Independent New Hampshire Journalism · Northfield, NHNew college graduates now face a higher jobless rate than the workforce as a whole, AI is thinning entry-level hiring, and adults under 30 trust their feeds about as much as they trust newsrooms. Granite State Report is recruiting the people the squeeze hits first to help build the fix.
Full disclosure: this is a recruiting call. I founded and edit Granite State Report, and the people this piece describes are the people it asks to join.
In June, 5.7 percent of Americans ages 22 to 27 with a bachelor’s degree were out of work and looking. Among all workers ages 16 to 65, the rate was 4.1 percent. Those are the Federal Reserve Bank of New York’s numbers, built from the government’s monthly jobs survey, and they say something a diploma was supposed to prevent. The newest graduates are now more likely to be jobless than the workforce as a whole, and that has been true every single month since January 2021.
For most of the past 36 years it ran the other way. In the New York Fed’s series, which starts in 1990, a fresh degree usually bought you a lower jobless rate than the national average. The gap has flipped, and six of the seven widest gaps on record in the wrong direction have come since March 2025.
Here is where I stand. The people getting shut out of the first rung of work are the same people the old news business stopped talking to, and they are the people best equipped to build what replaces it. Granite State Report is looking for some of them.
Related coverage. This piece is part of GSR’s Generations Series. New Hampshire Needs Young Workers. It Built a State They Can’t Afford. priced the front door for young adults here, and Trump Said the American Dream Was Dead. For Children Born the Year He Was, the Odds Were 86 Percent. ran the mobility record by birth year. Outside the series, They Can’t Count the Rules. We’ll Count Who Answers. laid out the verified Granite Stater plan. This piece adds 2026 graduate jobless data, Stanford’s payroll study on AI and young workers, and new data on where young adults get news. Its housing figures are newer than the July piece’s, so the numbers differ by design.
The first rung is missing
The jobless rate is only half of it. The New York Fed also tracks underemployment, the share of graduates working jobs that typically do not require a degree. For recent graduates it stood at 42 percent in the second quarter of 2026. Four in ten recent graduates with jobs hold ones that never needed the degree.
The labor market as a whole does not look broken, which is why this gets missed. The Bureau of Labor Statistics put the national unemployment rate at 4.2 percent in September and the rate for college graduates 25 and older at 2.5 percent. A degree still pays once you are inside. The trouble is getting inside.
New Hampshire’s headline number looks even better than the nation’s. The state’s unemployment rate was 2.8 percent in August, according to federal data published by the St. Louis Fed. A low rate does not make a state easy to start out in. The median single-family home sold for a record $580,000 in July, according to the New Hampshire Association of Realtors, as reported by the New Hampshire Bulletin. New Hampshire Housing puts the statewide median rent at $2,011 a month, which takes $80,440 a year to afford, according to NHPR’s account of the agency’s numbers.
This is a state with a median age of 43.4 in 2023, one of the three oldest in the country, per the New Hampshire Fiscal Policy Institute’s reading of Census estimates. It needs young workers. It charges them full price at the door.
What AI is doing, and what it isn’t
You will hear that AI is about to replace everybody. The best payroll evidence so far says something narrower, and for young people it is worse in a specific way.
In August, Stanford economists Erik Brynjolfsson, Bharat Chandar and Ruyu Chen updated a study built on ADP payroll records covering millions of U.S. workers through June 2026. Their first finding is that there is no evidence of widespread, economy-wide job displacement. Their second is the one that should keep a 23-year-old up at night. Employment of workers ages 22 to 25 in AI-exposed occupations now sits 19 percent below where it would be had it kept pace with less-exposed peers, and experienced workers in the same jobs show no comparable gap.
In raw terms, employment of 22- to 25-year-olds in the two most AI-exposed fifths of occupations fell about 11 percent between November 2022 and June 2026. The same age group in the three least-exposed fifths grew about 10 percent. The gap has widened since the authors first reported it in August 2025, and it comes mostly from companies hiring fewer young people, not firing more of them. Nobody gets a layoff notice. The door just doesn’t open.
The authors are careful about what that proves. They call these early descriptive indicators, not causal estimates, and they report the pattern weakens when they control for education. Still, the direction is plain enough to plan around. And one finding points somewhere useful. Where AI mostly substitutes for human tasks, young employment fell. Where it mostly helps people do their jobs, employment held flat or rose.
