The polls say Jocelyn Benson leads John James by about six and a half points. The Michigan governor prediction market on Kalshi says she wins roughly nine times out of ten. Those two numbers look like they describe different races, and the most common explanation I hear is the third-party vote: if a Libertarian, a Green and a U.S. Taxpayers candidate are pulling from one side, maybe the two-way polls are hiding the real shape of the contest. It is a reasonable guess, and once you check the ballot, the polling tables and Michigan's own recent history, it turns out to be mostly wrong. The third-party vote in this race is small, the "other" number in the polls is mostly something else entirely, and the gap between a six-point lead and a 90-cent price is explained by what a prediction market actually measures. I will walk through all three, and by the end you will have a much cleaner way to think about those twelve unassigned poll points.
The Quick Answer: Does The Third-Party Vote Explain The Gap?
No. The three minor-party tickets on Michigan's November ballot have taken between about 1.6% and 3% of the vote combined in the last two governor races, nowhere near the 12 points the polling average leaves unassigned. That unassigned share is almost certainly mostly undecided voters: the polls do not break it out, but the minor-party ceiling in Michigan's last two governor races was 2.9%, so at most about a quarter of it can be third-party, and undecideds are not a candidate. The distance between a six-point poll lead and a roughly 90-cent Kalshi price is the difference between vote share and win probability. The ballot, the poll table, the 2018 and 2022 minor-party totals, and the one independent contract still listed, whose candidate quit in May, are all below.
Free: The Weekly PM Market Brief
The 8-model panel's graded record, the week's biggest market-vs-model gaps, and what's spiking next. One email, Sundays.
Unsubscribe anytime, one click. We never share your email.
New to this? A price on Kalshi is a probability in cents: what the price actually means covers the mechanics in five minutes, and this page is the Michigan layer on top of it.
The Gap, In Numbers
The whole argument fits in one table. The poll side is the Pollsmax average of 20 public general-election polls, as of its August 11 update; the market side is Kalshi's GOVPARTYMI-26 board, fetched August 28, 2026, around 9:40 a.m. ET. That board is written as a party market: each line asks whether a given party's nominee wins the governorship, and Duggan's independent bid is listed as its own candidate-scoped line: the rules pay out only if a representative of the "Mike Duggan (Independent) party" is inaugurated, so the line is about him, not about any independent.
| Line | Party | Kalshi last trade | Kalshi bid / ask | Polling average (20 polls) |
|---|---|---|---|---|
| Jocelyn Benson | Democratic | 92.9¢ | 88¢ / 92.7¢ | 47.3% |
| John James | Republican | 12¢ | 9.1¢ / 12¢ | 40.8% |
| Mike Duggan (Withdrew May 2026) | Independent | 0.2¢ | 0.2¢ / 1¢ | not polled |
| Unassigned (100 Minus Benson Minus James) | Libertarian, Green, U.S. Taxpayers, undecided | no contract | no contract | 11.9% |
Prices shown are market quotes at the time of the fetch; our own model reads are model estimates, not predictions of fact and not financial advice. The contract settles on the certified winner of the November 3, 2026 election, and Kalshi can settle it early once a consensus of media organizations projects the winner. Kalshi is a CFTC-regulated event-contract exchange; it sets its own minimum age at 18+, and availability varies by state as of August 2026.
Two things jump out. First, the polls leave almost 12 points of the electorate unassigned, and if you are hunting for a third-party story, that is the number you are staring at. Second, what buyers are offering for Benson and what sellers are asking sit just under five cents apart (88¢ against 92.7¢), which tells you this board is thin. Hold that thought, because the thinness matters later.
That gap between what buyers offer and what sellers want is the market's way of saying the price is a range, not a point, and it is the row I keep coming back to. Treat "Benson at 93¢" as "Benson somewhere between 88¢ and 93¢" and every comparison below gets more honest.
