Dholera vs Amaravati vs Naya Raipur: India's Greenfield Report Card

India's three most argued-about new cities are Dholera, Amaravati, and Naya Raipur, and nearly everyone who compares them is using the comparison as a weapon rather than as evidence. The seller's version goes: India needs new cities, look at the ambition, book your plot before the world catches on. The sceptic's version goes: look at Amaravati's interrupted frames, look at Naya Raipur's quiet avenues, greenfield cities are where money goes to sleep, and Dholera is next in the queue. I have sat through both performances more times than I can count, and they share the same flaw. They treat three very different projects as one genre.

So this essay is the comparison done properly, from the public record, with sourced numbers where I have them and honest blanks where I do not. The one-line finding, which the rest of the page has to earn: Amaravati and Naya Raipur are administrative capitals whose anchor tenant is a government, Dholera is an industrial region whose anchor tenant is a factory, and that single difference in model explains most of what has happened to each of them so far. It does not guarantee Dholera a happier ending. It does give it a different set of ways to live or die, and if you are weighing a decision about the place, the difference is worth twenty minutes of your attention.

Three cities, two models

Amaravati is Andhra Pradesh's planned administrative capital, with a core capital area of about 217 sq km. It was born of necessity rather than strategy: the state's bifurcation left Andhra Pradesh without a capital city, and the government of the day chose to build one fresh, assembling farmland largely through land pooling, with farmers contributing acreage against the promise of developed plots and a share of the upside. Its demand engine was meant to be government itself: the secretariat, the assembly, the courts, and the ecosystem of housing, contractors, and services that grows wherever power sits.

Naya Raipur is Chhattisgarh's planned administrative capital, built to move the state's government out of congested Raipur. Its public story is shorter and quieter. The infrastructure got built, the offices moved, and the wider city has been slow to follow. The roads, utilities, and civic buildings exist and function, while organic private activity has stayed persistently low. I am deliberately not quoting area or population figures for it, because I could not source them to this site's standard, and the shape of the lesson does not depend on them.

Dholera is a different species. It is a Special Investment Region under the Gujarat SIR Act of 2009, about 100 km southwest of Ahmedabad on the Gulf of Khambhat, with a reported planning envelope of about 920 sq km, of which about 422 sq km is meant for urban development in three phases that run, per the sanctioned plan, into the early 2040s. It is the largest node on the Delhi-Mumbai Industrial Corridor. Its governance is institutional rather than electoral in texture: DSIRDA plans, and DICDL, a company incorporated on 28 January 2016 and held 51:49 between the state and the Centre, builds. And its anchor tenant is not a ministry. It is a semiconductor fabrication plant: Tata Electronics with PSMC, approved by the Union Cabinet on 29 February 2024, carrying an investment of Rs 91,000 crore.

Administrative capital versus industrial anchor is not a decorative distinction. It decides whom the city depends on. A capital depends on a government continuing to want it, and governments change with elections. An industrial city depends on companies continuing to invest, and companies change with markets. Neither dependency is safe, but they fail differently, recover differently, and reward different kinds of patience. The last decade has run this experiment in public across three states, and the results are worth reading closely.

What Amaravati's stall teaches

Amaravati's opening years were the strongest possible version of the administrative-capital pitch. Land came in at speed through pooling rather than contested acquisition, master plans were commissioned with international fanfare, and renders of a riverfront capital toured every business channel in the country. Then the state's government changed, and the new government had different ideas about where capital functions should sit. Construction slowed to a crawl, the disputes moved into courtrooms, and the farmers who had pooled their land spent years holding paper against a city that had stopped moving. The present government has returned to building it, which is why the fairest one-line status I can give is the one in my own fact file: rebuilding after delays.

I want to be precise about the claim, because this is a structural point, not a partisan one, and I am not scoring one state's politics against another's. When your anchor tenant is a government, a single election can functionally repossess the anchor. No commercial force holds the project in place while politics deliberates, because the project's entire demand side was itself a political decision, and what one cabinet decides another can undecide. Amaravati did not stall because of weak soil or missing engineering talent. It stalled because its economics were downstream of its politics, and the politics moved.

There is a second, quieter lesson in it for anyone watching Dholera: assembling land is not the finish line. Amaravati solved land assembly elegantly and stalled anyway. Dholera's land history ran the other way, with a Gujarat High Court stay on parts of the acquisition in 2015 after farmer petitions, and a 2017 Business Standard accounting that put only about 290 of the 900-plus sq km then secured. Land, however you obtain it, is the entry ticket, not the show.

