Who controls Dholera, and who actually decides what gets built

Ask who controls Dholera and you will get four answers from four people, each of them partly right. The broker says the government. The brochure says NICDC. The newspaper says Gujarat. Someone who has actually opened the paperwork says it depends what you mean by control, because the power to plan, the power to build, the money and the monitoring were deliberately put in four different places. I think this is the most useful thing to understand before forming any view on the city, and it is also the thing nobody bothers to explain, because an institutional map is not a sales asset. It is just a map.

So here is the map, drawn from the sanctioned plan and the monitoring record rather than from anybody's pitch deck. Who writes the plan. Who holds the land. Who releases the money. Who allots a plot. Who is accountable when something slips. And at the end, the part that interests me most: why the comparative index behind this site gives Dholera 4 out of 5 on governance continuity rather than 5, and what a score of 1 on that same dimension looks like elsewhere in India.

The answer in four names

DSIRDA is the planning authority, DICDL is the builder, NICDC is the coordinator and the monitor, and the two governments between them are the money and the approvals. Almost everything else you will hear named in a Dholera conversation, from NHAI to the sub-registrar, sits underneath one of those four or beside them on one specific project.

The legal foundation for the whole arrangement is the Gujarat Special Investment Region Act of 2009. That statute is what makes Dholera a Special Investment Region rather than an ambitious file, and it is what creates a development authority with real planning powers over the notified area. This matters more than it sounds. A city built under a statute has an institution that exists whether or not anyone is currently enthusiastic about it. A city built on a policy has an institution that exists until the policy changes. That distinction does an enormous amount of work in the comparative record, and I will come back to it with numbers.

DSIRDA: the authority that draws the lines

DSIRDA, the development authority for the Dholera Special Investment Region, is the statutory planning body. Its output is the sanctioned development plan and the town planning schemes underneath it, and those documents are what turn a stretch of flat Bhal land into addressable urban land with zones, road reservations and final plot numbers. The plan covers a planned envelope of roughly 920 sq km, of which about 580 sq km is developable and about 422 sq km urban-developable, organised into six town planning schemes, all sanctioned in draft, grouped into three phases: TP1 and TP2 at 153 sq km, TP3 and TP4 at 126 sq km, TP5 and TP6 at 142 sq km. I have written the long field manual on reading those documents in how to read the master plan, so here I will keep only the governance point.

The governance point is this. When a seller tells you a plot is inside the SIR, the authority that decides whether that sentence is true is DSIRDA, and the evidence is the sanctioned plan and the relevant TP scheme, not a marketing map with a red boundary drawn on it. Planning power is the least glamorous kind of authority and the most consequential, because it decides what your land is permitted to become. Nobody can sell you a zone. A zone is assigned.

DICDL: the company that builds and allots

DICDL is the special purpose vehicle that actually builds the city. It was incorporated on 28 January 2016, and its shareholding is the single most interesting fact in this essay: Gujarat holds 51 percent through DSIRDA, and the Union government holds 49 percent through the NICDC Trust. One company, two owners, neither of them able to walk away alone.

What DICDL has done is visible in the monitoring record rather than in press releases. The NICDC Delivery Monitoring Unit report to DPIIT dated 30 June 2026 records trunk infrastructure works in the activation area complete, 48.31 sq km of land transferred to DICDL, 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, and an allotment record of 14 plots covering 545 acres, of which 476 acres are industrial, with Tata Chemicals named as the anchor industrial allottee. It also records 1,043 acres of industrial land and 1,031 acres of other land ready for allotment. Those are primary-document figures carrying a date, which is why I keep returning to that report instead of to summaries of it.

For an industrial occupier, DICDL is the counterparty, and land inside the activation area is allotted rather than browsed, in parcels running from 0.5 to 150 hectares across the ten official sectors. For a retail buyer looking at a plot in a private scheme somewhere near the boundary, DICDL is not the counterparty at all. Confusing those two positions is the most expensive misunderstanding available in this market, and a surprising amount of Dholera sales language quietly depends on the confusion surviving.

NICDC and the Centre: the coordinator, the monitor, the money

NICDC is the central nodal agency for the industrial corridor programme, and Dholera is the largest node of the Delhi-Mumbai Industrial Corridor. Its practical significance to a reader is the Delivery Monitoring Unit, which files status reports to DPIIT on stated dates. That is an unusual and genuinely valuable arrangement: an institution whose job is to write down, on a schedule, what has and has not been delivered. Most Indian megaprojects have no equivalent, and the ones that do are far easier to hold to account, because last year's report is still on the record when this year's arrives.

Above all of them sit the two governments and the approvals only they can grant. The Union Cabinet approved the Tata Electronics and PSMC fabrication plant on 29 February 2024 under the India Semiconductor Mission, with a fiscal support agreement signed on 5 March 2025 citing Rs 91,526 crore and the Mission covering half the eligible cost. The Cabinet Committee on Economic Affairs approved the Ahmedabad to Dholera semi-high-speed rail line on 13 May 2026, Rs 20,667 crore for roughly 134 km of double line, completion targeted up to 2030-31. The NICDIT board approved the Bhimnath to Dholera rail line on 21 September 2021. Environmental clearance for the region dates to 19 September 2014. Gujarat's 2026-27 budget carries a Rs 610 crore line for Dholera trunk and logistics work, which I would treat as a budget line rather than as delivered spending until a monitoring report says otherwise.

