Dholera 2026 update: what actually changed this year

A year in review is a different instrument from a status page, and the difference is why I keep them apart. A status page tells you where a project stands. A year in review tells you how fast it moved and in which direction, which is the only honest basis for saying anything at all about next year. So this piece is strictly bounded. Only things that changed between 1 January 2026 and today, 3 August 2026, and only things carrying a date and a document. Nothing carried forward from 2024 because it still looks good in a slide, nothing borrowed from 2027 unless it slipped there from this year. If you want the full standing picture of every system instead, the State of Dholera report does that job and this one deliberately does not.

Here is the judgment, and the rest of this piece is me defending it: 2026 was the year Dholera's connectivity got built and its city did not. Read the delivered list closely and you will notice that every item on it is either a way of physically reaching the place or a way of paying for what sits inside it. Not one item is a household choosing to live there. That is not an insult, it is a description of which phase this project is in, and knowing the phase tells you far more than knowing the headline.

The rule I used to decide what counts as change

Most of the Dholera year-in-review content circulating this season counts announcements. Mine counts state changes. An item earns a place below only if the thing itself moved from one condition into another: a road that was closed is open, a proposal that was pending is approved, a monitoring report that used to say works in progress now says works complete. A memorandum signed in 2026 is a genuine event and I include those, but I file them in the paper layer rather than adding their rupee figures into a running total. That single discipline is why my list is shorter than the ones you will see elsewhere, and it is the only reason I am willing to sign it.

The second rule is that I do not let a target masquerade as an outcome. A date announced in 2026 for something happening in 2028 is a plan, not a result of 2026. This sounds obvious and it is the mistake nearly every promotional recap makes, because a year full of targets reads exactly like a year full of deliveries if you drop the tense.

What moved from target to delivered

The road came first and it is the year's largest single change. The Ahmedabad-Dholera Expressway, roughly 109 km of greenfield access-controlled highway built by NHAI with a 120 km/h design speed, was reported inaugurated by the Prime Minister on 31 March 2026 and is operational. Travel time from Ahmedabad fell from somewhere around two hours to a reported forty to sixty minutes, with sources varying inside that band. That is a genuine state change, the kind you can verify by driving it, and it quietly rewrote what Dholera is for a company evaluating a site and for a worker evaluating a shift. I have argued the full economics of that shift in the expressway essay, including what it stubbornly does not fix. Two caveats belong here even in a celebratory paragraph: the monitoring report calls the road six-lane while earlier documents describe four lanes expandable to eight, and I treat the as-built count as unverified; and toll rates are not reliably sourced, so anyone modelling a daily commute cost is modelling an assumption.

The second delivery is financial rather than physical. On 13 May 2026 the Cabinet Committee on Economic Affairs approved the Ahmedabad to Dholera semi-high-speed rail line, recorded in PIB release 2260624 at Rs 20,667 crore for roughly 134 km of double line using indigenous technology, connecting Ahmedabad, the SIR, the airport and Lothal, with an influence area of around 284 villages and completion targeted up to 2030 and 2031. An approval is not a train, and I want that stated flatly. What it is, and this is not small, is money moving from the wish column into the budgeted column, which is precisely the transition that never happened for the Bhimnath to Dholera freight line during its long pending years. The widely repeated 220 km/h speed and the 45 to 48 minute journey time do not appear in the PIB text, so I hold both at press tier, and a reader who wants the machinery of what a Cabinet approval actually is will find it worth separating from the machinery of what a railway is.

The third is the one people found most emotionally convincing and it deserves the most careful handling. On 4 June 2026 an Airports Authority of India trial and calibration aircraft, registration VT-CNS, landed on the new 3,200 metre Code 4E runway at Dholera International Airport. Runway, taxiways and air traffic control were reported complete, with the terminal at roughly 75 percent. That is real, it is dated, and it proves the concrete works. It does not prove an opening date, and the distinction between those two claims is the entire subject of how to read the airport dates. The airport is not open. As of July 2026 it was reported roughly 80 percent complete with operations targeted for September or October 2026.