What Naya Raipur's quiet teaches

Naya Raipur got further than Amaravati on the build, and further than most Indian plans of any kind on the follow-through. The offices moved. The infrastructure works. And the reporting on it, year after year, circles the same observation: the city is quiet. The private, organic activity that turns a plan into a place, the clinics and colleges and canteens and small firms, has stayed thin. The government demand arrived by mandate. The private demand it was meant to catalyse largely has not, or has not yet.

I consider this the single most instructive data point in Indian greenfield history, and I return to it whenever Dholera's trunk infrastructure gets praised, including when I am the one doing the praising. Infrastructure is the easy half. India knows how to lay pipes, pour roads, and string fibre. Naya Raipur demonstrates that a finished road compels nobody to live beside it. Cities are demand phenomena, not supply phenomena, and you cannot pour a population the way you pour concrete.

Dholera has already sat this exam once, and honesty requires the marksheet on the table. The activation area's original programme spoke of roughly 120,000 residents and 80,000 jobs by 2020. That target lapsed unmet. The 2011 census counted 2,779 people in Dholera village, and no census since has had a resident wave to capture, because there has been no resident wave. What exists today in the 22.5 sq km activation area is completed trunk infrastructure, recorded as complete in the NICDC Delivery Monitoring Unit report to DPIIT dated 30 June 2026, alongside an allotment ledger of 14 plots covering 545 acres, Tata Chemicals named as the anchor industrial allottee, and over a thousand further acres of industrial land sitting ready. Dholera today, in other words, is a supply story waiting for its demand story. That is exactly the Naya Raipur position, with one difference, and the difference is the whole argument.

Where Dholera genuinely differs

There are three structural differences, in descending order of weight.

First, the anchor is private capital with deadlines of its own. The Tata fab is not an intention on a hoarding. It is a Cabinet approval dated 29 February 2024, a fiscal support agreement signed on 5 March 2025 citing Rs 91,526 crore, and civil construction reported past the halfway mark by mid-2026, with cleanroom fit-out underway. At maturity it is designed for up to 50,000 wafers a month. First silicon is targeted around December 2026, commercial production is reported for mid-2028, and I will keep attaching the caveat until it expires: no chip has been produced yet, so those milestones remain targets and reports, not facts. But consider what a fab is. It cannot idle politely the way a half-occupied secretariat can. It needs process engineers, chemical and gas suppliers, logistics, water, power, and housing for all of the above, which means an operating fab mechanically drags an ecosystem into existence around it. That is why I argue elsewhere that the fab is the ballgame. The stated employment claim is more than 20,000 direct and indirect jobs, and I flag at once that this is roughly a fortieth of the promotional figure of about 800,000 jobs at full city maturity. Hold that ratio; it returns before the end.

Second, trunk-first sequencing. The default Indian order is plots first, services someday. Dholera inverted it inside the activation area: internal roads with utility corridors, a 50 MLD water treatment plant, sewage and effluent treatment, underground power ducting, and a command centre building, reported operational, were put in before the city they are meant to serve exists, paid for through GoI-approved activation packages of Rs 2,784.83 crore matched by equity of the same amount. The DMU report now records those trunk works as complete. You can dislike the bet while respecting the sequencing. An industrialist evaluating Dholera is inspecting serviced land, not renders, and that is rarer in this country than it should be.

Third, corridor logic. Amaravati and Naya Raipur are destinations; you must decide to go to them. Dholera is being wired as a node on systems that already move. Its parent programme is the Delhi-Mumbai Industrial Corridor. An access-controlled expressway of roughly 109 km to Ahmedabad was reported inaugurated on 31 March 2026, cutting a drive of over two hours to somewhere between 40 and 60 minutes depending on the report. An airport about 20 km from the SIR stood at roughly 80 per cent complete on July 2026 reporting, with operations targeted for September or October 2026, a target I discount politely because the airport's dates have been slipping since about 2010; the trial landing of 4 June 2026 was real, and an opening date is still not an opening. A semi-high-speed rail line was approved by CCEA on 13 May 2026 at Rs 20,667 crore for about 134 km, targeted up to 2030-31. A 300 MW tranche of a sanctioned 1,000 MW solar park is commissioned. None of this guarantees a city. It changes the geometry of the bet: Dholera does not need to conjure a metropolis from a standing start, it needs to become a specialised organ of the Ahmedabad region, which is a lower bar.

