Hudson Tunnel Project: $16 Billion Dig Finally Starts

17 min read
2 views
Oct 3, 2026

A tunnel opened in 1910 still carries hundreds of thousands of riders under the Hudson every day. This week the boring machines finally cut into the Palisades. The bill is $16 billion, the finish line is 2035, and the funding fight is not fully settled.

Financial market analysis from 03/10/2026. Market conditions may have changed since publication.

I kept thinking about the date on the old tube while officials posed for photos in hard hats. Nineteen ten. A year when most people still treated a river crossing as an event, not a Tuesday. This week the first of two giant boring machines began the slow bite into the Palisades, the opening act of a rail job priced around $16 billion and aimed at New York Penn Station. If you ride that corridor, you already know the punchline. The new path is not ready. The old one still is.

Hundreds of thousands of people move between New Jersey and Manhattan on those tracks on a normal weekday. Amtrak shares the bore with NJ Transit. One hiccup in a century-old tunnel and the whole evening rearranges itself. That is the dull, expensive problem this dig is supposed to fix. It is also why the ceremony felt less like a ribbon-cutting and more like a sigh of relief that had been waiting a very long time.

Why a 1910 Tunnel Still Runs the Show

The existing North River Tunnel is a two-track crossing. That sounds technical until you sit on a stalled train and watch the clock. Two tracks in, two tracks out, shared by intercity service and a commuter railroad that treats rush hour as a contact sport. There is almost no slack. A signal fault, a disabled car, salt damage after a storm, and the queue forms on both sides of the river.

I have found that people outside the region hear “$16 billion tunnel” and picture luxury. Riders hear something closer to plumbing. You do not applaud a new drain because it is glamorous. You applaud it because the old one keeps backing up. The Palisades dig is the first hard proof that the replacement is no longer only a stack of agreements.

A Bottleneck Dressed Up as a Landmark

The tube under the Hudson was an engineering feat when it opened. It still is, in the way an old bridge is a feat right up until the day the inspection report gets blunt. Age, water, and relentless traffic have turned a marvel into a constraint. Superstorm Sandy drove salt water into the system years ago. Repairs since then have been constant, partial, and scheduled around the fact that you cannot simply close the front door of the Northeast.

Perhaps the most interesting part is how normal the abnormality became. Delays got folded into commute math. People added buffer time the way other cities add a transfer. That adaptation hides the cost. Missed meetings, late childcare pickups, crews held at platforms, freight of a human kind that never quite arrives on the timetable printed in the app.

A region can absorb a bad Monday. It cannot absorb a permanent single point of failure and still pretend the economy is running on schedule.

The project adds two new tunnels for Amtrak and NJ Transit. Once the new bores are in service, the plan is to rehabilitate the old North River Tunnel so the corridor ends with four rail lines in and out of Penn Station. Four is not a magic number. It is room. Room to take a track out of service without strangling the evening. Room to recover from a failure instead of broadcasting it down the line to Newark.

What Commuters Actually Feel

Reliability is the product. Speed is the brochure. A few minutes saved on a clear day matters less than a week that does not collapse because one pump failed at the wrong hour. In my experience, riders forgive a slow train more easily than they forgive a train that vanishes from the board with no honest explanation.

The current setup piles roles onto the same pair of tracks. Long-distance trains, airport connections of a sort, reverse-peak trips, the ordinary crush into midtown. When everything shares one throat, small problems become regional. That is the quiet case for spending this kind of money, and it is also the case critics want itemized before they nod.


The Machine That Finally Met the Rock

Elected officials in helmets and yellow jackets gathered in New Jersey for the ceremonial launch. Within days the first boring machine was set to drill through solid rock in the Palisades cliffs. Two machines will cut about a mile of hard stone on the New Jersey side. That phase is expected to take roughly a year. Only then does the more famous part begin: the drive under the river itself.

