Who will win the Democratic nomination for President in 2028?

TL;DR

Kamala Harris is my most likely individual nominee at 21%, narrowly ahead of Gavin Newsom at 19%. Other is the largest category at 23% because several credible governors and senators are not listed. Harris's national polling and South Carolina path narrowly outweigh Newsom's stronger evidence of intent.

Context

As of August 1, 2026, no major candidate has declared, and the visible field remains a contest among former national candidates, governors, senators, and several plausible late entrants (Reuters). The question resolves to the person formally nominated, so a candidate who leads the primaries but withdraws before nomination would not count.

The DNC Rules and Bylaws Committee has proposed South Carolina for January 22, 2028, followed by Nevada, New Hampshire, New Mexico, Michigan, and Virginia; the full DNC must still approve the calendar in August (DNC proposal, proposed dates). South Carolina first raises the value of support among Black Democrats and reduces the value of building initial momentum in New Hampshire alone.

Evidence

The historical record says early polling identifies viable candidates better than winners. In six open Democratic contests between 1960 and 1992, the early national leader won only once (Pew Research Center). A broader study of every competitive nomination from 1972 through 2016 found that candidates polling between 20% and 35% in the first half of the year before voting won 36% of nominations, while those between 10% and 20% won 9%; only one of 84 highly recognized candidates polling below 10% eventually won (FiveThirtyEight historical analysis). The current snapshot is earlier than that study's main window, so I shrink today's differences further.

Across eight national polls published from July 6 through July 28, an unadjusted equal-weight average gives Harris about 28%, Newsom 17%, Buttigieg 12%, Ocasio-Cortez 9%, and Shapiro 6% (RealClearPolling poll list). The useful signal is the tiering, not the exact shares. Harris ranged from 8% to 37% across those polls, far beyond ordinary sampling error. Ballot construction is a major cause: Emerson newly added Ossoff and found Buttigieg at 19%, Newsom at 17%, Ossoff and Ocasio-Cortez at 13%, and Harris at 8%; its overall sample was 1,100 likely 2026 voters, while the Democratic subset was smaller and its effective sample size was not reported in the release (Emerson College).

The proposed calendar gives Harris a real edge. A February Harvard-Harris survey using a restricted candidate list placed her at 59% among Black respondents, versus 15% for Newsom; the full survey covered 2,000 registered voters, but the smaller relevant subgroup was not separately reported (Harvard-Harris results). New Hampshire shows the opposite coalition: a July survey of 648 likely Democratic-primary voters put Ocasio-Cortez at 22% and Buttigieg at 21%, while Newsom had 8% and Harris 5% (UNH Granite State Poll). This is why I treat Harris as the narrow favorite rather than a dominant one.

Entry signals favor Newsom and Buttigieg. Newsom is term-limited, has built a national political operation, traveled to early states, published a memoir, and openly acknowledged that he is considering a run; Buttigieg has more than $5 million in his PAC and has traveled to Iowa and New Hampshire (Reuters candidate review). Buttigieg also headlined the Iowa Democratic Party's July 12 fundraiser, but South Carolina first directly exposes the Black-voter weakness that ended his 2020 path (Iowa Public Radio). Newsom's federal investigation adds a wide tail: it could damage him, fade away, or strengthen his image as Trump's principal Democratic antagonist (Reuters).

Shapiro has the strongest conventional dark-horse case. His reelection campaign reported raising $50 million from January 2025 through June 2026, including $12 million during the second quarter (Philadelphia Inquirer). Pennsylvania polling has him well ahead of Republican Stacy Garrity, creating a plausible post-midterm argument that he can win and govern in a decisive battleground (Franklin & Marshall poll coverage). His present national support is modest, and his positions on Israel create friction with part of the primary electorate, so I keep him near Buttigieg rather than in the top pair (Reuters candidate review).

Ocasio-Cortez has a distinct progressive lane, strong small-dollar potential, a New Hampshire lead, and a growing record of national travel and successful endorsements (Axios campaign activity, Axios endorsements). Her main discount is entry: a 2028 New York Senate campaign remains a credible alternative. If she chooses the presidential race, her conditional chance is higher than her 10% unconditional forecast suggests.

Ossoff is the largest downward adjustment from recent attention. A July 13–16 survey of 1,060 Georgia likely voters put him ahead in his Senate race, 52% to 43%, and his new inclusion in Emerson's national ballot immediately produced 13% (AARP Georgia survey, Emerson College). But he has repeatedly said he has zero interest in running, calls the discussion a distraction, and has no visible presidential PAC or early-state organization (Atlanta Journal-Constitution). I allow for a reversal after a strong reelection, but not enough to put him near AOC or Shapiro.

Whitmer's résumé is stronger than her probability. She said on May 28 that a presidential run was not on her list and that she would not be among the candidates, then softened this to never say never later that day (Michigan Public). With little visible preparation and low-single-digit national polling, most of her probability comes from a reconsideration scenario.

Other remains the largest bucket. It contains Mark Kelly, Andy Beshear, JB Pritzker, Wes Moore, Raphael Warnock, Cory Booker, Ro Khanna, Ruben Gallego, Chris Murphy, Rahm Emanuel, and possible post-midterm breakout candidates; several are already traveling, fundraising, or leaving the door open (Reuters candidate review). No one in that group is individually as likely as Harris or Newsom, but their combined path is substantial.

I modeled each candidate as roughly proportional to

$$P(\text{serious campaign}) \times S(\text{nomination strength if running}),$$

where strength combines polling, organization, coalition fit, early-state path, and room to grow. I then shrank the polling component toward the historical prior and reserved a separate probability for the unlisted field. That method favors Harris on coalition and polling, Newsom on entry and organization, and Shapiro on upside after the midterms.

What's non-obvious

Harris and Ossoff generate opposite forecasting errors. Harris is easy to underrate because she is familiar, lost in 2024, and faces donor demand for a new face; she still has the best current combination of national support and a favorable first contest (Associated Press). Ossoff is easy to overrate because one newly constructed poll and a strong Senate campaign created sudden national attention, while the harder evidence—intent, staff, PAC activity, and early-state travel—still points away from a presidential campaign.

The national leader and the likeliest runner are different people. Harris leads most polls, but Newsom has made more costly preparations and has fewer attractive alternatives. That distinction is why the gap between them is less than two percentage points in my forecast.

Uncertainties

• The calendar is still provisional. Full DNC approval, New Hampshire's response to being placed third, and any later rule changes could alter the value of the early states (DNC).
• There is no high-quality public South Carolina candidate poll, and Nevada's major May survey examined voter characteristics rather than a candidate horse race (Third Way). State polling from those two contests would materially improve the estimate.
• The November 3, 2026 elections are major gating events. Shapiro's Pennsylvania margin and Ossoff's Georgia result will test their electability arguments, fundraising capacity, and national demand.
• Endorsements, early-state staff hires, donor commitments, and 2027 fundraising reports remain mostly unobserved. Those signals usually carry more information than hypothetical polling during the invisible primary (University of Chicago Press).