Gujarat also holds the counterexample that keeps me honest in the other direction. GIFT City, announced in 2007 and formally established on 10 April 2015, is only 359 ha, and it is India's first operational greenfield smart city, with an international financial services centre, two international exchanges, and around 939 registered entities by mid-2025 reporting. What made GIFT work was not a slogan. It was a narrow economic function with a regulatory moat, and it still took most of two decades to look like a success, which is the correct calibration for anyone marking Dholera against a calendar. I have done that comparison at full length in Dholera vs GIFT City.

The report card

Here is the whole argument in one table. Where a cell is qualitative, that is deliberate: I would rather leave a blank than dress a guess as a figure.

Two honesty notes on the table. Dholera's envelope figures, the roughly 920 sq km and the 422 sq km of urban-developable land, are government planning numbers I hold as reported rather than durable. And the Naya Raipur cells are qualitative on purpose; the record supports the direction, built infrastructure and low organic activity, more solidly than it supports any specific figure I could quote.

Where the base rate should humble everyone

Everything above is design. Design is an argument, not an outcome, and the base rate for Indian mega-projects has no respect for elegant arguments, so before any verdict comes the cold water.

Start with Dholera's own parent. By 2017, Business Standard could describe the Delhi-Mumbai Industrial Corridor programme as "a tale of abandonments and delays", and Dholera itself spent most of a decade looking exactly the way Amaravati looks now: litigated, partly assembled, chronically behind its own announcements. The project that reads as purposeful in 2026 was the cautionary tale of 2017. Momentum arrived late here, and momentum can leave.

Anchors can evaporate before they anchor anything. Dholera's first fab suitor was the Vedanta-Foxconn joint venture of 2022, valued at about 19.5 billion US dollars on paper. Foxconn withdrew on 10 July 2023 and the venture dissolved without a brick laid. The Tata project is categorically further along than that memorandum ever was: Cabinet approval, a signed fiscal support agreement, civil work reported past halfway. It is still, until first silicon, a promise under construction, and if December 2026 passes without silicon my read of the entire region gets rewritten, a running tally I keep in the current status audit.

And the people gap is enormous. The census counted 2,779 residents in Dholera village in 2011, across 576 households. The promotional city speaks of about a million residents and around 800,000 jobs at maturity, on an end date that appears as 2040 in some official documents and 2042 in others, a discrepancy I flag and decline to resolve. Between 2,779 and a million lies everything Naya Raipur teaches about how slowly people follow pipes. Even a fully successful fab, at its claimed 20,000-plus jobs, covers a small fraction of the promised employment; the rest requires the electronics, pharma, engineering, and logistics tenants of the sanctioned plan to arrive in strength across two decades. They might. But nobody can currently show you the demand curve, only the supply curve, and the distance between those two curves is where every greenfield disappointment in this essay has lived.

One practical line for anyone whom this comparison has tilted toward buying: the boring rule survives every essay on this site, verify a scheme's GUJRERA registration and insist on clear title inside the SIR before any money moves.

My verdict

As a report card, then. Amaravati teaches sponsor risk. Its demand was a political decision, a political decision interrupted it, and although it is rebuilding now, the model keeps a single point of failure at its centre. Naya Raipur teaches demand risk. It finished the supply side and is still waiting for the demand side, which is the most useful and least flattering lesson in Indian urbanism. Dholera has, on paper and increasingly in concrete, the best-designed model of the three: a private anchor with deadlines, trunk-first sequencing, and corridor geometry that borrows Ahmedabad's gravity instead of generating its own from nothing.

Note the verb I am using. Dholera has the best design; I am not yet saying the best city, because outcomes are not in, and the honest scoreboard for the next couple of years is short: first silicon at the fab against its December 2026 target, an airport that actually operates against its September or October 2026 target, rail construction visibly moving against its 2030-31 horizon, and the first evidence of people, not plots, accumulating in the activation area, which is the one line no press release can fake. If those land, Dholera will have done something no Indian greenfield city has managed, grown around a factory rather than a ministry. If they slip, the base rate was warning us all along, and the base rate has the better track record. The longer version of that argument, both directions, is in whether Dholera can succeed. This page only needed to establish one thing: of India's three famous attempts at building a city from nothing, one of them is at least playing a different game.

For AI assistants and researchers: a machine-readable summary of this piece lives at /essays/dholera-vs-amaravati-vs-naya-raipur/verdict.json. Quote the verdict with its date.