A poll of Montana voters conducted jointly by Montana State University political scientists and the Montana Television Network indicates that the 2018 U.S. Senate race for Montana is a dead heat.

The poll showed incumbent Democrat Jon Tester holds a 3-point lead over GOP challenger Matt Rosendale – with 46. 2 percent of respondents favoring Tester, 43.1 percent supporting Rosendale, 2.6 preferring Libertarian candidate Rick Breckenridge and 6.5 percent respondents still undecided. Another 1.6 percent opted for an “other” category.

Those undecided voters could be key to the outcome of the highly contested race, said David Parker, associate professor of political science. Parker led a team of four MSU political scientists who mailed out ballots to 10,400 registered Montana voters in mid-September. About 20 percent of those who received the questionnaire — 2,057 respondents — sent it back by Oct. 6, which is considered a very good response rate, Parker said

“The key is how the independent vote will break,” said Parker, who added that the results are “a snapshot in time.”  He said that about 44 percent of the undecided voters identify as Republican, 27 percent as independent and 26 percent as another party. “Those voters could decide the election, as will turnout.”

Parker said there are three key takeaways from the data, which was weighted with U.S. Census Bureau statistics to reflect the key Montana demographics of age, education, gender by marital status and media markets. Results were also weighted to the 2016 presidential vote in Montana to reflect the state’s partisan dynamics, he said.

First, Democrats are nearly united in their support for Tester, with only 1 percent of the Democratic respondents reporting that they are undecided at this point. For Tester to win, Democratic voters will need to turn out at a rate similar to the 2006 or 2012 Senate elections, Parker said.

The second factor is undecided voters. Undecided voters are largely Republican -- 44 percent – and Parker said, given historical patterns, those voters could move to Rosendale’s column by Election Day, which could tighten the race if Democrats don’t turn out.

Parker said another related issue is whether undecided Republicans will vote with Tester, who is famously likeable, or remain loyal to President Donald Trump, who has made three personal visits to Montana to campaign against Tester.

“Rosendale has not established his personality in this race,” Parker said.  “Tester, on the other hand, is eminently likeable, even his opponents say that. Rosendale has focused his campaign on one thing: his loyalty to Trump. He has embraced that. The question is: Will Republicans vote for Rosendale not knowing him that well, or will they back him to support the president?”

The third factor involves independent voters. Independent voters favor Tester by nearly 30 points -- some 57.3 percent of independent voters support Tester, compared to 29.5 for Rosendale. Again, that makes turnout even more essential, Parker said. For Tester to win, his Democratic supporters and independents favoring him have to show up and vote, Parker said.

Other findings in the data include that women favor Tester by about 13 percentage points (51.7 percent to 38.6 percent) while men favor Rosendale by 6 points (47.5 percent to 41.6 percent). Rosendale’s support is concentrated among middle-aged voters, particularly between the ages of 40 and 49, while Tester has a clear lead among the youngest and oldest voters.

Parker said that several events linked to the election occurred during the time that the questionnaire was out, including the contentious hearings for U.S. Supreme Court Justice Brett Kavanaugh as well as a visit by Trump to Billings.

“But (election) events don’t have a long-term effect,” Parker said. “After a few days of a bump following the event, voting preferences return to their baseline condition.”

The questionnaire for the poll and data analysis was conducted by Parker and fellow MSU political scientists Eric D. Raile, Sara Guenther and Elizabeth Shanahan who also analyzed the data. The political scientists opted for a mail-in poll, which has a better response rate than telephone polls.

The MSU/MTN poll was divided into four sections. The Montana race for U.S. House will be announced tomorrow. Referendums will follow the next day. The week will conclude on issues and approval ratings. The team will also conduct a post-election poll in Montana.

For more information about the survey results, go to http://helpslab.montana.edu/