That is the opening. Work that uses AI as a tool and keeps a person on the judgment is the work that is holding up. Checking a claim against the record, or deciding what a town needs to know: AI can speed up that work, and it cannot own it.
Where the news went
The same generation has moved its news off the old platforms. Pew Research Center reported last December that 76 percent of adults under 30 get news from social media at least sometimes, against 28 percent of those 65 and older. Forty-three percent of under-30s regularly get news on TikTok, and 38 percent regularly get it from news influencers. In a September 2025 survey, 51 percent of 18- to 29-year-olds said they have at least some trust in information from national news organizations, and 50 percent said the same of social media. For every older group, those two numbers sit far apart. For the youngest adults, they are one point apart.
The Reuters Institute at Oxford found the same shift across 18 countries. Among 18- to 24-year-olds, social media was the main source of news for 39 percent in 2025, up from 21 percent in 2015. News websites and apps fell from 36 percent to 24. Forty-two percent of that age group say they sometimes or often avoid the news. Young avoiders are more likely than older ones to say the news has little to do with their lives, 21 percent against 16, and far more likely to say it is hard to follow, 15 percent against 5.
I don’t read that as a generation that stopped caring about facts. I read it as one that got tired of news written for someone else. Meanwhile, the local newsrooms that might have spoken to them kept closing. Northwestern’s Medill School counted 213 U.S. counties it classifies as news deserts in its 2025 report, and about 50 million Americans with limited or no access to local news.
What GSR is, and what it is not
Granite State Report is an independent New Hampshire newsroom I founded in 2025 and run out of Northfield. Every piece carries a fact-check table that tells you which claims were verified against a primary source and which can only be attributed. We link the source so you can check it yourself. When we get something wrong, we fix it on the page and say what changed.
I have called this project remaking government in the form of a business. GSR is not a government and should never act like one. A newsroom that wants to hold power to account cannot also want to hold power. What GSR can be is a disciplined business in service of self-government: pull the records, publish the numbers, show the work, and let Granite Staters decide. That is what “we the people” looks like on an ordinary Tuesday.
On Oct. 14, GSR announced that only verified Granite Staters, adults confirmed by a code mailed to a New Hampshire address, will take part in its political side: public questions to legislators, bill polls and a shared log of records requests. It needs builders.
Who I’m looking for
I’m writing this to Gen Z and millennials because the numbers above land on them first. If you’re older and this describes you, write anyway.
I want people who put the truth ahead of their own side, who would rather publish a correction than defend a mistake, and who can show their work. Your education can come from a classroom or from your own reading. I care what you can do with it. If you’re between jobs, underemployed or stuck in work that wastes you, that is not a mark against you. It may be why you see the problem so clearly.
Here is the work:
- Builders who code, automate and know AI tools well enough to know where they fail. GSR runs on free tools and homemade software, and there is more to build.
- Researchers and fact-checkers who read the primary source before the summary.
- Records people willing to learn RSA 91-A and file requests.
- Video and social producers who can make a long investigation travel without bending it.
- Reporters who live here and know a town or a beat.
About money, straight: tell me what you need to make this work, and I will tell you plainly what GSR can offer right now. I won’t promise what I can’t deliver.
How to raise your hand
Email granitestatereport@gmail.com with “I’M IN” in the subject line. Tell me where in New Hampshire you live and what you’re good at, and send one link to something you made. Then pick one claim you’ve seen online about New Hampshire, check it, and tell me in a few sentences what the record says and how you found it.
That last part is the application. It is also the job.
New Hampshire’s motto is about refusing to live on someone else’s terms. The generation handed the worst entry terms in decades also holds the best tools anyone has ever had to check the record and publish it. I would rather build with that generation than keep writing about it. If that’s you, the inbox is open.
— Dexter Dow, Granite State Report
Your Turn
Poll: What would make you trust a news source?
A) It links its sources · B) It corrects itself in public · C) People I know work there · D) Nothing would
Poll: What is the biggest thing pushing young Granite Staters out?
