When two candidates share almost the same name, election law runs headlong into its oldest tension: protecting voter choice while preventing tactics that muddle voter intent. Alaska’s “two Dan Sullivans” Senate race put that tension on center stage — and the Department of Justice’s move to condition certification on an after-the-fact confusion probe made the stakes unmistakable.
The Short Version
- The DOJ asked a federal court to halt Alaska’s Senate certification if incumbent Dan S. Sullivan loses, citing likely confusion with another candidate named Daniel J. Sullivan Jr.
- Alaska officials tried to mitigate confusion by labeling the incumbent with a middle initial, “registered Republican,” and “incumbent,” while listing the other Sullivan without a party label.
- Alaska courts kept Daniel J. Sullivan on the ballot; the state’s ranked-choice structure arguably cushions confusion by letting voters rank both Sullivans.
- The dispute spotlights a core doctrine: courts balance broad ballot access against sham or “nuisance” candidacies designed to distort voter intent, but they demand concrete proof, not conjecture.
What the DOJ’s lawsuit actually argues
The Justice Department’s complaint hinges on a specific, plausible mechanism of harm: that some voters who support the incumbent Dan S. Sullivan would mistakenly select Daniel J. Sullivan because of their strikingly similar names, potentially distorting a ranked-choice tabulation. The filing states that “some supporters of Senator Sullivan likely will inadvertently vote for Daniel J. Sullivan,” and seeks to enjoin Alaska from certifying the race if the incumbent loses until an investigation determines whether votes for Daniel J. were intended for Dan S.
That theory draws its force from two facts few would contest: the names are unusually similar and they appear in the same contest. Precedent recognizes that election administration must guard against confusion that defeats the electorate’s will — but remedy requires proof, not intuition. Here, the DOJ asks a federal judge to delay certification and supervise a post-election inquiry into voter intent, a remedy that is extraordinary precisely because it tries to reconstruct mental states after the votes are cast.
How Alaska structured the ballot to separate two near-identical names
Alaska’s administrators anticipated the similarity problem and changed the information architecture of the ballot to help voters distinguish the two Sullivans. The incumbent appears as “Dan S. Sullivan,” with “registered Republican” and “incumbent” flags; Daniel J. Sullivan Jr. appears by full name without a party label. These are classic design interventions — salience cues that signal identity, officeholder status, and partisan alignment — intended to lower the cognitive burden at the moment of choice. Multiple independent outlets reported these specific label choices.
Courts allowed Daniel J. Sullivan to remain on the ballot. The Alaska Supreme Court’s action did not validate the wisdom of the candidacy; it reflected a familiar line in election law: access to the ballot is presumptive, and claims of “nuisance” or sham entry must be substantiated. Where the evidence of orchestrated confusion is thin, courts default to voter choice while encouraging election officials to mitigate risk through labeling and layout.
Ranked-choice mechanics: why confusion may be blunted, not amplified
Ranked-choice voting (RCV) changes the damage function of confusion. Under RCV, voters can rank multiple candidates in order of preference; if a voter mistakenly ranks the “wrong” Sullivan first but the “right” Sullivan second, that second preference can still count once the first-choice candidate is eliminated. In a same-name scenario, this redundancy can serve as a buffer: it localizes the cost of confusion to early rounds and lets later transfers realign with true intent. Analysts who looked at Alaska’s setup have argued that RCV, paired with labeling, reduces the room for misattribution to swing a final outcome.
That is not to say confusion is harmless; it never is. But the system’s transfer logic means the DOJ’s direst scenario — inadvertent first-choice votes flipping a final winner — requires more than casual voter error. It requires error at scale without corrective ranking, plus a margin narrow enough that corrected transfers would have changed the finish. Evidence of such a pattern, if it exists, must be empirical, not hypothesized.
Where the burden of proof belongs — and why courts are cautious
American election law recognizes two overlapping commitments: keep the ballot accessible and keep it intelligible. Courts and administrators distinguish ordinary candidacies by similar-named citizens from “nuisance” entries meant to degrade the signal of voter intent; the latter requires showing more than mere resemblance — indicators of bad faith, coordination to confuse, or other facts that transform similarity into subversion. Alaska’s own administrative record shows officials scrutinized Daniel J. Sullivan’s filing for good faith, but judicial review ultimately left the choice to voters, fortified by clarifying labels.
Given that posture, a federal injunction that withholds certification absent a showing of actual confusion is a hard sell. Federal courts are generally reluctant to preempt or second-guess state certification processes absent a clear violation of federal law, especially when state courts have already set the ballot and the election is underway. Put bluntly: remedy follows proof. Without data showing miscast ballots at a scale that would alter the outcome, a sweeping certification hold risks overshooting the problem the system already tried to solve.
🚨 ALASKA SENATE SHOWDOWN!
The DOJ is fighting to block certification of a narrow defeat for Republican Sen. Dan S. Sullivan, warning that a second candidate with nearly the SAME NAME could confuse voters.
Two Dan Sullivans. One Senate seat. A major legal battle.
Is this voter…
— Xi Znping (@XiZnping) October 27, 2026
Mechanism and measurement: how you would actually test confusion
Confusion is measurable, but only with discipline. Analysts would look for anomalous vote patterns inconsistent with known partisan baselines — for example, precincts that heavily back the incumbent in other races yet register outsized first-choice support for the similarly named challenger, without commensurate second-choice rankings correcting the pattern. Survey instruments, post-election audits, and undervote/overvote diagnostics — long used in ballot design research — can identify where design fails voters and how often. Those methods, not speculation, are what convert a theory of harm into a record courts can act on.
Conversely, if ballots clearly differentiate the candidates and RCV transfers collapse back to expected alignments, any remaining confusion signal is likely too weak to justify setting aside certification timelines. The law tolerates some level of voter error; it intervenes when process defects make it impossible to discern the electorate’s will.
The broader lesson: design early, litigate late — if at all
The Alaska case underscores a durable lesson for administrators and campaigns. The most reliable safeguard against same-name distortions is not post-election litigation; it is pre-election design — name formatting, incumbency flags, party labels, and clear instructions — paired with voter education that anticipates edge cases. Where RCV is used, teaching voters to rank all acceptable candidates, including near-name duplicates, further hardens the system against confusion.
Litigation has a role when evidence shows the ballot failed its central job: translating voter intent into a tabulation that reflects it. But courts will demand specificity — the measurable, not the merely imaginable. In Alaska, the DOJ has articulated a real risk mechanism. The counter is not denial; it is the record Alaska built: explicit labels, an RCV structure that cushions misfires, and judicial review that kept the ballot open while urging clarity. If confusion truly threatens to invert the will of the voters, the numbers will show it. If not, certification should proceed on schedule.
Sources:
npr.org, goodmorningamerica.com, cbsnews.com, alaskasnewssource.com, cnn.com, usatoday.com, anycase.ai
