EAST LANSING, Michigan --- Former Secretary of State Hillary Clinton maintained her 3 percent lead last night according to the latest Fox 2 Detroit/Mitchell Poll of Michigan. In the four-way ballot question that includes Libertarian Party candidate former New Mexico Gov. Gary Johnson and Green Party candidate Dr. Jill Stein, it is Clinton 47% - Trump 44% - Johnson 4% - Stein 3% while 2% are undecided. Clinton has a 5 percent lead in the two-way race where she leads Trump 51%-46%. The percentages are almost exactly the same as they were the night before.
The IVR (automated) poll of 1,150 likely voters in the November 2016 General Election was conducted by Mitchell Research & Communications on Wednesday night, November 2, 2016 and has a Margin of Error of + or - 2.89% at the 95% level of confidence.
“Clinton’s weakness with 65 and older white women continued again last night and Trump maintained his small lead with men. Clinton did take back most of the support she lost with Democrats on Tuesday night. Although she stanched the bleeding, Clinton’s problems are taking a toll on her candidacy in Michigan and the state is now in play. That is likely the reason we have we have seen most of the Trump family including the candidate in Michigan along with former President Bill Clinton. Tomorrow, Hillary Clinton will be in town campaigning, more evidence of a close race,” Steve Mitchell, CEO of Mitchell Research & Communications said.
In terms of key demographics:
Clinton’s lead stayed identical to Tuesday night’s polling (50%-42%) while Trump’s 46%-44% lead dropped a point to 45%-44% last night.
Clinton rebounded with 18-44 year olds, moving from 45%-39% up to 51%-36% last night. Two percent fewer younger voters are voting for third party candidates compared to the night before. Johnson fell 1 percent to 5% and Stein dropped 1 percent to 6%. Clinton’s big lead with 65 + voters is at 25%, up from 15% Tuesday. Trump’s lead with 45-64 year old voters is the highest ever (55%-38%).
Three nights ago both Clinton and Trump were getting exactly 91.2% of their party’s vote in the four-way ballot test. Two nights ago, Clinton’s support with Democrats dropped to 86% while Trump’s stayed about the same (91.5%). Last night Trump dropped from 91% to 86% and Clinton rebounded to 90%. Trump continues to have a huge 4:1 lead (60%-16%) with ticket-splitters.
By race, Trump’s lead with white voters (48%-42%) is almost identical to his lead on Tuesday, while Clinton’s lead with African-Americans Tuesday (83%-14%) dropped last night (79%-19%). But, she expanded her lead with Hispanic/Asian/ and other ethnic groups to 62%-32% from 55%-38%.
By area, Clinton leads in the city of Detroit (85%-14%). Clinton has re-taken the lead in the Tri-County area (Wayne outside of Detroit/Oakland/Macomb) (48%-43%) and Trump leads outside the metro Detroit area (48%-43%).
Other key findings:
• Clinton’ favorable moved up 3 percent and her unfavorable dropped below 50% (43% Favorable-49% Unfavorable). Trump’s stayed about the same as the night before (40%-56%).
We also tested to determine the awareness of 17% premium increases in the individual market at Healthcare.gov. Only half the voters were aware, about a third of Clinton voters and two-thirds of Trump voters. When told the statement on the premium increases was true and after asking if that information would make them more or less likely to vote for Trump, the answers were partisan, with Trump’s supporters saying it made them more likely to vote for him and Clinton’s voters saying they were less likely.
“It is easy to get wrapped up in changes within demographic groups, but the bottom line is the race solidified at 3% last night. Now, we will have to see if Trump will be able to gain enough momentum to take the lead, or whether Clinton will be able to regain her momentum,” Mitchell concluded.
Mitchell Research & Communications used a sample of registered voters in the November 2016 Michigan General Election. Our goal is to spread as wide a net as possible to assure we survey voters who may not have voted in elections for a long time. A double filter was used to determine that we were surveying only likely voters.
• First voters had to say they were registered voters. If they were not, the phone call ended.
• Then, they were asked if they were definitely voting, probably voting, not sure yet, or definitely not voting in the November General Election, or if they already voted by absentee ballot. If they were “definitely not voting” the phone call ended.
Federal law only permits us to call land lines when using automated phoning. Any surveys that doesn’t specify if they use cell phones do not use them. Data is weighted to reflect likely voter turnout by gender, age, and race.
(Steve Mitchell is CEO of Mitchell Research & Communications, an East Lansing, Michigan based national polling and consulting company. He is currently political pollster for FOX 2 Detroit. He has polled with great accuracy for the media in Michigan, South Carolina, Florida, Missouri, Illinois, and California.
An examination of RealClearPolitics.com will show that in final polls before the election in races for president, U.S. Senator and governor in Michigan, his final poll results have been off by an average of only 2.75% from actual results 2008-2014.
Mitchell can be reached at 248-891-2414; firstname.lastname@example.org; @stevemitchell40 on Twitter)