Who Is Actually On The Ballot Besides Benson And James
The Michigan Bureau of Elections' unofficial candidate list for November 3, 2026, pulled August 28, shows five tickets for governor and lieutenant governor:
| Ticket | Party | How the ticket got here |
|---|---|---|
| Jocelyn Benson / Winnie Brinks | Democratic | Benson won the August 4 primary |
| John James / Jay DeBoyer | Republican | James won the August 4 primary |
| Anthony Hudson / Beau Parmenter | Libertarian | Party convention, June 29 |
| Donna Brandenburg / Robert Donald Cowper II | U.S. Taxpayers | Party convention, July 20 |
| Douglas Campbell / Bobbie Clay | Green | Party convention, July 20 |
The name that is not on that list is the one the market still prices. Mike Duggan, the former Detroit mayor who launched an independent run in December 2024, ended his campaign on May 21, telling supporters that being down 11 points and behind in fundraising left no realistic path. He was the only third-party candidate in this race with real money, real name recognition and a real polling floor. With him gone, the "third-party vote" in Michigan is three convention-nominated tickets with no statewide polling presence. Duggan himself was an independent, not a third party, and the distinction matters for everything below: the minor parties are on the ballot, and the one independent who could have split the race is not.
That is the first half of the answer. If the polls were hiding a serious third-party candidate, the market would have to price the chance that candidate wins, and it would have to price how that candidate splits the two-way vote. There is no such candidate. The Duggan contract that sits on the board at 0.2¢ is the paperwork of a finished story, and I will come back to why it still rivals James for the most-traded line on the whole market.
What Third Parties Actually Got In 2018 And 2022
The second half of the answer is history. Michigan runs the same minor parties in most cycles, and their combined governor-race share is small and stable.
| Year | Democrat | Republican | Combined minor parties |
|---|---|---|---|
| 2022 | Gretchen Whitmer 54.47% | Tudor Dixon 43.94% | about 1.6% |
| 2018 | Gretchen Whitmer 53.31% | Bill Schuette 43.75% | about 2.9% |
Sources: Wikipedia, 2022 Michigan gubernatorial election; Wikipedia, 2018 Michigan gubernatorial election.
Three percent is the recent ceiling. To get from a 2-to-3-point third-party share to the 12 points the polls leave open, you would need a Duggan-sized candidate on the ballot. There isn't one. So the 12 points cannot be third-party voters in any meaningful number, which forces the real question: what are they?
Where The Twelve Points Actually Go
They are undecided voters, and the poll table makes that obvious once you look at the "other" column poll by poll instead of in the average.
| Pollster | Date | Benson | James | Unassigned (100 minus the two) |
|---|---|---|---|---|
| Fabrizio Ward / Impact Research (For AARP) | Aug 9-11, 2026 | 49% | 44% | 7% |
| EPIC-MRA | Jul 31, 2026 | 47% | 38% | 15% |
| Mitchell Research | Jul 29, 2026 | 46% | 44% | 10% |
| Glengariff Group | Jul 24, 2026 | 44% | 39% | 17% |
| Mitchell Research | Jun 13, 2026 | 51% | 40% | 9% |
| TIPP Insights | May 23, 2026 | 49% | 38% | 13% |
Sources: Pollsmax Michigan governor polling table; the AARP release on the August survey. The unassigned column is my arithmetic, not the pollsters' label; it bundles undecided, refused and any minor-party responses together.
The unassigned column swings from 7% to 17% depending on the pollster, and it swings for reasons that have nothing to do with the Libertarian ticket. Glengariff's July survey left 17% unassigned, and its own release shows what that bundle was made of: about 16% did not know, and 1.6% said someone else. The AARP survey in August left 7%. The March MSU survey, which polled 1,000 adults rather than likely voters and offered Duggan as a choice, left 28% off the two major names in Pollsmax's two-way row. That one is the exception that proves the rule: the number is high because a credible independent was on the card, and he is the candidate who no longer exists. If the minor parties were collecting a large share, it would show up as a stable floor across pollsters, the way the 1.6% and 2.9% totals above did on Election Day; instead the number moves with the pollster's method. The number is a measure of how hard the pollster pressed, not a measure of minor-party support. The Glengariff row is the one that tells the story: 17% undecided in late July, and the named responses went to Benson or James.
And undecideds are not a contract. When you buy Benson at 90¢, you are not buying her share of the vote; you are buying the probability she wins. An undecided voter who breaks toward James in October does not create a third outcome. They just move the margin, and the margin is what the market is pricing.
Why A Six-Point Lead Trades Like A 90-Cent Contract
This is the mechanism that actually explains the gap, and it is the part the third-party theory skips.
