FLIPR Shifts Wisconsin Governor Race to Lean D After Crowley’s 2-Poll Edge

Q: Can just two early polls accurately predict the outcome of a highly volatile gubernatorial race after a chaotic primary season?

No. Two early polls can justify only a tentative forecast, not a reliable prediction in a race this unsettled. Silver Bulletin’s change from Toss-Up to Lean Democratic is understandable because the nominee changed and the only available head-to-head polling shows David Crowley narrowly ahead of Tom Tiffany. But those polls are a very thin sample. In a state like Wisconsin, where statewide races are often decided by less than a point, a small polling edge can easily fall within normal survey error. The broader context makes early polling even less stable. Crowley won the Democratic primary by roughly four-tenths of a point after a chaotic campaign in which candidates dropped out, re-entered, and absentee voting was already underway. Pre-primary polling did not capture that late movement well: Francesca Hong had led many surveys, yet Crowley still pulled out a narrow upset. That is a warning that voter preferences were fluid and not fully settled. There are also major unknowns that two polls cannot fully measure yet: whether Democratic voters unify after the close primary, how independents react once both nominees are defined, how much outside money enters the race, and whether Wisconsin’s national environment shifts over the fall. So the two polls are useful as an early signal, not a dependable forecast. They support calling Crowley slightly favored at this moment, but the race remains highly volatile and could move quickly as more polling and campaign data come in.

3 articles · Updated · natesilver.net · Aug 12

Wisconsin moved from Toss-Up to Lean D in FLIPR’s updated 2026 midterm forecast after David Crowley’s upset Democratic primary win replaced Francesca Hong as the modeled nominee.
Two polls now in FLIPR’s database test Crowley against Republican Tom Tiffany, and both show Crowley narrowly ahead — consistent with the new Lean D rating.
FLIPR said the switch reflects how its forecast updates candidate assumptions as primaries finish; Hong had remained in the model earlier because pre-primary polling showed her with a large lead.
The broader forecast page tracks Senate, House and gubernatorial races jointly, using about 40,000 simulations and updated polling inputs as Democrats look for gains in 2026.