Democrats’ NAME Trick Could Flip the Senate

Voter fills out ballot at a polling booth — Photo: Rob Crandall / Shutterstock

Ballot engineering thrives in the gray space between hard law and hardball: in Alaska’s 2026 Senate race, published reporting depicts a deliberate recruitment of a same‑name candidate—Daniel J. Sullivan—to alter how votes would be cast and counted under ranked‑choice rules.

At a Glance

Multiple outlets reported a coordinated effort by aides linked to Democrat Mary Peltola and DSCC staff to recruit Daniel J. Sullivan to run against incumbent Republican Sen. Dan S. Sullivan, with interviews, texts, and draft legal documents cited as evidence.
The project reportedly carried an internal label—the “ballot composition project”—and aimed to draw votes from the incumbent via a same‑name candidacy.
Daniel J. Sullivan, a retired teacher from Petersburg, entered the race and qualified for Alaska’s four‑candidate ranked‑choice general election.
Peltola’s campaign and the DSCC publicly denied recruiting any candidate; a former Peltola campaign manager said Peltola was “definitely aware,” while not personally directing the effort, according to reporting.

What the reporting establishes: a coordinated recruitment to shape the ballot

The core facts, as described by multiple outlets, are straightforward. Based on interviews with a dozen individuals, text messages, and documents prepared with litigation in mind, the New York Times’ findings—summarized by CNBC—reported that operatives tied to Rep. Mary Peltola’s Senate bid and senior staff at the Democratic Senatorial Campaign Committee discussed and participated in recruiting a candidate named Daniel J. Sullivan to run in Alaska’s Senate race. Good Morning America reported the organizers referred to this as a “ballot composition project,” a term of art in campaign circles for strategically adjusting the field of candidates to optimize downstream vote dynamics.

Two elements matter here. First, the intent attributed to the project was explicit: find a spoiler capable of siphoning first‑choice votes from Republican incumbent Dan S. Sullivan. Second, the execution keyed on identical name recognition. A retired schoolteacher from Petersburg, Daniel J. Sullivan, entered the race following these efforts and ultimately advanced into Alaska’s four‑way ranked‑choice general election—precisely the condition under which any name‑based confusion could move real votes before transfers.

How the tactic intersects with Alaska’s ranked‑choice voting

Recruiting spoilers is not new; what’s different is the terrain. Alaska uses a nonpartisan top‑four primary followed by ranked‑choice tabulation. In a classic plurality system, a same‑name entrant can divert enough votes to determine a winner without majority consent. Ranked‑choice voting (RCV) is meant to blunt that spoiler effect by redistributing ballots from eliminated candidates to voters’ next preferences in successive rounds. But the protection is not absolute. If a same‑name candidate survives into later rounds or if ballots exhaust—when voters rank too few choices and their ballots drop out of later counts—strategic first‑choice siphoning can still matter before transfers wash it out. Experienced operatives understand that the first round sets the table; shift those first‑choices and you can reshape the path of later reallocations.

That calculation explains the reported logic of the “ballot composition project.” It sought to alter the composition of first‑choice votes cast for “Dan Sullivan,” banking on a subset of voters—through inattention, miscue, or protest voting—to select the non‑incumbent Sullivan as their first mark. Even if many such voters later ranked the Republican incumbent second, the intervening arithmetic of eliminations and ballot exhaustion can produce non‑intuitive outcomes in close contests. Campaigns test for these edges because small percentages can be decisive under RCV’s round‑by‑round math.

Who did what: roles, denials, and the evidentiary spine

The reporting attributes participation to aides connected with Peltola’s operation and to DSCC personnel, and it anchors those claims to documentary materials and interviews—twelve sources, texts, and litigation‑drafted documents, per the CNBC account of the Times’ work. Good Morning America’s description of an internal name for the project adds specificity consistent with an organized, not ad‑hoc, undertaking. The Washington Examiner and others identified Daniel J. Sullivan as a retired teacher from Petersburg and placed his entry after the recruitment push. CNN’s analysis noted he cleared the primary threshold, formally positioning him in the general election where the tactic could operate.

Public pushback came swiftly. Alaskans for Mary denied involvement “with either Sullivan campaign,” and the DSCC likewise said it was not involved with Dan J. Sullivan. Reporting also conveyed a former Peltola campaign manager’s statement that Peltola did not direct the project but was “definitely aware” of it—an acknowledgement that, while insulating the candidate from operational control, situates the activity within the campaign’s horizon of knowledge. In practice, that distinction matters for ethics, optics, and potential compliance questions, but it does not erase the underlying strategic behavior the documents and interviews describe.

Mechanism and intent: why same‑name recruitment can work

Same‑name candidacies exploit cognitive shortcuts. Voters under time pressure often scan for familiar names and party cues; when two candidates share a name, the risk of a selection error rises, especially in down‑ballot or multi‑candidate contexts. Alaska’s ballots identify candidates by full name and party, and in the 2026 Senate race the incumbent carried both the “incumbent” status and Republican label; Daniel J. Sullivan appeared as a distinct entry with his own identifiers. Careful voters are unlikely to err. But campaigns typically need only a sliver—one or two percentage points of misallocated first‑choices—to reshape early rounds. When combined with ballot exhaustion and uneven ranking behavior across regions, this sliver can change who faces whom in the decisive instant‑runoff round.

That is why a “ballot composition project” is an apt moniker. You are not persuading mass opinion; you are composing the menu and nudging the sequencing of votes to maximize a probabilistic edge. In pluralities, that edge is direct. In RCV, it is subtler and depends on survival thresholds, drop‑off rates, and the distribution of second and third choices. Still, as Alaska’s adoption of top‑four RCV has matured, partisans on both sides have experimented with candidate‑field shaping precisely because it is one of the few remaining levers that can be pulled before persuasion and turnout do the rest.

Why this episode matters beyond Alaska

The reported Alaska recruitment illustrates three durable truths about modern campaigns. First, field‑shaping is now a standard competency, not an aberration; parties and aligned groups map out hypothetical slates months in advance and stress‑test the math. Second, RCV changes incentives but does not abolish them. The spoiler problem is mitigated but not cured, leaving openings for tactics that exploit voter confusion or low‑information behavior in the first round. Third, the line between legal hardball and unlawful deception is fact‑bound and context‑specific. Public denials will persist, but the documentary trail reported—texts and lawsuit‑ready drafts—signals that operatives are formalizing these strategies in ways that leave marks.

For election administrators and reformers, the response is not rhetorical but mechanical. Clearer ballot design, voter education that emphasizes full ranking, and uniform presentation of incumbency and party cues reduce the yield from same‑name gambits. For campaigns, the lesson is simpler: if one side can change the composition, so can the other. And for voters, the countermeasure is to slow down and rank thoroughly—because RCV’s safeguards activate only when ballots carry complete preferences.

Bottom line

Based on multi‑source reporting, aides tied to Peltola’s campaign and DSCC staff engaged in a coordinated effort—their “ballot composition project”—to recruit a same‑name contender, Daniel J. Sullivan, into Alaska’s 2026 Senate race, aiming to redirect first‑choice votes from the Republican incumbent. Peltola’s campaign and the DSCC denied recruiting anyone, while separate reporting quoted a former campaign manager saying Peltola was aware but not directing the operation. In an RCV system built to temper spoilers, candidate‑field engineering endures, and identical‑name tactics remain one of the few levers that can still move margins at the edge of close contests.