In a contentious twist to the Michigan Senate race, Tony Fabrizio, a prominent pollster for former President Donald Trump, has dismissed recent polls showing progressive candidate Abdul El-Sayed ahead of Republican Mike Rogers as inaccurate and misleading. According to Fabrizio, these polls disproportionately include Democratic-leaning demographics, thereby skewing results in favor of El-Sayed.

A newly released internal memo obtained by POLITICO reveals Fabrizio's critique of the polling methodology, particularly taking issue with the over-sampling of college-educated and Black voters. He points to a Susquehanna poll which places El-Sayed seven points ahead and a TIPP poll giving him a four-point lead as key examples of this alleged bias. In contrast, Fabrizio's own research for AARP, conducted with the Democratic firm Impact Research, shows the race as nearly tied.

The debate over polling accuracy has intensified, with Fabrizio citing "some of the most egregious examples of public polling failures" in recent Democratic primaries across Michigan, Wisconsin, and Florida. These instances underscore the Republicans' growing concerns over a series of unfavorable polls for Rogers, whose performance could be critical in determining which party controls the Senate.

In defense of their methodology, Jim Lee, CEO of Susquehanna Polling and Research, stated to POLITICO that their sampling did not intentionally over-represent any demographic group. Lee suggested that the real issue might be a significant enthusiasm gap between Democratic and Republican voters this cycle. Meanwhile, TIPP has not yet responded to requests for comment on their polling methods.

Further critiquing the Susquehanna poll, Fabrizio argued that the representation of college-educated voters was unrealistically high. "50% of the sample in this survey is college-educated – making them massively overrepresented," Fabrizo noted, adding that according to 2022 and 2024 Exit Polls and VoteCast data, such a demographic is unlikely to constitute even 40% of Michigan's electorate. He concluded that by overestimating the number of college-educated voters, these polls likely inflate El-Sayed’s projected support.

This ongoing dispute highlights the complexities and challenges inherent in political polling, where demographic assumptions can significantly influence projected election outcomes. As Michigan prepares for this crucial Senate race, both sides are scrutinizing the polls, each hoping for an edge in what is shaping up to be a tightly contested battle.