Wisconsin’s Democratic primary for governor is producing the kind of political drama that can make a poll look more decisive than it is. Francesca Hong has gained support among progressive voters, Mandela Barnes has left the race, David Crowley has returned, and many voters remain undecided. The numbers are a snapshot of a moving field.
The Associated Press reported that a Marquette Law School poll found Hong leading the Democratic candidates, while roughly one-third of primary voters remained undecided. AP also reported that Barnes withdrew less than two weeks before the August 11 primary. The poll was conducted before the field fully settled, so its topline cannot be read as a final vote prediction.
A primary poll answers a limited question. It measures preference among the people who were interviewed at that moment, under the turnout assumptions and candidate list used by the pollster. It does not automatically tell us which candidate can win a general election, how late-deciding voters will break or whether a candidate’s supporters will turn out at the same rate as the other side.
The Wisconsin race adds another complication: withdrawals change the meaning of earlier responses. A voter who named Barnes may now choose among Hong, Crowley and the remaining candidates, stay home or select a different candidate without the poll having captured that transition. The campaign messages will also change because candidates no longer need to divide attention among the same opponents.
Enthusiasm is relevant, but it is easy to misuse. A voter can say the race is important and still fail to vote. A voter can express uncertainty and still participate. Early voting and absentee returns may reflect campaign organization, convenience and timing as much as intensity. A good analysis therefore looks at the question wording, sample type, field dates, undecided share and likely-voter screen.
The general election will pose a different test. The eventual Democratic nominee is expected to face Republican Rep. Tom Tiffany, who has President Trump’s endorsement. Voters who agree on a primary label may disagree over electability, public safety, taxes or the role of state government once the contest becomes two-sided.
The useful conclusion is not that Hong has won or that the primary is meaningless. It is that the field changed faster than the old poll can measure. The next survey should disclose how it handles Barnes’s departure, Crowley’s return and the large pool of undecided voters.
Numbers are evidence, not prophecy. In Wisconsin, the next ballot—not the loudest interpretation of the last poll—will decide the nomination.