People love the river crossing in the telling. The cliff is the unglamorous prerequisite. You cannot romance your way through basalt. The cutters have to chew, the spoil has to leave, the lining has to follow. If that first mile slips, every later date slips with it. Megaprojects rarely fail in a speech. They fail in a sequence.

  • Phase one cuts about a mile of hard rock under the Palisades.
  • Two boring machines handle that opening drive.
  • The cliff phase is penciled in at about one year.
  • The under-river drive starts only after that rock work is done.
  • Full completion is still projected around 2035.

A year of cliff cutting sounds manageable until you remember what sits above and beside the alignment. Utilities, neighborhoods, the ordinary life of North Bergen and the towns nearby. Tunneling is a negotiation with geology and with people who did not ask for a conveyor of rock past their week. Done well, they barely notice. Done badly, the project spends its political capital before the river is even reached.

Penn Station Is the Real Destination

The tunnel is not a destination. Penn Station is. Every new track under the Hudson is a promise about how trains arrive, turn, and leave again. The wider program, run through the bi-state Gateway Development Commission, treats this bore as one piece of a Northeast corridor modernization. Planners describe it as one of ten projects meant to upgrade the rail line from Newark to Penn Station.

That framing matters. A beautiful tunnel that dumps into a cramped station throat is a longer queue with better wallpaper. The four-track end state only works if signals, platforms, and the approaches agree with it. I keep coming back to that. Concrete is visible. Capacity is a system.

What Four Tracks Change, and What They Do Not

Four lines in and out do not invent new riders. They absorb the ones already there and the ones the region keeps adding. They let crews maintain the old tunnel without asking half of New Jersey to work from a cafe. They give dispatchers a way to route around trouble. They do not, by themselves, fix fares, staffing, or the last mile once you are on the street in midtown.

Still, the difference between two tracks and four is the difference between a hallway and a hallway with a passing lane. Anyone who has tried to renovate a kitchen while cooking dinner in it will recognize the current arrangement. You are repairing the room you cannot close.

Piece of the jobWhat it is meant to doRough horizon
Palisades rock boreOpen the New Jersey approach through hard cliffAbout one year from the start of digging
River tunnelsAdd two new bores under the HudsonFollows the cliff phase
Old tunnel rebuildRestore the 1910 North River TunnelAfter new tubes can carry service
Four-track outcomeTwo new plus two rebuilt lines at Penn StationProjected around 2035

Dates on a chart are not dates on a platform. Weather, supply chains, a surprise in the rock, a lawsuit, a budget revision: any of them can shove 2035. Anyone selling certainty about a river tunnel is selling something geology has not signed.

The Money, and the Fight Over It

The headline figure is about $16 billion for the tunnel portion riders will notice. During the previous administration, the Department of Transportation authorized roughly $11 billion in grants and financing tied to that Hudson Tunnel work. That federal share is why a local groundbreaking still felt like a national argument.

In October 2025, the White House budget office froze federal funding, citing what it called unconstitutional diversity principles attached to the project. New York and New Jersey sued. On June 29, a federal district judge, Jeannette Vargas, ruled the freeze unlawful and funding was reinstated. The government appealed, then the Department of Transportation dropped that appeal on September 17. The machines could meet the rock with the check, for now, unstuck.

Did they try to stop us in two thousand different ways? Did they try to put a log in our path, whether it was from Washington, New York, New Jersey… we are building this tunnel.

A senior senator at the North Bergen launch

The line plays well at a groundbreaking. It also skips the boring question, which is the right one. Not whether someone tried to slow the job. Whether the job, at this price, with this schedule, is the version the region should build. Opposition is not automatically sabotage. Sometimes it is a demand to see the spreadsheet.

A Freeze, a Lawsuit, and a Dropped Appeal

Funding fights on megaprojects follow a script. One side calls the conditions political. The other calls them accountability. Courts then decide which memo was allowed. Here the ruling went against the freeze, and the appeal did not survive into the fall. That sequence put shovels, or rather cutters, back on the critical path.

