The Saint Anselm College Survey Center polled 1,614 New Hampshire voters online on 24 and 25 June. Inside that sample sat 398 first district Democrats. The margin of error there is 4.9 points. Shaheen took 22 percent, Sullivan 15, and State Representative Heath Howard 9. Then came Carleigh Beriont at 3, Christian Urrutia at 3, and Sarah Chadzynski at 1.

And 47 percent were unsure.

That last figure is the story. Nine Democrats filed for this seat in June, and the poll tested only six of them. Chris Pappas gave up the seat in April 2025 to run for the Senate seat Jeanne Shaheen is leaving, so there is no incumbent and no clear heir. That is the setup that blows a primary wide open six weeks out.

Sullivan’s campaign calls the June numbers proof she is closing the gap. That claim does more work than the data allows. In January, a University of New Hampshire poll had her third, behind both Shaheen and Howard. She is second now. But two surveys from two different firms do not make a trend. Pollsters differ in method and sample for reasons that have nothing to do with voters changing their minds. Here is the narrower truth. She sits second in the newest poll, seven points back, and undecided voters outnumber the leader two to one.

Sullivan has run here before. In 2018 she went for this same open seat. She raised close to $2 million, more than anyone else in an eleven way field. EMILY’s List backed her. So did VoteVets. She had the Marine record, the Obama era post, and the biggest bank account in the race.

She lost to Chris Pappas.

Pappas had the state party behind him, including Jeanne Shaheen, Maggie Hassan, and Annie Kuster. He had spent years working up through New Hampshire politics. Sullivan had moved to the state only months before. An early campaign post called her a lifelong Granite Stater, which handed her rivals the carpetbagger attack free of charge.

For a site that argues veterans outperform, this case has to be answered rather than skipped. Sullivan in 2018 comes close to a clean test. Keep the money steady, along with the national endorsements and the service record. The party’s pick still won with room to spare. If service moved Democratic primary voters, it should have moved them then.

The strongest case against Sullivan is not about Sullivan. It is about what the veteran edge actually is, and where it works.

Most proof that veterans outperform comes from general elections. The effect shows up with independents and soft partisans. Those voters read a service record as a shortcut. It tells them a Democrat is sound on security and close to the middle. Democratic primary voters are not that crowd. They have already picked a party. So they choose among Democrats on ideology, on local roots, on who knocked their door, and on which surname they know. A Marine record answers none of that.

Then there is the seat itself. Democrats have held New Hampshire’s first district since 2019. It is not a toss up that needs a candidate who can pull Republican leaning independents. Whatever edge makes a veteran worth more in Michigan’s seventh district counts for less in a seat the party expects to keep.

Put those together and the honest read follows. The veteran profile is a general election asset. Here it has to survive a primary crowd it was never built for. Sullivan failed that same test once.

Still, the June poll holds findings worth sitting with. Among voters with a graduate or professional degree, Sullivan leads at 24 percent to Shaheen’s 18. In the Merrimack Valley, which takes in Manchester, she leads 20 to 16. Shaheen’s edge comes almost all from voters 65 and older. There she draws 33 percent to Sullivan’s 19.

These are small samples inside a small sample, so read them as texture rather than proof. Even so, they show a race that is not splitting on age or ideology alone. Sullivan is not losing to a left insurgent. Howard, the populist option, sits at 9 percent. He takes 23 percent of voters under 35 and almost nothing above that. Sullivan is losing to a known name among older voters. That is just what a Shaheen carries in New Hampshire.

The difference matters for reading the result in September. Losing to Howard would say something about ideology, while losing to Shaheen says something about who voters already know. And being known is the one thing a candidate can still build in six weeks.

One thing has changed since 2018. It does not show up anywhere in the match above.

The same June survey asked about the American military action in Iran. In the first district, 60 percent disapproved. Among Democrats statewide the figure hit 97 percent, with 90 percent strongly against. That is near total agreement among the voters Sullivan needs. It sits next to a related finding. A quarter of New Hampshire voters named elections and democracy the top issue facing the country, ahead of prices.

Being against the war does not separate these candidates. Every Democrat in the field opposes it. What separates them is who can make that stance land as more than a position paper. Sullivan served in Iraq as a Marine officer. Urrutia is a Guard captain who defends soldiers at courts martial. Shaheen, Howard, and Beriont argue from belief, which is fair and also open to anyone.

Does that gap turn into votes? Nobody knows yet, and that is the narrow question 8 September answers. Note what it is not. This is not the general election effect the site usually tracks, where a service record calms voters who might otherwise break Republican. It is a claim about standing inside a primary, on one issue, at a moment when that issue sits at the front of Democratic minds.

Nobody should treat this race as called. Seven points is a real lead, and it comes from the freshest numbers available, which makes Shaheen the favorite.

But the result will be readable either way, which is rarer than it sounds. Say Sullivan wins. She will have beaten a family name in a low turnout primary, with a war in the news and a service record only Urrutia can match. That would be proof the veteran profile does work inside a primary, not just in November. Say she loses to Shaheen by a similar margin. Then the 2018 pattern has repeated against a new opponent. And this site should say plainly that the profile looks like a general election tool only.

Either finding is useful. The second is less comfortable, so it belongs on the record now rather than after the fact. Alpha Democrat has argued that veteran candidates give Democrats a structural edge in swing seats. New Hampshire’s first district is not a swing seat in November, and the voters going to the polls on 8 September are not the voters that edge was measured on. So a loss here would not sink the thesis. A win would stretch it further than the evidence has reached before.

Six weeks. Seven points. Forty seven percent still deciding. Watch older voters in Rockingham and Strafford counties, because that is where Shaheen’s margin lives right now.