Published: July 6, 2026

On July 1, the New England Submarine Shipbuilding Partnership passed 10,000 workers trained.  SENEDIA, the intermediary that runs it, carries a $98.3 million contract to train 8,600 more by 2029, and it crosses the milestone under new leadership: Stephen Guertin took over as executive director in February from longtime CEO Molly Donohue Magee, who built the alliance into what it is.  Down in Camden, Arkansas, Lockheed Martin is running its Camden Manufacturing Apprentice Program with Southern Arkansas University Tech while it builds out a munitions campus.  Within a year, every workforce board in a defense manufacturing state will probably be told to copy one of these models.  Some copies will work; most will not.  What will separate them?

Whether the copiers reach for the visible part or the invisible one.  What travels in the case study is the visible part: the curriculum, the training center, the governance chart, the ribbon cutting.  What made these programs work is mostly invisible, and some of it cannot be packed in a box.

Start with what the New England model actually is.  It is a demand signal with a school attached.  The Navy awarded Electric Boat a $15.38 billion contract modification in March for Columbia- and Virginia-class work running to 2035, and the yard committed to 8,000 hires this year; the training pipeline exists because a named employer with a named contract needs named trades in named quantities.  Camden is the same story at different scale.  Lockheed, General Dynamics, and a Rafael-Raytheon joint venture are all expanding within one county, and the apprenticeships map to specific production lines.

That suggests a rule.  Three things in these models transfer: an accountable intermediary that owns the pipeline rather than coordinating it to death; training tied to a specific contract and start date rather than to general labor market projections; and employers who co-design the curriculum because they will eat the consequences of a bad one.  Two things do not transfer: anchor demand at multi-billion-dollar scale, and the accident of geography that put a submarine yard or a munitions cluster in your region seventy years ago.

The obvious objection: is this an argument for giving up?  If your region lacks an Electric Boat, are you finished?  No.  It is an argument for sequencing.  The failed version of replication builds the training center first and prays demand shows up.  (There is no limit to what you can dream when you are not responsible for the results.)  The working version secures the demand commitment first, in writing, with numbers, and sizes the training to match.  If the employers in your region will not commit to hires, that is not a reason to build a smaller training center.  It is the answer to whether you should build one.

So before your region copies SENEDIA, ask the operator’s questions.  Which employer, which contract, how many hires, by when?  Who owns the pipeline, and what happens to them if it fails?  Do the wages at the end of the training clear the local market, or are we training people for jobs they will leave in a year?  (Pascagoula is running that experiment for all of us right now.)

The model is not the curriculum.  The model is the customer.

---