How 153 Voters Could Have Changed Maine’s 2nd District
A razor-thin ranked-choice elimination may have determined not just the order of finish, but the Democratic nominee
Abstract: Matt Dunlap won the Democratic nomination in Maine’s 2nd Congressional District after surviving Jordan Wood by only 304 votes at the decisive elimination point. A shift of just 153 voters could have produced a different nominee, and the result offers an important preview of November’s contest against former Governor Paul LePage—and of the opening ranked-choice voting might create for a credible independent.
Matt Dunlap’s victory in the Democratic primary for Maine’s 2nd Congressional District offers an unusually clear illustration of both the advantages and the peculiarities of ranked-choice voting.
Dunlap appeared to be running third in early election-night returns. But those returns were incomplete. Once all first-choice votes were counted, the official first round was:
Joe Baldacci: 24,966
Matt Dunlap: 22,933
Jordan Wood: 22,712
Paige Loud: 8,194
Baldacci finished first, but with less than one-third of the vote. Because no candidate received a majority, the ranked-choice process began.
Loud was eliminated first. After her ballots were transferred to each voter’s next available choice, the remaining candidates stood at:
Baldacci: 25,923
Dunlap: 25,681
Wood: 25,377
Only 546 votes separated first place from third.
More importantly, Dunlap led Wood by just 304 votes.
Wood was therefore eliminated. His supporters strongly preferred Dunlap to Baldacci: 10,243 of Wood’s ballots transferred to Dunlap, while 6,632 transferred to Baldacci. Another 8,502 Wood ballots did not contain a continuing preference for either finalist.
The result was:
Matt Dunlap: 35,924
Joe Baldacci: 32,555
Dunlap won with approximately 52.5 percent of the votes remaining in the final round.
That explains how Dunlap won. But it does not fully capture how close the race came to following an entirely different path.
The Decisive 304-Vote Margin
Suppose that 153 voters counted for Dunlap at the decisive stage had instead supported Wood.
Each switched vote would subtract one from Dunlap and add one to Wood, changing the difference between them by two votes. A shift of 153 voters would therefore have erased Dunlap’s 304-vote advantage and placed Wood narrowly ahead.
The approximate totals would then have been:
Baldacci: 25,923
Wood: 25,530
Dunlap: 25,528
Dunlap—not Wood—would have been eliminated.
The later preferences of Dunlap’s supporters would then have determined whether Baldacci or Wood won the nomination.
Wood would have entered that hypothetical final round only 393 votes behind Baldacci. Among Dunlap ballots that ranked remaining candidate, Wood would have needed to receive only 394 more transfers than Baldacci to move ahead.
That would not have required an overwhelming preference for Wood. A relatively even division, tilted modestly toward Wood, could have made him the nominee.
Would Wood Have Won?
There are reasons to believe that Wood would have had a strong chance.
Wood, Dunlap and Loud all ran to Baldacci’s left and supported Medicare for All. They also criticized the Democratic establishment’s support for Baldacci. When Wood was eliminated, his supporters preferred Dunlap to Baldacci by a substantial margin.
It is therefore plausible that Dunlap’s voters would have reciprocated by preferring Wood to Baldacci.
But the available totals do not prove that conclusion.
Voters do not rank candidates solely along a left-to-right ideological scale. Biography, geography, age, political experience, personality and perceptions of electability can all affect second choices.
Dunlap and Baldacci were both familiar Maine officeholders with strong roots in the Bangor–Old Town area. Dunlap had served as secretary of state, state auditor and a state legislator. Baldacci was a lawyer, former Bangor city councilor and state senator.
Wood, by contrast, was a younger former congressional staffer from Auburn who had worked as chief of staff to California Representative Katie Porter. Some Dunlap voters may have been ideologically closer to Wood but culturally or institutionally more comfortable with Baldacci.
The public round totals show how Wood’s voters divided between Dunlap and Baldacci. They do not show how Dunlap’s voters would have divided between Wood and Baldacci.
The appropriate conclusion is therefore not that Wood certainly would have won. It is that Wood might well have won, and that an extraordinarily small change in the vote would have tested a completely different final matchup.
What This Says About Ranked-Choice Voting
Ranked-choice voting is often described as a system that identifies the candidate with the broadest support.
In one important sense, it accomplished that goal. When Dunlap and Baldacci became the finalists, Dunlap received most of the ballots that continued into the final round.
But ranked-choice voting does not compare every candidate directly against every other candidate.
