Who will win the Democratic nomination for President in 2028?

TL;DR

My forecast makes Kamala Harris the most likely individual nominee at 19%, followed by Gavin Newsom at 16% and Pete Buttigieg at 13%. Other is the largest option at 27%, because Mark Kelly, JB Pritzker, Andy Beshear, Wes Moore and a candidate who breaks out after the midterms together outweigh any single listed contender. Harris leads today, but no candidate controls the race; entry decisions and the South Carolina-first calendar matter more than another point or two in a hypothetical national poll.

Context

The field is still pre-declaration. None of the seven named contenders appears as a 2028 presidential candidate in the Federal Election Commission’s candidate records. The first sanctioned contest is South Carolina on January 22, 2028, followed by Nevada, New Hampshire, New Mexico, Michigan and Virginia, according to the official DNC calendar. South Carolina is 495 days from this forecast’s cutoff. (democrats.org)

National polls show a real Harris lead, but not a settled race. The current RealClearPolling average, based on fieldwork ending between July 19 and September 2, has Harris at 24.5%, Newsom at 13.5%, Buttigieg at 12.8%, Ocasio-Cortez at 11.0%, Ossoff at 8.8% and Mark Kelly at 6.8%. A newer BGSU/YouGov poll, fielded August 25–September 1 among a full sample of 1,200 likely voters, found a much flatter Democratic field: Harris 17%, Buttigieg 16%, Ocasio-Cortez and Newsom 14% each, and Kelly 10%. (realclearpolling.com)

Evidence

History says to discount early polling sharply. Gallup’s review of 12 non-incumbent Democratic nomination races from 1956 through 2016 found that only four eventual nominees led continuously from the year before the election; eight became sustained leaders only during the election year. At a horizon closer to today’s, Pew found that the early leader won only one of six open Democratic contests from 1960 through 1996. Harris’s lead matters, but these base rates rule out treating 24.5% polling support as a 24.5% nomination probability.

The officeholding reference class is unusually balanced. Across the ten open or successor contests from 1972 through 2020, the nominees included three governors, three sitting senators, three sitting or former vice presidents, and Hillary Clinton, a former senator and secretary of state. That history gives Harris a résumé advantage, but it also leaves clear routes for Newsom, Shapiro and the senators. Ocasio-Cortez and Buttigieg have less conventional profiles: no modern Democratic nominee has moved directly from the House, and no nominee has matched Buttigieg’s mayor-plus-Cabinet path. The nominee record comes from the American Presidency Project’s convention archive. (presidency.ucsb.edu)

The polling itself is noisy. In the eight component polls behind the recent national batch, Harris ranged from 8% to 42%, Buttigieg from 7% to 26%, and Ossoff from 4% to 21%; the ballot lists, voter screens and treatment of undecided voters differed. The latest BGSU result is valuable because it corroborates a competitive top four, but its published margin of error covers the full 1,200-person sample rather than a separately reported Democratic subgroup. I read the polls as evidence of Harris’s floor, Buttigieg’s resilience and AOC’s real constituency—not as a direct forecast. (realclearpolling.com)

The calendar is the largest structural change. South Carolina’s Democratic electorate can be more than half Black, and it now votes before Nevada and New Hampshire (Associated Press). That helps Harris and forces Buttigieg and Ocasio-Cortez to prove broader appeal sooner. It also reduces the value of New Hampshire, where the current polling average is Buttigieg 22%, Ocasio-Cortez 17%, Kelly 10.3%, Newsom 9% and Harris 5.3%. New Hampshire still matters, but it votes third rather than first. (realclearpolling.com)

Harris remains the strongest single candidate. She publicly said in April that she was considering another run (Reuters), and her national recognition and Black-voter support fit the opening calendar. I use roughly a 72% entry probability. Her liabilities are the 2024 defeat, weak New Hampshire numbers and visible reluctance among some former major donors to finance a repeat campaign (Los Angeles Times). She is the favorite conditional on running, but not close to inevitable. (latimes.com)

Newsom has the clearest campaign-in-waiting. His network raised more than $1.7 million for 14 state parties from over 74,000 donors and held more than $14 million across affiliated entities at the end of June (Axios). His September South Carolina trip included eight public events, private meetings and his third visit there during the midterm cycle; he also left open making a decision before year-end (Associated Press). I put his entry probability near 90%. His main problem is not intent. It is whether a California governor can build a durable Black, Southern and working-class coalition. (axios.com)

Buttigieg is likely to run and has the best verified New Hampshire position. He said on August 3 that he was “more inclined than not” to enter (RealClearPolitics), while preserving a national organization and returning to South Carolina to court the Black voters who rejected him in 2020 (Associated Press). I use an entry probability around 83%. The South Carolina-first sequence is why his nomination forecast remains below his organizational and New Hampshire strength. (realclearpolitics.com)

