Michigan Senate Race Tight as Voters Divided on Candidates, Poll Shows
Abdul El-Sayed holds a slim 1-point lead over Mike Rogers in a U.S. Senate contest marked by voter concerns about both contenders.
Abdul El-Sayed has a narrow 50-49% lead over Republican Mike Rogers among likely voters in Michigan's U.S. Senate race, according to a Fox News poll released Wednesday. This marks the first Fox News Poll of likely voters this season, preventing direct comparison with previous surveys.
Among registered voters, El-Sayed leads 51% to 48%. This represents a shift from August, when Rogers held a 4-point advantage among registered voters.
The tight contest is attributed to voter concerns about both candidates. Approximately equal numbers of Michigan residents express significant worry that Rogers is too closely aligned with former President Donald Trump (39%) as express concern that El-Sayed's views are too liberal (36%).
Favorable ratings for the candidates are also close, with El-Sayed viewed positively by 47% of respondents and Rogers by 44%. Since August, positive perceptions of El-Sayed have increased by 5 points, while Rogers' favorability has decreased by 4 points.
El-Sayed draws strong support from demographic groups typically favoring Democrats, including women, Black voters, those with college degrees, voters under 45, and union households. Rogers' core support comes from men, rural voters, White evangelical Christians, married voters, and White individuals without a college degree.
Both candidates command strong party loyalty, with Rogers securing 95% of Republican support and El-Sayed receiving 93% from Democrats. Rogers also garners 99% of MAGA Republicans and 88% of non-MAGA Republicans, while El-Sayed has 98% of DSA Democrats and 90% of non-DSA Democrats. Independents narrowly favor El-Sayed.
The broader political environment favors El-Sayed, as Democrats hold a 5-point advantage in the preference for Senate control (52% to 47%). However, El-Sayed faces the challenge of consolidating support after a contentious primary, with 11% of likely voters who supported his primary opponent, Haley Stevens, indicating they plan to vote for Rogers.
"The Michigan Senate race is unique in that short-term political forces favoring the Democrats are somewhat offset by concerns about the Democratic candidate," said Republican pollster Daron Shaw, who conducts the Fox News Poll with Democrat Chris Anderson. "The bad news for Rogers is El-Sayed has made progress shoring up Democratic support and has room to grow given voter concerns about the economy."
Financial concerns significantly influence voter preference. Four in ten Michiganders who feel they are falling behind financially back El-Sayed by 25 points. Conversely, those who describe their financial situation as holding steady, nearly half of respondents, favor Rogers by 13 points, while a smaller group experiencing financial gains prefers Rogers by 36 points.
The cost of living is a critical issue for 70% of voters, who favor El-Sayed by 13 points. El-Sayed also holds advantages among voters focused on U.S. policy toward Israel (8 points) and artificial intelligence technology (17 points). Immigration is a key issue for 44% of respondents, who favor Rogers by 26 points.
Donald Trump's importance to the Senate vote is considered extremely high by half of Michigan voters, who favor El-Sayed by 22 points. In 2024, Trump won Michigan by less than two percentage points, and Rogers lost to Democrat Elissa Slotkin by less than half a point.
Disapproval of the current administration's performance, cited by nearly 60% of voters, poses a challenge for Rogers. While 40% approve of the president's performance, not all of them support Rogers, who still garners support from 19% of those who disapprove of Trump. However, 28% of those who only "somewhat" disapprove of Trump favor El-Sayed.
In the gubernatorial race, Democrat Jocelyn Benson leads Republican John James 54% to 45% among likely voters.
The poll was conducted September 24-28, 2026, surveying 1,203 Michigan registered voters. The margin of sampling error for the registered voter sample is ±2.5 percentage points, and for the likely voter subsample of 1,028 respondents, it is ±3 percentage points.