A) Housing costs · B) No entry-level jobs · C) Childcare and cost of living · D) Nothing for young people to do
You tell me: Check one claim about New Hampshire you saw online this week and send me what the record says. granitestatereport@gmail.com
Fact check
| # | Claim | Status | Source |
|---|---|---|---|
| 1 | Unemployment, June 2026: 5.7% for recent college graduates (22–27), 4.1% for all workers (16–65) | VERIFIED | NY Fed unemployment data file (seasonally adjusted, 3-month moving average), released Aug. 6, 2026: 5.698 and 4.061. The Fed’s quarterly summary rounds the Q2 graduate rate to “about 5.6 percent.” |
| 2 | Recent-graduate rate above the all-worker rate every month since January 2021 | VERIFIED | GSR computation on the NY Fed monthly series, 1990–June 2026; last month at or below was December 2020 |
| 3 | Six of the seven widest graduate-over-all-worker gaps since 1990 came since March 2025 | VERIFIED | GSR computation on the NY Fed series; top seven: Apr. 2025, Mar. 2025, Sept. 2025, June 2026, Oct. 2025, Feb. 2026, Aug. 2020 |
| 4 | Recent-graduate underemployment 42% in Q2 2026 | VERIFIED | NY Fed, The Labor Market for Recent College Graduates (page and data file) |
| 5 | U.S. unemployment 4.2% in September 2026; 2.5% for bachelor’s degree holders 25+ | VERIFIED | BLS, The Employment Situation, Oct. 2, 2026 |
| 6 | NH unemployment 2.8% in August 2026 | VERIFIED | BLS state data via St. Louis Fed ALFRED, series NHUR, vintage Sept. 18, 2026 |
| 7 | NH median single-family home price a record $580,000 in July 2026 | ATTRIBUTED | NH Association of Realtors, as reported by the New Hampshire Bulletin; NHAR release not opened directly |
| 8 | NH median rent $2,011; $80,440 income needed to afford it | ATTRIBUTED | New Hampshire Housing figures as reported by NHPR, Oct. 7, 2026; agency report not opened directly |
| 9 | NH median age 43.4 in 2023, among the three oldest states | ATTRIBUTED | NH Fiscal Policy Institute, citing Census Bureau 2020–2023 estimates (tied with Vermont for second, behind Maine) |
| 10 | Stanford: no economy-wide displacement; 22–25-year-olds in AI-exposed jobs 19% below kept-pace level; about −11% vs. +10% from Nov. 2022 to June 2026; driven by reduced hiring; flat or rising where AI complements | VERIFIED | Brynjolfsson, Chandar and Chen, “Canaries in the Coal Mine?” Stanford Digital Economy Lab, August 2026 PDF |
| 11 | Pew: 76% of under-30s get news from social media at least sometimes (28% of 65+); 43% TikTok; 38% news influencers; trust 51% national news vs. 50% social media | VERIFIED | Pew Research Center, “Young Adults and the Future of News,” Dec. 3, 2025 |
| 12 | Reuters Institute: 18–24 main source social media 39% (2025) vs. 21% (2015); websites/apps 24% vs. 36%; 42% avoid news; 21% vs. 16% low relevance; 15% vs. 5% hard to follow | VERIFIED | Reuters Institute, “Understanding Young News Audiences at a Time of Rapid Change,” March 2026 |
| 13 | 213 news-desert counties; about 50 million Americans with limited or no local news | VERIFIED | Medill State of Local News 2025, Northwestern University, Oct. 20, 2025 |
| 14 | RSA 91-A:4, I gives every citizen the right to inspect and copy governmental records | VERIFIED | Statute text, NH General Court |
| 15 | GSR announced Oct. 14, 2026 that only verified Granite Staters (mailed code to an NH address) will use its political features | VERIFIED | GSR, “They Can’t Count the Rules. We’ll Count Who Answers.,” Oct. 14, 2026 |
Reach the editor directly — confidentiality respected where possible.
granitestatereport@gmail.com · 603-931-9264
Editor’s note. Every factual claim above was checked against the sources listed; see the fact-check table. GSR computed the graduate-versus-all-worker comparisons directly from the New York Fed’s published data file, which uses seasonally adjusted three-month moving averages and excludes people enrolled in school; the Fed’s own quarterly summary rounds the second-quarter graduate rate to about 5.6 percent, while the June monthly value used here is 5.7. The Stanford study is descriptive, not causal, and its authors say so. Housing and median-age figures are attributed to the organizations that published them. Photographs are by Carol M. Highsmith from the Library of Congress collection, which lists no known restrictions on publication. The narration is a computer-generated voice reading this text. This piece is also a recruiting call from GSR’s editor, as disclosed at the top. Corrections: Granite State Report corrects verified errors promptly and appends a note identifying what changed and when.
