Democrats Hold Lead on Statewide Generic Congressional Ballot
Democrat Abdul El-Sayed leads Republican Mike Rogers in the race for U.S. Senate in Michigan. The Wolverine State Poll was conducted by BIG DATA POLL as part of the Public Polling Project for September 2026.
Dr. El-Sayed leads former Representative Rogers 44.3% to 39.0% among registered voters and 45.9% to 40.5% among likely voters on the initial ballot. When undecided voters are asked to lean, the Democrat’s 5-point lead among likely voters holds on the margin to 46.7% to 42.1%. Undecided likely voters break for Rep. Rogers, though not enough to close the gap.
“The undecided voters in the race for U.S. Senate in Michigan remind me somewhat of the Virginia election in 2025” Director BIG DATA POLL Rich Baris noted in a statement. “A not-so insignificant share of undecided Democrats are choosing someone else when leaned over a controversial nominee, though both candidate’s appear to have equal support among the rest of their base.”
“In the end, they voted for the Democratic candidate.”
Generic Congressional Ballot
Democrats hold a lead over Republicans on the statewide Generic Congressional Ballot 46.4% to 39.8% among registered voters and 48.3% to 40.7% among likely voters. When undecided voters are leaned, Democrats slightly expand their lead to 7 points, 49.5% to 43.1% among likely voters.
Trust to Handle Issues
Democrats are more trusted than Republicans to handle all three of the top five voting issues to Michiganders in 2026, while Republicans are trusted more on two of all seven issues tested. On Cost of Living / Inflation (Rank 1), Democrats have a 47.7% to 34.6% edge; 48.3% to 36.5% on Economy / Jobs (Rank 2); and 53.3% to 33.2% on Healthcare / ObamaCare (Rank 3).
Republicans lead 42.7% to 39.8% on Crime / Safety (Rank 4) and 43.9% to 41.0% on Immigration / Border Security (Rank 6). Democrats lead on Corruption / Accountability (Rank 5) 45.3% to 32.9%.
Methodology
The Public Polling Project conducted by BIG DATA POLL interviewed 725 registered voters and 678 likely voters in Michigan from September 22 – 24, 2026. Interviews conducted online (≈40%) are sourced through Lucid (CINT) and live-agent phone interviews (≈60%) including P2P SMS and text-to-online are sourced from either the L2 National Voter File Database. Participants who opted for text-to-online were given 24 hours to complete the interview. Interview details plotted on maps can be reviewed by hovering and clicking on the locator pins. Results were weighted for sex, age, race and ethnicity, education, and geography. The overall sampling error is ±4.0% for both the registered voter sample and the likely voter sample at a 95% confidence level. It is important to note that sampling errors for subgroups are higher. All BIG DATA POLL publicly conducted surveys are crowdfunded via the PUBLIC POLLING PROJECT, supplemented if necessary by BIG DATA POLL and are NOT funded by or affiliated with any candidate, campaign, committee, or political entity. Full and interactive crosstabs can be viewed on MarketSight.