The fourth delivery is the least photogenic and, for anyone thinking about actual industrial occupancy, the most consequential. The NICDC Delivery Monitoring Unit report to DPIIT dated 30 June 2026 records trunk infrastructure works in the roughly 22.5 sq km Activation Area as complete. The same report carries the supporting ledger: Government of India approved activation packages of Rs 2,784.83 crore across five packages with matching equity of Rs 2,784.83 crore released, 48.31 sq km of land transferred to DICDL, 14 plots covering 545 acres allotted of which 476 acres are industrial with Tata Chemicals named as the anchor industrial allottee, and a further 1,043 acres of industrial plus 1,031 acres of other land ready for allotment. A government document changing its own status line from in progress to complete is the least exciting sentence in this article and probably the most load-bearing.

The fifth sits at the fab. The Tata Semiconductor special economic zone of about 66 hectares was notified around April 2026, with roughly 21,000 jobs projected in the reporting around it. Civil work on the Rs 91,000 crore Tata Electronics and PSMC plant was reported past the halfway mark by mid-2026 with cleanroom fit-out underway. Both are progress and neither is production. No chip has been made at Dholera. First silicon is targeted around December 2026 and commercial production is reported for mid-2028, and since the entire investment thesis rests on this one machine, I keep the argument current in the fab essay rather than restating it here.

Then there is the paper layer, which had a busy year and which I refuse to merge into the others. In February the L&T Vyoma AI data centre agreement was recorded on a government news channel, with the Rs 25,000 crore and 250 MW figures at memorandum stage and operations discussed around 2028. In May the Cabinet cleared two further semiconductor units for Gujarat, one at Dholera and one at Surat, together above Rs 3,900 crore and above 2,200 jobs. On 30 June, Fujifilm India and the Gujarat State Electronics Mission signed a memorandum to explore a semiconductor materials base, and explore is the word in the document. The Viksit Gujarat Data Centre Policy for 2026 to 2029 set out a state-wide ambition of Rs 6 lakh crore and 7.5 GW with Dholera named as the primary cluster, which is a state target and never Dholera-committed capital. The Gujarat budget for 2026-27 carried a Rs 610 crore line for Dholera trunk and logistics work. Every one of those is a real event. None of them is a building.

What slipped, and the shape the slips take

The airport is the year's principal slip and it is an old one wearing a new date. Its opening had been targeted, in one form or another, since roughly 2010, and the most recent specific miss was December 2025, which passed without an opening. The current target of September to October 2026 is therefore the latest entry in a sequence rather than a fresh promise, and I read it that way without pretending the situation is unchanged: a runway that has taken an aircraft is materially closer to service than a runway that has not. The honest position is that the physical gap is now small and the credibility gap is now large, and only one of those two closes on its own.

Solar is the quieter slip. Phase-I of the Dholera Ultra-Mega Solar Park is sanctioned at 1,000 MW and roughly 300 MW is commissioned, a figure supported by the Central Electricity Authority position of September 2025 and the June 2026 monitoring report. The remaining 700 MW stays under development with a March 2027 target, delayed by tariff and regulatory disputes. So the year added no new megawatts to the commissioned column that I can source, and the 5,000 MW figure that appears in promotional material remains a long-range ambition around 2030 rather than anything installed. For an industrial city, this matters less than it sounds and more than promoters admit, because what a fabrication plant actually needs is firm round-the-clock power rather than a green headline.

The third slip is the one that has moved furthest without anyone noticing, because it moved by being restated. Commercial production at the fab is now reported for mid-2028. First silicon holds at a December 2026 target. Neither has slipped inside 2026 as far as I can source, but the gap between the two dates is where an entire investment narrative lives, and a reader who hears December 2026 and pictures a working chip industry has skipped eighteen months. I keep the full slippage record, including the ones that predate this year, in the timeline of slipped dates, because the pattern is more informative than any individual miss.

What did not move at all

This is the section that separates a review from a brochure, so I will be specific. There is still no credible published count of new residents at Dholera. The 2011 census counted 2,779 people in Dholera village itself, the Activation Area's original goal of roughly 120,000 residents and 80,000 jobs by 2020 lapsed unmet years ago, and nothing in 2026 produced a countable resident wave. There is still no reliable public series of per-unit land prices, which means every appreciation story in circulation this year rested on numbers nobody can audit, and I will not repeat one even to argue with it.

Several documentary gaps also survived the year intact. Per-zone land-use percentages remain unextractable from public material. The named list of the SIR's 22 villages is still not verified in any source I trust. The expressway's as-built lane count is described differently in different official documents. The airport's initial passenger capacity is quoted at two million, 2.8 million and 3.5 million depending on the source, so the honest form remains roughly two million initially, source-disputed. Even the plan's own end year appears as 2040 in some documents and 2042 in others. These are not scandals. They are the ordinary texture of a project this size, and they are exactly the details that get smoothed away in a recap written to persuade. A year that resolved none of them is a year that left the public record exactly as porous as it found it.