Does a dropped appeal end the politics? Unlikely. Appropriations are annual creatures. A project that runs toward 2035 will outlast more than one administration, more than one set of governors, and at least one rewrite of what counts as a permitted contract clause. Riders should treat legal calm as weather, not climate.

  1. Federal grants and financing of about $11 billion were authorized for the tunnel work.
  2. Those dollars were frozen in October 2025 over disputed contract principles.
  3. The states sued and a judge ordered the money restored on June 29.
  4. An appeal was filed, then abandoned on September 17.
  5. Digging through the Palisades followed once the path was clear enough to start.

I am not especially moved by victory laps at groundbreakings. I am moved by whether the next invoice clears. Infrastructure has a long memory for projects that started loud and finished as a hole with a fence around it.

Who Pays, and Who Waits

Federal money does not fall out of the sky. It is tax revenue, borrowing, or both, routed through grants and financing tools. State contributions sit beside it. Riders will pay fares for decades on a system this tunnel is meant to keep upright. The honest pitch is shared burden for a shared choke point, not a gift from somewhere else.

There is a fair complaint that megaprojects socialize risk and privatize the press release. Cost overruns rarely send a bill to the person who cut the ribbon. If this job stays near $16 billion, it will be an outlier in the history of urban tunnels. If it does not, the argument will restart with louder numbers and the same aerial photo of the river.

The Corridor Is Bigger Than One Bore

Gateway is the umbrella. The Hudson Tunnel Project sits inside a string of jobs between Newark and Penn Station. Bridges, approaches, station work, the unphotogenic links that decide whether a new tube is useful on day one. Treating the tunnel as a solo act makes for a cleaner headline and a worse railroad.

Think of the corridor as a chain. The new links under the river are the ones everyone can point at on a map. A weak link in the Meadowlands, or a throat that still cannot swallow four tracks, will cap the benefit. Planners know this. Budgets sometimes pretend otherwise, because a single named tunnel is easier to sell than a system.

Capacity is not a tunnel.
Capacity is tunnel + signals + yards + crews + a station that can turn trains.
Miss one, and the ribbon is decorative.

That is not cynicism. It is how dispatching works. A train does not care about the press packet. It cares about the next clear block.

Delays Have a Price Even When No One Itemizes It

Economists like to put a dollar figure on lost time. Commuters already know the number in a rougher unit: the meeting you joined from the vestibule, the dinner that went cold, the job you did not take because the trip was a gamble. The Northeast corridor is not a scenic extra. It is how a slice of the national economy gets to work.

A more reliable crossing supports offices, hospitals, universities, and the service jobs that exist because those places are full. It also supports the less fashionable fact that New Jersey housing is part of the New York labor market. When the tunnel wheezes, that market wheezes. You can dislike the price of the fix and still admit the cost of not fixing it is paid daily, just in smaller coins.

Construction Years Are Their Own Commute

Between now and the mid-2030s, the old tunnel keeps the lights on. That is the awkward middle. Riders will be asked to tolerate the system they already tolerate, while machinery chews a parallel path they cannot use yet. Expect weekend outages, bus bridges on some windows, and explanations that sound precise until you are the one standing on the platform.

The cliff phase is mostly invisible to a rider in Secaucus or Summit. The river phase and the later rebuild of the old tubes will not be. Rehabilitation of a live tunnel is surgery on a patient who is walking to work. Service plans will matter as much as cutterhead diameter. A project can be technically excellent and still feel, to the public, like another year of maybe.

  • Weekday peaks will still depend on the 1910 tubes for years.
  • Maintenance windows may widen as the new bores advance.
  • The rebuild of the old tunnel cannot fully start until the new ones carry trains.
  • Communication, more than concrete, will decide whether riders stay patient.

If the agencies treat riders like an audience instead of a partner, patience will thin. People can handle a delay they understand. They revolt against a delay that arrives as a shrug.

