Who will win the Democratic nomination for President in 2028?

TL;DR

The most likely option is Other at 23%; among named individuals, Kamala Harris leads at 20%, narrowly ahead of Gavin Newsom at 19%. That bucket combines several credible candidates already making national moves rather than one person. The key fact is that eight of twelve non-incumbent Democratic nominees from 1956 through 2016 emerged as sustained polling leaders only in the election year, while South Carolina votes first on January 22, 2028.

Context

No major contender had filed a 2028 presidential candidacy by the September 7, 2026 cutoff; the FEC's 2028 list contained only minor or unknown candidates. This is still two forecasts in one: who enters, then who wins if they enter.

The field should change after the November 3, 2026 midterms. Josh Shapiro's Pennsylvania margin and coattails will test his electability case, while Jon Ossoff's Georgia result will determine whether any presidential path remains plausible; the latest high-quality state polls had Shapiro ahead 55% to 39% among 760 likely voters surveyed August 17–21, published August 25 and Ossoff ahead 52% to 43% among 1,060 likely voters surveyed July 13–16, published July 23.

Evidence

The historical prior is broad. Gallup's 2019 full 12-contest series found only four nominees who held a sustained national lead from the year before the election; eight emerged as sustained leaders during the election year. The complete series is below.

Harris fits part of the successful early-leader pattern: she is a former vice president and prior nominee (Associated Press). But her lead is not dominant, and she lost the 2024 general election (Associated Press). The base rate gives her an edge, not inevitability. It also argues for a wide field rather than a simple Harris-versus-Newsom race.

The current national data says Harris has a real lead, but also shows large design effects. The RealClearPolling average for fieldwork from July 19 through September 2 was Harris 24.5%, Newsom 13.5%, Buttigieg 12.8%, Ocasio-Cortez 11.0%, Ossoff 8.8%, Shapiro 6.5%, and Whitmer 2.5%. These are preference shares among varying registered- or likely-voter samples, not nomination probabilities.

A dash means omitted or not reported, not zero. Harris led six of these eight polls, yet her range was 8% to 42%. That spread is too large to treat sampling error as the main cause. Candidate menus, voter screens and survey modes are moving the result.

The calendar changes the meaning of those national numbers. The DNC's official order is South Carolina, Nevada, New Hampshire, New Mexico, Michigan and Virginia. The two newest public New Hampshire polls instead favor Buttigieg and AOC: Saint Anselm's August 17–18 poll of 733 likely Democratic-primary voters had Buttigieg 25%, AOC 17%, Kelly and Ossoff 12%, Newsom 7% and Harris 3%; the University of New Hampshire's July 15–20 poll had AOC 22% and Buttigieg 21%. No public candidate-choice poll was located for South Carolina in the major public primary-poll index, so Harris's first-state advantage is an inference from coalition and relationships, not a measured lead.

Harris has the best national coalition floor. In April she said she was considering another run, and at the National Action Network convention she drew the largest crowd and the only standing ovation among the prospects there (Associated Press). Newsom has the clearest revealed intent: in early September he held eight public events in South Carolina, and people close to him called a White House campaign the next thing (Associated Press). Buttigieg is also behaving like a likely entrant. His August South Carolina trip followed Nevada and targeted the Black-voter weakness that ended his 2020 campaign (Associated Press).

Shapiro is the strongest lower-polling listed candidate. His campaign had raised more than $50 million by June 2026, and his Pennsylvania operation had opened 27 offices and hired nearly 100 full-time organizers (campaign filing summary; Washington Post). AOC has a clear progressive lane and led one New Hampshire poll, but she has also left open a 2028 Senate challenge in New York (Axios). Ossoff has rising attention and a favorable reelection position, but he said he had zero interest in running and called the talk fantasy football (Atlanta Journal-Constitution). Whitmer led an April Michigan test at 20.9%, then said a 2028 run was not on her list before adding never say never (Detroit Regional Chamber; Michigan Public).

Other deserves 23%, not a token residual. Mark Kelly was at 6.8% in the current national average and 12% in the latest New Hampshire poll. Andy Beshear, JB Pritzker, Wes Moore and Cory Booker are also building donor, travel or party networks, while Kelly had more than $22 million in his federal campaign account in the July Reuters review (RealClearPolling; Reuters).

I combined four views: current polling with heavy early-cycle shrinkage, candidacy intent and organization, historical/open-field structure, and the audited mean of several independent specifications. The calculation was:

$$P_i = 0.30P_{i,\mathrm{poll}} + 0.30P_{i,\mathrm{intent}} + 0.20P_{i,\mathrm{history}} + 0.20P_{i,\mathrm{ensemble}}$$

The polling term measures current support and name recognition. The intent term uses entry odds, organization and calendar fit. The history term spreads late-emergence risk across both listed and unlisted candidates. The ensemble term limits dependence on any one set of subjective assumptions. Dropping any one component leaves Harris at 20%–21%, Newsom at 18%–19%, and Other at 22%–24%. The extra decimals in the probability object come from this ensemble arithmetic, not from measurement accuracy.

What's non-obvious

The obvious pro-Harris read overstates her 24.5% average. The obvious anti-Harris read also goes too far. Her lead is noisy, but former vice presidents and prior nominees are exactly the kind of high-profile Democrats who have sometimes held early leads. I therefore keep her first among individuals, while placing the aggregate Other option above her.

The late-emergence base rate does not all belong in Other. Shapiro, AOC and even Buttigieg could be the candidate who rises during 2027. Assigning the full historical dark-horse rate to unlisted names double counts uncertainty. The small FEC clues are also weaker than they look: AOC's House committee paid $7,500 to an Iowa data firm between April 7 and June 11, 2026, and Ossoff's Senate committee paid $34,800 to a Nevada fundraising consultant between January 8 and June 10, 2026, but those payments fit current-office work better than hidden presidential campaigns (AOC FEC committee; Ossoff FEC committee).

Uncertainties

• The biggest gap is South Carolina. The state votes first, but no public candidate-choice poll was located in the major public primary-poll index; private polling may exist (DNC calendar).
• Candidate entry remains unresolved. Harris, Newsom, Buttigieg, Shapiro and AOC could all move several percentage points when they make final decisions; Ossoff and Whitmer could move sharply only by reversing their public posture (AP on Harris; AJC on Ossoff; Michigan Public on Whitmer).
• The November 3 midterms can reorder the race. A large Shapiro win with Democratic coattails would raise him; an Ossoff loss would nearly close his path (Inquirer; AARP).
• Poll quality is uneven. Several releases report a large total sample while the nomination question uses a much smaller party subgroup, and candidate lists differ enough to produce Harris results from 8% to 42% in the same broad period (RealClearPolling).