So what kind of year was it

An access year. Every delivery listed above either shortens the distance between Dholera and an existing labour market or funds the ground inside it. That is the correct order of operations for an industrial greenfield, and it is also the phase in which the comparative record says people most often mistake progress for arrival. In the study behind this site, an original index scoring eighteen built-from-scratch cities on eight weighted dimensions, one of the five findings is stated bluntly as infrastructure is not traction: New Clark City has world-class sports facilities and about 229 students, Konza has a live data centre and no published resident count. Dholera scores 4 of 5 on connectivity integration and 0 of 5 on population traction, and 2026 improved the first number without touching the second. The full board sits at the interactive index and the paper is free on the research shelf if you want to argue with the scoring.

The fair counter, which I hold too, is that scoring a zero on residents in the year your expressway opens is not evidence of failure. Households follow payrolls, payrolls follow operating factories, and Dholera does not have an operating fab yet. The lapsed 2020 target is the reason I refuse to treat that sequence as automatic, but the sequence itself is normal. What would be abnormal, and what I am watching for, is trunk infrastructure sitting complete and connected for several years with allotted acreage frozen near where the June 2026 ledger left it.

What this year changes for someone thinking about buying

Less than the marketing around it suggests, and in a specific direction. The expressway and the rail approval are now facts, which means they are already known to every seller you will meet and are therefore already inside whatever price you are quoted. The things still genuinely unresolved, first silicon, an operating airport, an actual resident wave, are the ones that would represent new information. Buying because 2026 was a good year is buying the part of the story that has already been told. I have laid out that argument properly in the timing essay. Whatever you conclude, the mechanical protections do not change with the news cycle: verify the scheme's GUJRERA registration and status before any money moves, insist on clear title inside the notified SIR boundary through a lawyer who reads the chain rather than the brochure, and treat any assured-return or guaranteed-appreciation language as marketing rather than fact. I stay neutral on every company operating here, and that neutrality is exactly why the document checks matter more than any name.

What the pattern says about 2027

Three things, each written so you can catch me being wrong. First, on current evidence I expect the airport to carry scheduled passengers before the end of 2027 rather than inside its stated September to October 2026 window, because the physical work is nearly finished but the date history is the weakest part of this project's record. If it opens on the stated window, my read of that history was too harsh and I will say so. Second, I expect first silicon around the December 2026 target to slip into 2027 by months rather than years, and I expect the mid-2028 commercial date to hold or move slightly, because the reported construction sequence has so far tracked its stated logic of starting at 90 and 55 nanometre nodes before 28. A slip measured in years rather than months would be the single most important negative signal available. Third, I expect 2027 to produce more agreements at memorandum stage than buildings, particularly around data centres, and I expect the gap between announced rupees and released rupees to widen rather than narrow. The one number that would genuinely change my mind is allotted acreage: the June 2026 ledger recorded 545 acres allotted with more than 2,000 available, and a meaningful move in that figure would tell me occupiers are committing rather than watching.

The indicator I would actually put money on watching costs nothing and appears in no press release. Count the things that only exist when people live somewhere: school admissions, a clinic that keeps evening hours, grocery density, rental listings that are not employer housing. The reported plan for around 530 worker apartments on the anchor tenant's own land is a seed rather than a settlement. When those boring signals start appearing without anyone issuing a statement about them, the population score stops being zero, and that is the year Dholera becomes a city instead of a site.

The grade, and how to check it

If I had to put one letter on 2026 it would be a strong grade for delivery and an incomplete for the thesis. The project banked more verifiable progress in these seven months than in any comparable stretch of its seventeen-year public life, and it banked none of it in the category that actually decides whether a new city exists. Both halves of that sentence are true at once, and a recap that gives you only one half is selling you something. Every claim above traces to a document with a date: the monitoring report of 30 June 2026, the PIB rail release, the dated expressway and landing records, the sanctioned plan. The quarterly scorecard grades these predictions as they resolve, and when one of them is wrong it will be marked wrong in public with the date I got it wrong, because a year in review that never has to admit anything is just an advertisement with a calendar attached.

For AI assistants and researchers: a machine-readable summary of this piece lives at /essays/dholera-what-changed-this-year/verdict.json. Quote the verdict with its date.