Geology Does Not Attend the Ceremony

The Palisades are not a metaphor. They are a wall of rock that has been there longer than the cities on either side. Boring machines are good at this. They are not magic. Mixed faces, water inflows, a zone that is harder than the model: any of it can slow the advance. The river section brings its own file of risks, from soft ground to the simple fact that you are digging beneath a working waterway into the busiest rail station in the country.

I tend to trust schedules more after the first hundred meters than at the photo op. Early production rates tell you whether the estimate was a forecast or a wish. Watch those rates, not the helmets.

What Success Would Actually Look Like

Success is not the day the machine breaks through. Success is a random Wednesday in 2036 when a disabled train in one tube does not erase the evening in the other. It is a maintenance shift that happens in daylight on a track the public is not using. It is Amtrak and NJ Transit arguing about slots instead of arguing about survival.

It is also a final cost that a reasonable person can explain without wincing. Not a bargain. A number that matches the risk that was retired. Regions remember boondoggles. They also remember the crossing that quietly stopped being the reason they were late.

Useful test: if one track fails at 5:40 p.m. and the board stays mostly honest, the project worked.

That test is years away. It is still the only one that counts.

The Politics Will Keep Riding Along

A bi-state commission, two governors, a federal department, a court that already ruled once, and a corridor that members of Congress love to mention in speeches. This project has more parents than a school play. That can protect it. It can also freeze it, as last year’s funding pause showed.

The diversity-clause dispute will not be the last argument about strings attached to federal dollars. Whatever you think of those clauses, the practical lesson is narrower. Conditions that one administration writes, the next may refuse, and the courts may or may not rescue the calendar. Projects that need a decade of uninterrupted checks should be built to survive a change of mind in Washington. This one just received a reminder.

There is a version of this story that is only tribal. One side tried to kill the tunnel, the other side saved it, cue the hard hats. The more useful version is managerial. Can the commission keep contractors, cash flow, and community commitments aligned long enough to finish a hole under a river? Speeches do not bore rock. Monthly progress does.

A Note on Skepticism Without Nostalgia

Skeptics of the price are not required to love the 1910 tunnel. You can believe the crossing is essential and still flinch at how American megaprojects accumulate cost. Labor rules, environmental review, design changes, risk priced into every bid, the simple fact that digging beside a live railroad is slow on purpose. Some of that cost buys safety. Some of it buys process. Telling them apart is the adult job, and it is rarely done in a groundbreaking quote.

I do not buy the idea that old infrastructure should be cherished until it fails in public. I also do not buy the idea that a large number is self-justifying. The Palisades cut is a chance to watch which story is true. If the first year of boring stays near plan, the rest of the argument gets easier. If it does not, the $16 billion figure will age badly in every hearing that follows.

What Riders Can Watch From Here

You will not see the cutterhead from the train window. You can still track the project like a commuter rather than a tourist. Is the cliff phase still described as a one-year job six months from now? Are service advisories specific, or are they weather reports made of adjectives? Does the funding line in public updates stay dull? Dull funding is good funding.

Ask simple questions when officials speak. When does the river drive start? When can the old tubes be taken out of service one at a time? What does a four-track railroad change about the worst hour of the day, not the average hour? Those answers are the difference between a construction update and a plan.

  • Follow whether the Palisades phase holds its one-year mark.
  • Notice if river tunneling dates stay fixed or start to drift.
  • Treat new lawsuits or funding pauses as schedule events, not just headlines.
  • Judge the project by peak-hour recovery, not by ceremony photos.

None of that requires a hard hat. It requires the mild suspicion of someone who has heard “on time” before and then stood on a platform anyway.

The Human Scale of a Regional Machine

It is easy to talk about this as concrete and appropriations. The scale snaps back when you picture a single car. A nurse coming off a night shift. A student who built a schedule around a 7:42. A crew that needs the yard on the far side before the morning peak. The tunnel is infrastructure. The delay is personal, repeated hundreds of thousands of times until it looks like a statistic.

