El-Sayed Defeats Stevens in Michigan's Democratic Senate Primary by 1 Point, on Record Turnout, Despite $30M+ AIPAC Onslaught

Abdul El-Sayed, the former Michigan health director, defeated Rep. Haley Stevens in Michigan’s Democratic U.S. Senate primary on August 4, 2026, by a margin of roughly one percentage point — 14,893 votes out of more than 1.5 million ballots cast statewide. The Associated Press called the race for El-Sayed on the morning of August 5, once nearly all ballots were counted; the outcome was not certain on election night itself. El-Sayed advances to face Republican former Rep. Mike Rogers, who ran unopposed, in the November general election.

Turnout was historic. Per Bridge Michigan, the Democratic Senate primary drew more than 1.5 million ballots statewide and is being described as the highest-turnout primary contest in Michigan’s history for a state summer primary; over 1.3 million of those ballots were cast before Election Day, the large majority absentee. Bridge Michigan’s separate turnout analysis put the Democratic primary specifically at roughly 1.2 million voters, up from about 933,000 in the 2024 Democratic primary — a 29 percent increase.

El-Sayed’s win was concentrated rather than broad. He carried 30 of Michigan’s 83 counties, drawing close to 60 percent support in college-town counties (Washtenaw, Ingham, Kalamazoo, Isabella, Kent), topping 70 percent in some Ann Arbor precincts and 80 percent in parts of Grand Rapids, and winning nine counties in the Upper Peninsula plus nine north of Grand Rapids. He lost all of metro Detroit and underperformed in the state’s traditional Black-population strongholds — Stevens won seven of twelve majority-Black counties by an average 8.5-point margin. Election Day same-day voters were decisive in metro Detroit specifically, where El-Sayed took 59 percent of same-day ballots to Stevens’ 37.9 percent, according to Bridge Michigan’s precinct analysis. State Sen. Mallory McMorrow, who suspended her campaign in early July but remained on the ballot, drew about 4 percent of the statewide vote — more than El-Sayed’s full margin of victory over Stevens.

The result came despite a historic outside-spending imbalance. The American Israel Public Affairs Committee’s super PAC, United Democracy Project (UDP), spent roughly $30.6 million opposing El-Sayed and supporting Stevens — AIPAC’s largest single-race investment on record — and total spending favoring Stevens (campaign plus outside groups) exceeded $61 million against roughly $15 million for El-Sayed, a combined disadvantage of better than four to one. El-Sayed had modestly outraised Stevens on candidate-committee fundraising alone (about $14.5 million to $11.9 million through mid-July, per FEC filings) before the outside-money gap opened.

Post-election coverage from The Intercept, Al Jazeera, and Jewish Currents frames the result as a setback for AIPAC’s spending model and connects it to the candidates’ sharply divergent positions on Israel and Gaza — El-Sayed opposed U.S. military aid to Israel, Stevens supported it — with The Intercept reporting El-Sayed “turned those millions of dollars into a key campaign issue,” using UDP’s spending itself as evidence of the “danger of lobby spending and corporate money in politics.” Former Rep. Andy Levin, quoted in the same coverage, argued primary voters read a candidate’s refusal to criticize Israel’s conduct in Gaza as a marker of being “bought and paid,” tying that judgment to broader anti-corporate, cost-of-living messaging rather than treating Israel/Palestine as a stand-alone issue.

El-Sayed’s campaign also relied on an on-the-ground volunteer operation described in multiple outlets as roughly 12,000 door-knocking volunteers, run alongside a Medicare for All and cost-of-living-focused message.

Note on the data-center wedge-issue framing (see also 2026-07-29–michigan-data-centers-divide-democratic-primary-el-sayed-stevens): data-center siting and utility-cost-shift policy remained a live, tier-1-documented point of disagreement between the candidates through election day, and El-Sayed’s position (utility-style regulation, “terms of engagement”) was more aggressive than Stevens’. But none of the post-election coverage surveyed for this entry — CNN, CNBC, Bridge Michigan, NBC News, Detroit News, The Intercept, Al Jazeera, Jewish Currents, Roll Call, or E&E News/POLITICO — credits data-center opposition as a decisive or even clearly-measurable factor in the outcome. No exit poll was conducted in the Senate primary, so no outlet has demographic or issue-attribution data to isolate the data-center effect from turnout, geography, or the Israel/Gaza divide. E&E News/POLITICO’s pre-election framing (University of Michigan’s Tyler Simko) had already flagged that data centers would “matter to some voters much more than others” and would compete with Gaza, Medicare, and corporate-spending messaging rather than stand alone — a caveat post-election reporting has not resolved either way. See the forthcoming postmortem note for a fuller assessment.

Open leads: Garfield Township’s board vote on its data-center moratorium-enabling ordinance is scheduled for August 11, 2026, a week after this primary — track separately. Confirm whether any post-election precinct-level analysis (e.g., a future academic or think-tank study) attempts to isolate the data-center issue’s electoral effect, since no outlet has done so as of this writing.