It follows a sequence. The lowest candidate is eliminated in each round, and that elimination determines which comparison occurs next.
The Maine count eventually compared Dunlap with Baldacci. It never produced a Wood–Baldacci final or a Wood–Dunlap final.
Wood’s 304-vote deficit prevented those comparisons from taking place.
The outcome was therefore path dependent. A very small change in the elimination order could have created a different final pairing and possibly a different winner.
That is not necessarily an argument against ranked-choice voting. Every electoral system has rules that become decisive in close contests.
Under ordinary plurality voting, Baldacci would have won despite receiving less than one-third of the first-choice vote. Ranked-choice voting instead produced a nominee supported by most of the active final-round ballots.
But the phrase “majority winner” requires some qualification. It does not always identify the only candidate who could have assembled a majority. It identifies the candidate who assembled a majority after a particular sequence of eliminations.
Matt Dunlap won under Maine’s established rules. Yet a shift of fewer than two-tenths of 1 percent of the ballots could have eliminated him—and might have made Jordan Wood the Democratic nominee.
November: A Two-Candidate Race—For Now
Dunlap will now face former Republican Governor Paul LePage in the general election.
The contrast is unusually sharp.
Dunlap has identified Medicare for All as one of the policies he intends to advance. LePage, during his years as governor, repeatedly opposed Medicaid expansion and resisted implementing it even after Maine voters approved expansion in a statewide referendum.
That leaves substantial political space between the two candidates.
It is disappointing that voters currently have no centrist or pragmatic option to succeed Jared Golden, whose political appeal rested in part on his willingness to depart from national party orthodoxy. The district is now being offered a much sharper ideological choice at a time when many voters may prefer continuity, moderation and practical problem-solving.
A pragmatic independent could reject both the Republican impulse to reduce public health care assistance and the Democratic proposal to replace most existing insurance with Medicare for All. Such a candidate could support improved universal coverage while preserving private insurance, state exchanges, employer contributions and individual choice.
That middle-ground position is developed in my recent speech, Beyond Medicaid Cuts and Medicare for All. It is written for an independent or third-party congressional candidate running against a Republican who favors reducing health care assistance and a Democrat who supports Medicare for All.
Maine’s 2nd District would appear to offer a particularly interesting setting for such a candidacy.
The relevant standard is not whether an organized minor party has nominated someone. Organized third parties rarely win congressional elections. The more important question is whether a credible individual—with a substantive career, a strong local reputation and the personal standing to compete with the major-party nominees—could finish among the top two.
Under Maine’s ranked-choice system, such a candidate would not necessarily be a spoiler.
Suppose a credible independent finished second in first-choice votes. The third-place major-party candidate would be eliminated, and that candidate’s later preferences could propel the independent to victory.
The Democratic primary demonstrated the basic mechanism. Dunlap did not begin in first place. He won because he survived the elimination round and then received enough support from Wood’s voters to pass Baldacci.
A third candidate in November could follow a similar path—but only by reaching second place. A candidate finishing third would be eliminated and could affect the result only through the later preferences of that candidate’s supporters.
At present, however, no third candidate has qualified for the printed ballot in Maine’s 2nd Congressional District. The official nonparty candidate list contains no congressional candidate, and neither the Green Independent Party nor the Libertarian Party nominated one.
A declared write-in candidate can still enter by August 25. Such a candidacy would trigger ranked-choice voting in the federal race, but the candidate would face the enormous disadvantage of having no name printed on the ballot. Building the recognition and organization needed to finish second as a write-in would be exceptionally difficult.
Unless such a candidate emerges, November will be a direct Dunlap–LePage contest, and no ranked-choice redistribution will be required.
LePage enters the general election with significant advantages. He carried the 2nd District during his unsuccessful statewide campaign for governor in 2022, and Donald Trump carried the district by approximately nine points in 2024. National Republicans regard the open seat as one of their strongest opportunities to gain a Democratic-held district.
Dunlap must unite a Democratic electorate that was deeply divided in the primary while persuading independent voters that his progressive positions are compatible with the needs of a politically mixed and largely rural district. His success in attracting Wood’s later-choice voters demonstrates an ability to assemble a coalition within the Democratic primary. In a two-person general election, however, there may be no later rounds and no transferred votes to rescue either candidate.
The primary showed how a few hundred votes can determine not only who survives a ranked-choice count, but which political coalition is given the opportunity to form a majority. The general election will determine whether Dunlap can build a broader coalition against one of Maine’s best-known—and most polarizing—political figures.