Strip the undecideds out and Benson's share of the two-way vote in the average is about 53.7%, a lead of about 7.4 points on that two-way split (6.5 on the raw average). Now ask a different question: with about ten weeks to go, how often does a candidate up seven in the polling average lose? The answer depends on how wrong state polls tend to be. Michigan's last two governor races are instructive on that point, because in 2022 several late polls understated the Democrat badly: a Mitchell Research survey on October 19, 2022 had Whitmer 49, Dixon 47 and called it a statistical tie, and Whitmer won by 10.5 points three weeks later, after winning by 9.6 in 2018. A seven-point lead with a typical statewide polling miss of four or five points in either direction leaves the trailing candidate a real but small path, and small-but-real is what a 9¢-to-12¢ James contract describes.
That framing also explains the one place where our own seven-model read of this race and the market disagree. On August 7, with every public poll still predating the primary, we asked seven models to price Benson-James without showing them the market. They landed at 81% for Benson and 17.4% for James, against 86.5¢ and 15.5¢ at the time. Since then the Fabrizio poll landed at Benson +5 and the market has drifted higher on Benson, to the 88-to-93-cent range you see above. Read the 81% as context, not as a signal; the market price is the better forecaster more often than not, and gaps like that are on the graded midterm scoreboard for anyone who wants to check the record. The point for this article is narrower: neither the model nor the market moved on third-party math, because there is no third-party math to move on.
A Worked Example: Turning A Poll Lead Into A Price
Here is how I actually do the conversion, so you can run it on any governor board on the 2026 governor races page.
| Step | Michigan, August 28 |
|---|---|
| Two-Way Share (Drop The Undecideds) | Benson 47.3 / (47.3 + 40.8) = 53.7% |
| Two-Way Margin | Benson by about 7.4 points on the two-way split, 6.5 on the raw average |
| Typical Miss | Final statewide polls off by about 4.5 points on average; treat that as a standard deviation of roughly 5.5 |
| Chance The Lead Survives | 7.4 divided by 5.5 is about 1.35 standard deviations, which holds about 91 times in 100 |
| Same Arithmetic On The Raw 6.5 | About 1.2 standard deviations, roughly 88 times in 100 |
| What Kalshi Says | 88¢ bid, 92.7¢ ask |
The market lands inside the range the arithmetic produces, 88 to 91, and the whole spread between bid and ask is the argument over which margin to trust and how big a miss to fear. Note which direction Michigan's own recent misses ran: 2022's late polls understated the Democrat, so a repeat of that would push Benson's line up, not down. The bear case for her line has to lean on the presidential-cycle misses, where Michigan's polls have underestimated the Republican more than once in the last decade. That is the honest debate on this board, and it is a debate about polling error, not about minor parties. What the arithmetic never touches is the Libertarian, Green or U.S. Taxpayers share, because at 1.6% to 3% combined it moves the two-way margin by less than a point. Anyone expecting the third-party vote to be the missing variable can watch it disappear in this table.
The Dead Contract That Still Rivals James In Volume
Now the callback I promised. About 120,100 Duggan contracts have changed hands in the market's lifetime, a few hundred behind James's 120,800 and far more than Benson's 72,300. On a board where the candidate is no longer running, that looks absurd until you remember when it traded. Kalshi listed Duggan while he was still polling as a credible third option, with the March MSU survey describing his support as concentrated in the Detroit metropolitan area. Then he quit in May, and everyone who had bought Duggan to win had to sell to someone. Exit trades on a dead line pile up fast.
So the second most-traded contract on the Michigan governor board, a few hundred contracts behind James, is the ghost of an independent campaign, not evidence of a live one. The 0.2¢ bid on it is the price traders put on a re-entry that nobody, including Duggan, expects. For a single number that settles the third-party question, take that one: the only independent who could have split this race is priced at one-fifth of a penny. How liquidity works on these boards explains why a dead contract's volume never tells you anything about a live one's odds.
The Record Behind The Price
A price this lopsided invites the question of whether the market has priced in everything that could move it, so here is the documented record on both nominees, cited, and what it did or did not do to the number.