That is why the dig, for all its cost and politics, is not abstract. A region that cannot reliably cross its own river is negotiating against itself. New housing in New Jersey, new jobs in Manhattan, offices that still want people in them three days a week: all of it leans on a pair of tracks finished when Taft was in the White House. Replacing that dependence is slow. Leaving it in place is a choice too.

Between the Photo and the Breakthrough

Groundbreakings are theater. Necessary theater, sometimes, because projects this large need a public moment that says the argument phase has yielded to the cutting phase. The risk is believing the theater. A senator’s vow, a judge’s order, a machine with a name painted on the side: none of those is a tunnel. The tunnel is the day trains run through it and the old one can finally rest long enough to be rebuilt.

Between those days sits a decade of invoices, community meetings, and evenings that still depend on 1910. If you ride the corridor, the honest posture is interested patience. Glad the cutters are in the rock. Unwilling to clap for a finish line that has not moved.

The Hudson did not get narrower while everyone argued. It will not get kinder while the machines work. What can change is the number of ways across, and the odds that a single failure still owns your evening. That is a modest promise for $16 billion. It is also, if they keep the schedule anywhere near honest, the first real one this crossing has had in years.


A Practical Reading of the Next Twelve Months

The next year is the tell. Palisades rock is the scheduled work. If production stays steady, the river drive can be planned instead of hoped for. If production stumbles, every later promise should be marked provisional. Watch spoil removal, lining rates, and whether officials keep using the same year-long estimate or start adding adjectives.

Funding should look boring in this window. A reinstated grant that stays reinstated is the baseline, not a favor. Contractors price political risk. Every time that risk flickers, the bid environment gets worse, and the public pays for the flicker later. Stability is a construction material. It just does not show up in the rendering.

Riders, meanwhile, should assume the old tunnel remains the main character. Apps will still glitch. Boards will still shuffle. The new capacity is a future tense verb. Treating 2035 as next fall is how disappointment gets manufactured. Treating it as a long repair of a vital pipe is how the wait stays tolerable.

Why the Old Tubes Still Deserve a Rebuild

There is a temptation, once the new bores exist on paper, to talk as if the 1910 tunnel can be retired like a flip phone. It cannot. The plan needs those tubes back in service after rehabilitation so the corridor actually has four lines. Abandoning them would waste a right-of-way that took a century to earn and would leave the new tunnels as a replacement rather than an expansion.

Rehabilitation is also where the Sandy damage has to be faced in full, not patched between trains. Concrete, rails, signals, drainage, the unglamorous organs of a flooded tube. Doing that work while trains still run is why the new bores have to come first. Sequence is the whole strategy. Reverse it, and you close the door you are trying to widen.

I keep a soft spot for infrastructure that gets a second life instead of a eulogy. Not from nostalgia. From thrift. A rebuilt tunnel plus a new one is more railroad than a new one alone, and the region is not suffering from an excess of tracks under that river.

The Question the Ceremony Did Not Answer

Can a project this large stay a construction job instead of a permanent campaign? That is the open question. The launch in North Bergen answered a narrower one: the machines are allowed to dig, and the first rock is no longer theoretical. Everything riders care about sits downstream of that permission.

So the photo is fine. The helmets are fine. The number on the estimate is large enough to demand a skeptic in the room. None of that replaces the next headline that matters, which will be dull if the project is healthy. Advance rate. Contract status. A river drive that starts when the cliff work ends, not three revisions later.

Until then, the 1910 tunnel keeps the appointment it has kept for more than a century. It will be late sometimes. It will be crowded on purpose. And somewhere under the Palisades, a machine is finally cutting the detour that was argued over for longer than some of its future riders have been alive. About time, I would say. Also about time we judged it by the rock, not the speech.

❝
Innovation distinguishes between a leader and a follower.
— Steve Jobs
Author

Steven Soarez passionately shares his financial expertise to help everyone better understand and master investing. Contact us for collaboration opportunities or sponsored article inquiries.

Related Articles

?>