When two candidates share the same name, the law’s central tension snaps into focus: election administrators must deter deception and confusion, but courts are loath to disturb ballots already cast under rules set before voting began. Alaska’s recent “two Dan Sullivans” case shows where that line is drawn: protect the secret ballot, avoid late-breaking remedies that chill participation, and resolve name disputes before Election Day—not after.
The Short Version
- Federal court rejected the Justice Department’s bid to delay certification and segregate ballots in Alaska’s Senate race involving two candidates named Dan(iel) Sullivan.
- The judge found the requested remedy infeasible and incompatible with voters’ right to a secret ballot and with stable election rules.
- Alaska’s courts had already allowed the similarly named candidate onto the ballot, rejecting a non-statutory “good-faith” candidacy test.
- The episode underscores a durable principle: resolve potential voter-confusion issues ex ante; post-election ballot triage is a legal dead end.
What the court actually decided—and why it matters
A federal district court in Alaska dismissed the Justice Department’s lawsuit that sought to condition or delay certification of the U.S. Senate race if incumbent Senator Dan Sullivan lost to an opponent with the same first and last name, Daniel J. Sullivan. The court identified two fatal defects. First, there was no workable, lawful way to segregate ballots cast for Daniel J. Sullivan while preserving the constitutional guarantee of ballot secrecy. Second, ordering such a process days before the election would likely confuse voters and deter participation—precisely the harms election law tries to avoid. In plain terms: even if confusion were plausible, the cure the government requested was worse than the disease.
The opinion also emphasized feasibility. By the time the suit was heard, early votes had been cast, commingled, and could not be matched back to individual voters without compromising secrecy. The court characterized the remedy as “infeasible” for that reason, and because federal lawyers did not provide a concrete operational plan that avoided those pitfalls. In emergency election litigation, that practical dimension often decides cases; a clever legal theory does not rescue an unworkable remedy.
How Alaska ended up with two Dan Sullivans on the same ballot
Months before the ruling, Alaska’s Division of Elections tried to keep Daniel J. Sullivan off the ballot, concluding that his filing was not a good-faith candidacy and was instead intended to confuse or mislead voters. The Division de-certified him accordingly, issuing a formal determination to that effect. That decision did not stand. A state superior court rejected the Division’s reliance on a “good-faith” requirement that appears neither in the U.S. Constitution nor in Alaska’s statutes or regulations, and the Alaska Supreme Court affirmed his placement on the ballot. The appellate posture is significant: once state courts said the candidate met the legal prerequisites to appear, the federal court was not going to improvise new, midstream procedures to police voter intent after the fact.
Alaska also uses ranked-choice voting, which became part of the state’s rebuttal to the Justice Department’s theory of confusion. If a voter wished to support both Sullivans—or misselected the “wrong” one initially—ranking the similarly named candidates first and second still captures clear voter preference in the tabulation rounds. That design does not eliminate the possibility of confusion, but it does undercut the premise that a single first-choice mark must be treated as a miscast vote demanding federal segregation and investigation.
The legal architecture: prevention up front, restraint after ballots are cast
American election law distinguishes between ex ante regulation (who gets on the ballot and how names appear) and ex post remedies (what courts may order once voting is underway or complete). Courts allow considerable preventive regulation, so long as the rules are clear, neutral, and adopted in time for compliance; by contrast, courts grow skeptical of late-stage changes that risk disenfranchisement, compromise secrecy, or induce uncertainty at the polls. The Alaska case lands squarely on the latter side. The federal court declined to segregate ballots or condition certification on an investigation of voters who selected Daniel J. Sullivan precisely because doing so would pierce secrecy and deter turnout—costs the law treats as unacceptable absent a narrowly tailored, statutorily grounded mechanism that simply did not exist here.
That posture is consistent with a larger pattern in emergency election disputes: federal courts often defer to state processes if state courts can act promptly and if post-hoc interventions would upend settled expectations of voters and administrators alike. The more entangled a proposed remedy is with identifying individual voters or interpreting their subjective intent beyond what the ballot legally captures, the less likely a court is to grant it once ballots are in the box.
Similar names are a known problem—so fix them before Election Day
Election administrators and political scientists have long documented that ballot presentation can influence voter behavior. Ballot order effects are real in low-information settings; Alaska’s randomized rotation can, by chance, place similarly named candidates adjacent to one another on some ballots, modestly increasing the risk of selection error. That is an argument for careful design and proactive voter education, not for post-election excavation of voter intent. Scholarship on candidate-name regulation converges on a simple rule: permit candidates to use the names by which they are genuinely known, but guard against deception, fraud, or bad faith by adopting clear, pre-election standards and enforceable documentation requirements.
The Alaska Division of Elections tried to operationalize that guardrail by inferring bad faith from circumstantial evidence and removing the candidate. State courts, however, demanded a statutory or regulatory hook rather than a freestanding “good-faith” test. The lesson for administrators is procedural, not philosophical: if the state wants to screen out spoiler-name tactics, it must promulgate clear, lawful criteria—proof of identity usage, professional aliases, ballot descriptors, or disambiguators—well before filing deadlines, and then enforce them consistently.
Some of the people all the time-
Judge dismisses DOJ lawsuit seeking to block Alaska Senate certification if GOP’s Dan Sullivan loses— Thomas Barrows (@ThomasB08190712) November 1, 2026
Consequences and durable takeaways for future cycles
First, secrecy is non-negotiable. Any remedy that requires identifying which voters selected a particular candidate for the purpose of subsequent law-enforcement inquiry is almost certainly out of bounds. The Alaska ruling makes that boundary explicit. Second, feasibility matters as much as doctrine. If a remedy cannot be executed without disrupting tabulation, breaching secrecy, or changing rules midstream, courts will not order it—particularly when early votes are already mixed and counting processes are underway.
Third, ranked-choice voting subtly changes the equities. Because voters can express ordered preferences, the system blunts the most acute version of the “spoiler by confusion” claim. That does not immunize elections from mischief, but it shifts the burden back to pre-election rules about candidate qualification and ballot presentation rather than post-hoc forensic sorting of ballots. Finally, the right time to police name-related confusion is before petitions are accepted and ballots are printed. States that anticipate the risk can craft neutral disambiguation tools—middle initials, occupations, or jurisdictionally consistent descriptors—and codify what evidence establishes a permissible ballot name. Courts will back administrators who enforce clear law; they will balk when asked to improvise after voters have already spoken.
Sources:
cbsnews.com, san.com, supreme.justia.com, ucs.org