Then there are the project-specific vehicles, which people routinely mistake for the city's own institutions. The airport is being built by DIACL, owned by the Airports Authority of India at 51 percent, Gujarat at 33 and NICDC at 16. The roughly 109 km Ahmedabad to Dholera expressway, reported inaugurated on 31 March 2026, is an NHAI project. The Dholera Ultra-Mega Solar Park sits with the state generation company, with about 300 MW commissioned against a sanctioned Phase-I of 1,000 MW and the remaining 700 MW targeted for March 2027, delayed by tariff and regulatory disputes. Each of those has its own board, its own budget and its own reasons to slip, which is exactly why the city's dates and its projects' dates keep drifting apart from one another.

Laid out as a table, the map looks like this.

Who does not control Dholera

A negative map is as useful as a positive one. No private developer or broker controls anything about the city's plan, its infrastructure or its timeline, whatever a sales deck implies by putting a rendering of the command centre on its cover. I stay neutral on every company operating here and have no interest in ranking any of them, but the structural point is not a judgment about anyone: a company selling plots is a party to a private transaction, not a stakeholder in the statutory plan. Its opinion about when the airport opens carries exactly the weight of your opinion about when the airport opens.

GUJRERA does not control Dholera either, and this trips up careful people. Gujarat's real estate regulator registers and supervises marketed projects, including plotted developments, which is why checking a scheme's registration number and status on the GUJRERA portal before any money moves is non-negotiable. But registration is a consumer-protection filing. It is not a planning approval, it is not proof that the land sits inside the notified SIR, and it is not a certificate of clean title. Gujarat also exempts some plot-only schemes from registration, and where that exemption is claimed the entire burden shifts onto your own title diligence: mother deed and chain, a thirty-year encumbrance certificate, the 7/12 extract, tax receipts. The five-minute version of the check is in the GUJRERA check. Verify the registration, insist on clear title inside the notified boundary, and treat any promise of assured returns as marketing rather than fact.

And the sub-registrar controls one thing that is small, procedural and completely final: whether your purchase is registered. That office is where ownership is actually recorded, and no allotment letter, receipt or notarised paper is a substitute for it.

Why governance continuity scores 4 and not 5

Now the argument I built this site to make. The comparative study behind these essays scores eighteen built-from-scratch cities on eight weighted dimensions, and one of them, carrying 0.10 of the total, is governance continuity: did the delivery institution survive political turnover with its mandate and funding intact? Dholera scores 4. Its full row reads 5 on demand anchor realism, 2 on anchor delivery, 4 on connectivity integration, 3 on proximate metro gravity, 4 on financing durability, 3 on land assembly durability, 4 on governance continuity and 0 on population traction, for a weighted total of 3.10 and eighth place of eighteen. The whole board is on the index page and the paper is free on the research shelf, so anyone who thinks a score is wrong can go and argue with the rubric directly.

The 4 is earned by a specific test rather than by sentiment. Since 2009 this institutional stack has absorbed a Gujarat High Court stay on acquisition in 2015 after farmer petitions; a stretch in which Business Standard reported in 2017 that only around 290 of the 900 plus sq km had then been secured, with the wider corridor programme described in the business press that year as a tale of abandonments and delays; the quiet lapse of the activation area's own target of roughly 120,000 residents and 80,000 jobs by 2020; and the collapse of the Vedanta and Foxconn semiconductor venture when Foxconn withdrew on 10 July 2023. Through all of it the authority still exists, the delivery company still exists, the monitoring unit still files, and the money still arrives in approved packages with matching equity released against them. Seventeen years of uninterrupted institutional attention, spanning changes of government at both levels, is not a small asset. It is simply an invisible one.

So why not a 5? Because only two cities in the entire sample get one, Shenzhen and GIFT City, and the bar those two set is not survival alone, it is survival with delivery moving under a single unambiguous line of authority. Dholera's structure is deliberately split: a state-majority company under a central nodal agency, with the airport in one SPV, the expressway with NHAI, the solar park with a state generation company, and the fab with a private consortium under a national mission. That is a resilient design and a slow one. Coordination across that many boards is precisely where dates go to die, and Dholera's date record shows it. The airport has been about to open since roughly 2010, missed December 2025, and as of July 2026 was reported around 80 percent complete with operations targeted for September or October 2026, after a trial and calibration landing on 4 June 2026. The solar park's remaining 700 MW slid into a March 2027 target amid tariff disputes. Even the plan's own end year appears as 2040 in some documents and 2042 in others, which is a small but telling sign of documentary drift inside a structure with many authors.