- Benson's Campaign-Finance Finding. In May 2025, Attorney General Dana Nessel's office determined that Benson violated the Michigan Campaign Finance Act by holding her January 2025 campaign launch in the lobby of a state office building. No penalty followed, because the statute gives the attorney general no enforcement mechanism when the violation involves the Secretary of State herself; the office called that gap a likely legislative oversight. The finding predates every poll in the table above, and Benson has led all six of the most recent ones.
- James And The Debates. During the primary, James skipped an October 2025 forum in Sparta and would not commit to an April 30 debate, drawing a public rebuke from the Oakland County Republican chair. His stated reason was avoiding party infighting. He won the August 4 primary with President Trump's endorsement regardless.
- The Complaints Over James's Official-Office Spending. In August 2025, a constituent filed a complaint with the U.S. House Ethics Committee alleging James used taxpayer-funded congressional resources to promote his run for governor. A separate whistleblower complaint filed May 27, 2026 alleged that YouTube ads paid for by his House office carried campaign messaging to viewers outside his district; the bipartisan House Communications Standards Commission dismissed that complaint on July 2, saying no official funds were used for campaign purposes and finding "no reasonable justification" for the complaint. No discipline resulted from either, and the James line has traded in the same low-teens-and-below range through the summer.
None of those three items involves a third party, and none has moved the board more than the ordinary drift of a thin market, which is the payoff of the spread you saw at the top: a board this thin drifts a few cents on nothing at all, so a few cents of movement is not a verdict on anything. The event that did reshape this market was Duggan's exit, and that was a story about whether an independent could raise money, not about scandal.
One Caveat Before You Try To Trade From Michigan
This is a Michigan race, and Michigan is one of several states where Kalshi's access is in court, and the one with a judge's order and a daily fine on the record this summer. On June 29 an Ingham County judge ordered Kalshi to stop offering sports event contracts to Michigan residents, at the request of Attorney General Nessel, with fines of $120,000 a day for noncompliance, and the fight over that order has since drawn in the CFTC, the federal regulator that oversees Kalshi. The order was about sports contracts, not election contracts, but the practical upshot for a Michigan reader is that platform availability in the state changes month to month. Check Kalshi's own eligibility screen before assuming you can trade this board from inside Michigan, and see where Kalshi is legal for the state-by-state picture as of this month.
- 2026 Governor Races: Every Kalshi Board We Track
- Michigan Governor 2026 Odds: Benson Vs James For The Open Seat
- Is It Legal To Bet On Elections In Michigan? The 2026 Rules
- Michigan Senate Election 2026 Candidates: El-Sayed Favored
- Minnesota Governor And Senate Odds 2026
What Would Actually Reprice This Race
Since the third-party vote will not, it is worth naming what will. The first is the undecided break: the Glengariff-style 17% pool has to land somewhere by late October, and nobody has a clean read on which way Michigan's late deciders go in a governor race, which is the real uncertainty inside this price. Watch the next Glengariff or EPIC-MRA release: an unassigned share under 10% with Benson still in the mid-40s is the shape that pushes her line into the mid-90s, while a James number in the mid-40s is the shape that pulls her back toward the 80s. The second is polling error running the other direction for once. Michigan's presidential polls have underestimated the Republican more than once in the last decade, and any post-Labor Day poll putting James within three points would be the first since Mitchell's late-July 46-44 and the first of the post-primary stretch, the kind of print that reprices a 9-to-12-cent line into the high teens. The third, money and turnout in a midterm that leans against the president's party, is already inside the 90-cent price; it would take a reversal of that environment, not a confirmation of it, to move the board.
For the reader who came here asking whether the Libertarian or Green ticket explains the market, the honest answer is that the market barely knows they exist, and neither does the polling; how Kalshi's election markets are built explains why a party board like this one never lists a ticket that polls under the rounding error. The action in this race is between the two names at the top of the ballot and the 12% of voters who have not picked one. If you want to see how this board compares to the rest of the map, the 2026 governor races odds board lines every state up side by side, and the Michigan Senate board shows the same electorate priced far closer to a toss-up. That contrast is the last word on the third-party theory: same voters, same minor parties on the ballot, two very different prices, because the polls in one race are close and in the other they are not. The only independent who could have changed that is priced at one-fifth of a penny.
Browse every live election board on the OddsShopper prediction markets hub, and start with how prediction markets work if this is your first one.
