, 4 percent say it depends on who the Republican is, 2 percent favor some other candidate, and 16 percent are undecided.


In the race for U.S. Congress from the Second District, incumbent Republican Charlie Bass holds a comfortable lead over his primary Democratic challenger, Katrina Swett. Currently, 58 percent of likely voters say they will vote for Bass, 23 percent favor Swett, 2 percent favor some other candidate, and 8 percent are undecided.


Sununu continues to hold his own against Smith among the Republican party base while attracting moderates and independent (undeclared) voters. Among registered Republicans, Sununu leads Smith by 52 percent to 42 percent while he leads 59 percent to 33 percent among independent voters who say they will vote in the GOP primary. Similarly, Sununu and Smith are tied among self-described conservatives while Sununu holds a 61 percent to 30 percent lead among self-described moderates.

In trial heat match-ups versus Shaheen, Sununu also does better than Smith among registered Republicans (85% to 76%), conservatives (78% to 75%), self-identified Republicans (89% to 78%) and among undeclared voters (48% to 46%). Shaheen gets her strongest support from Democrats and liberals.

There is a gender gap developing in both the September GOP primary and the November general election. While GOP men are evenly split between Smith and Sununu (47% to 46%), GOP women favor Sununu by a wide margin - 60 percent to 31 percent. Women should be a strong source of support for Shaheen, New Hampshire's first woman governor. But Shaheen only holds a 49 percent to 46 percent lead over Sununu among women who plan to vote in November Sununu leads Shaheen among men by 56 percent to 34 percent.


These findings are based on the most recent Granite State Poll conducted by the UNH Survey Center from June 23-July 1, 2002. A random sample of 658 New Hampshire adults was interviewed by telephone. In 95 of 100 cases, the statewide estimates will be accurate to plus or minus 3.8 percent. Results reported for other subgroups have potential for somewhat larger variation than those for the entire population. The margin of sampling error for likely November voters (N=529) is +/-4 percent. The margin of sampling error for likely GOP primary voters (N=230) is +/-6 percent.

The data have been weighted to adjust for numbers of adults and telephone lines within households, respondent sex, region of the state and likelihood of voting in either the Sept. 10, primary or the November general election. In addition to potential sampling error, all surveys have other potential sources of nonsampling error including question order effects, question wording effects, and nonresponse.

Back to UNH News Bureau