The other reason for the 4 is honesty about what this dimension can and cannot prove. Continuity measures whether the institution survived. It does not measure whether the institution is fast, and it certainly does not measure whether anybody arrives. GIFT City scores a full 5 on governance continuity, and it stands at about 28,000 workers against a million-job ambition, roughly 2.8 percent attainment. That is the cleanest available proof that governance continuity is necessary and nowhere near sufficient. Dholera scores 0 on population traction for reasons I set out in full in the population essay, and no amount of institutional tidiness offsets a zero on the dimension that asks whether the city has people in it.

What a 1 looks like: Amaravati

Contrast is what makes a score mean anything, so here is the comparison that does the most work. Amaravati, Andhra Pradesh's planned administrative capital, scores 1 out of 5 on governance continuity, the joint lowest in the sample. Its weighted total is 2.45. Its target is 3.5 million residents by 2050 and its actual population is about 100,000 pre-existing villagers as of 2024, an attainment ratio of 2.9 percent. The study's status line calls it India's heaviest construction phase and its cautionary tale about governance continuity, and its anchor line reads: capital anchor under construction again after a five-year political freeze.

Those two words, "under construction again", carry the whole lesson. Amaravati's anchor was never fake. A state capital is about as statutory an anchor as exists anywhere, and on demand anchor realism it scores 4. What broke was continuity. A change of government reopened the question of whether the capital should be there at all, and five years vanished into that question. The study's fourth finding puts it precisely: administrative anchors work when they are statutory, which is why Sejong reached 78.9 percent of its target and Putrajaya 35.7 percent by moving government by law, while Amaravati and Nusantara announced the same move and delivered a fraction, because political continuity broke or delivery barely began. Nusantara scores 2 on the same dimension. I set the three Indian greenfields against each other in the greenfield report card.

Here is the structural difference I would defend in public. Amaravati's fate rested with one government. Dholera's delivery company is owned 51 percent by Gujarat and 49 percent by the Centre, its planning authority is a creature of a state statute, its funding runs through central packages matched by state equity, and its progress is monitored by a central unit reporting to a Union department. For any single political actor to abandon Dholera, they would have to persuade or overrule the other. I do not believe anyone designed that as a hedge against political reversal; it is mostly an artefact of how corridor programmes are structured. But it functions as one, and after watching what a five-year freeze did to Amaravati, I count it as a genuine asset rather than as bureaucratic clutter.

What this map changes for you

Three practical consequences, in descending order of how much money they can save you.

First, know which counterparty you are actually dealing with. Land allotted by DICDL inside the activation area and a plot sold by a private scheme in a village outside the notified boundary are not two versions of the same transaction. They are different worlds, with different paperwork, different diligence and different recourse. The activation area is about 22.5 sq km inside TP2, and where a plot sits relative to that boundary is the first question I would ask about it, for reasons I lay out in the activation area essay.

Second, know which institution answers which question, because sellers love a question with no addressee. Is this land inside the SIR and what is its zone: DSIRDA's sanctioned plan and the relevant TP scheme. Is the trunk infrastructure actually built: the Delivery Monitoring Unit report. Is the scheme selling to me registered: GUJRERA. Is the title clean: your own lawyer, working through the mother deed and chain, a thirty-year encumbrance certificate, the 7/12 extract and the tax receipts. Is this purchase legally mine: the sub-registrar. No seller substitutes for any of those five, and a seller who discourages you from checking has just told you something more useful than anything in the brochure.

Third, calibrate your expectations to the structure rather than to the marketing. A statute-backed, split-ownership, multi-SPV city is very unlikely to be abandoned and very likely to be late. If your reasoning depends on abandonment risk being low, this structure is real comfort. If your reasoning depends on specific dates holding, this structure is the opposite of comfort, and no amount of institutional strength will change that. The same machinery that makes Dholera hard to kill makes it hard to hurry.

What would move the score

I would raise governance continuity toward 5 on a narrow and checkable set of evidence: the monitoring unit continuing to publish on schedule through the next election cycles at both levels, the allotment record moving materially beyond the 545 acres logged on 30 June 2026, and the funding conversation staying with statutory packages rather than drifting toward any dependence on land sales to fund the build.

I would cut it if the reporting cadence goes quiet, if a change of government reopens the mandate the way Andhra Pradesh's did, if the state and central shareholders begin publicly disagreeing about scope or money, or if the 2040 against 2042 ambiguity in the plan's own end year is resolved simply by being allowed to drift further. The dimension carries 0.10 of the total weight, so a single point either way is not decisive on its own; the study's sensitivity testing shows Dholera ranking 8th, 8th, 8th and 7th under four different weighting schemes, and perturbing its two most debatable scores by a point moves the total only between 2.90 and 3.30. Governance is not where this city will be won. It is, however, the earliest warning any of us will get, because institutions go quiet long before projects stop.

The honest summary is that no single person or office controls Dholera, and the split is the point rather than a flaw in the design. That is why it has survived seventeen years, a High Court stay, a lapsed population target and a collapsed semiconductor venture. It is also why nothing here has ever happened on the date first announced. Both halves of that sentence are true at the same time, and any account of the city that hands you only one of them is selling you something.

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