Ocasio-Cortez has the clearest ideological lane and could benefit if several mainstream candidates divide their voters. Her team is being reshaped before a 2028 decision, with experienced Sanders-linked advisers available, but she is also considering a New York Senate campaign (Axios). I estimate a 58% chance that she chooses the presidential race. Her New Hampshire strength is real; the missing evidence is comparable support among older Black voters, union households and less ideological Democrats in the states voting before or immediately after New Hampshire. (axios.com)

Shapiro is the strongest underpolled breakout candidate. An August 17–21 survey of 760 Pennsylvania likely voters had him leading his Republican opponent 55%–39% (Siena); a separate August 17–23 poll of 501 registered voters had him ahead 50%–25% (Franklin & Marshall). A large victory plus down-ballot gains would create a clean electability pitch from a presidential battleground. His current national support and early-state activity remain modest, so I keep him just below 10%. (sri.siena.edu)

Ossoff has the largest gap between upside and present intent. His Senate campaign entered the fall with more than $42 million after raising $20 million in the second quarter (Atlanta Journal-Constitution), giving him a nationalizable donor network if he wins reelection. But he has repeatedly said he has “zero interest” in running and called the speculation “fantasy football” (Atlanta Journal-Constitution). I treat those categorical denials as stronger evidence than ordinary campaign-year evasions and use only about a 17% entry probability. (ajc.com)

Whitmer preserved an option, but little more. She said a presidential campaign was “not on my list,” then shifted the same day to “never say never” (Michigan Public/WCMU). I use a roughly 35% entry probability, higher than a literal reading of her initial denial but far below Newsom or Buttigieg. Low national polling and the lack of a visible presidential build keep her final probability near 2%. (wcmu.org)

Other is not a rounding bucket. Mark Kelly is already at 6.8% in the national polling average, while JB Pritzker has launched a political group planning roughly $20 million in midterm spending ahead of a possible bid (Washington Post). Andy Beshear, Wes Moore, Cory Booker, Raphael Warnock, Chris Murphy, Elissa Slotkin and an unforeseen post-midterm breakout add several smaller paths. No one in this group is individually favored, but together they warrant the largest option probability. (realclearpolling.com)

I combined three normalized lenses:

$$P_i = 0.40C_i + 0.30O_i + 0.30H_i$$

Here, $C_i$ is current polling adjusted for the early calendar and coalition fit; $O_i$ is entry probability, fundraising and organization; and $H_i$ is the historical reference class, including the chance of a late or unlisted nominee. I then made small candidate-specific adjustments for direct statements: down for Ossoff, up for Newsom and Buttigieg, and up for Other because the option contains several credible candidates. The resulting precision is bookkeeping rather than a claim that the race can be measured to a tenth of a point.

What's non-obvious

The apparent sample sizes in early presidential polling often overstate how much direct evidence exists. Rasmussen’s August poll advertised 1,222 likely voters, but its Democratic nomination question was asked of only 504 Democratic-primary-inclined respondents (questionnaire). BGSU reports a full sample of 1,200 likely voters but does not state the Democratic subgroup size on its main release. The national aggregate is useful, but less precise than its large headline samples suggest. (bgsu.edu)

The strongest intent signal is not an announcement. It is costly preparation. Newsom’s state-party fundraising and travel, Buttigieg’s choice to skip Michigan’s major 2026 offices, and Shapiro’s effort to produce Pennsylvania coattails reveal more than vague interview answers. The reverse is true for Ossoff: his money and media attention are well explained by a competitive Senate race, while his repeated denial is unusually direct. That is why I place Ossoff well below his current 8.8% national polling share and keep the broad Other bucket above every named candidate. (axios.com)

Uncertainties

• The largest data gap is the lack of current, well-documented statewide primary polling in South Carolina, Nevada, New Mexico and Virginia. Polls with large Democratic-primary samples and demographic crosstabs would materially sharpen the calendar adjustment. The absence of such data is not evidence that any candidate is weak there.
• Candidate entry dominates the forecast. Clear decisions from Harris, AOC, Ossoff or Whitmer would release large and distinct voter blocs rather than shifting support evenly to the remaining field.
• Shapiro’s and Ossoff’s midterm results are major gates. A decisive Pennsylvania victory with Democratic coattails would lift Shapiro; an emphatic Georgia win followed by softer presidential language would lift Ossoff. Their current launchpads are strong, but the elections have not occurred. (Pennsylvania polling; Ossoff fundraising). (sri.siena.edu)
• The campaign environment can still change completely through economic shocks, wars, scandals, health events, endorsements and debate performance. With the first votes more than a year away, these unknowns justify keeping every named candidate below 20%.