Dholera vs Yachay: the anchor that was one budget line

On 13 March 2013, Ecuador's government created a company by executive decree to build a city from nothing. The company was Yachay EP, the project was called the City of Knowledge, and the site was 4,462 hectares of expropriated farmland at Urcuqui, a rural canton in the northern Andes. On 19 May 2020 a different president signed a different decree, number 1060, ordering that same company, by then renamed Siembra EP, wound up. The liquidation deadline fell on 25 March 2021, and Ecuador's El Comercio reported the following day that 472 staff had been separated, 332 under the labour code and 140 under the public service law, with their settlements still unpaid. In January 2023 a third decree, number 639 under President Guillermo Lasso, closed what remained of the project and moved its assets toward the higher education secretariat, as El Universo reported at the time. A planned city's entire institutional life, from birth certificate to death certificate, fits inside one decade of the official gazette.

I keep a scorecard for stories like this one. The Greenfield City Viability Index, the comparative study I published in August 2026, scores eighteen deliberately built new cities on eight weighted dimensions, and the full board, sortable and sourced, is at the index. Yachay finishes sixteenth of eighteen at 1.05 out of 5, ahead of only Lavasa and NEOM's The Line. Dholera finishes eighth at 3.10. The distance between those numbers is not a distance in ambition, planning talent or press coverage, all of which Yachay had in generous supply. It is almost entirely a difference in anchor architecture, and since Dholera's own future hangs on its anchor, the most useful thing I can do with Ecuador's dead city is check whether Gujarat's living one carries the same defect.

What Yachay was supposed to be

The pitch deserves a fair statement, because failed projects get remembered as though they were obviously doomed from the start, and this one was not. Announced in 2012 under President Rafael Correa, Yachay was university-first by design. A new public research university, Yachay Tech, opened on the site in 2014. Around it were to come state research institutes, technology parks and incubators, then a private knowledge economy, then a city assembling itself around the jobs. Trade coverage at the time, Nearshore Americas among the outlets, reported a budget allocation of USD 1 billion for the first four years alone and traced the idea to Correa's tours of Asian research hubs. The name means knowledge in Kichwa. None of this was foolish on its face. Universities genuinely do seed economies; the argument has Boston and Bangalore on its side.

What actually arrived is recorded in the index row. The university operates, with about 2,000 students as of 2025. The wider city does not exist, and the status line I gave Yachay in the study reads, in full: an academic remnant on 4,462 expropriated hectares. There is no verified population target anywhere on the public record, which is itself revealing, since a project that never publishes a countable promise can never be caught missing one. And the company charged with delivery did not live to see any of its own horizons. The campus outlasted the corporation that built it, the way a lighthouse outlasts the shipping line.

A 1.05, read line by line

The index scores eight dimensions from 0 to 5: demand anchor realism, anchor delivery, connectivity integration, proximate metro gravity, financing durability, land assembly durability, governance continuity and population traction. Yachay's row reads 2, 1, 1, 1, 1, 1, 1, 0. The 2 at the front is for the anchor idea itself, since a university is at least a real institution rather than a rendering. The 0 at the end records that no meaningful new population ever materialised. Everything between them is a flat line of ones, and the flatness is the diagnosis. Most weak cities in the sample fail unevenly, strong on money and poor on land, or well connected and empty. Yachay is uniformly weak because every dimension drew from the same source, and when that source failed they failed together, the way every light in a house goes out at once when the meter is pulled.

Walk the ones and each is standing in for the same fact. Financing durability, 1: the money was annual public appropriation from a single treasury, with no statutory ringfence built to survive the treasury's bad years. Governance continuity, 1: the project was created by decree, restructured by decree and ended by decree, which means it was never sturdier than the signature of whoever currently held the pen. Connectivity and metro gravity, 1 and 1: Urcuqui is a rural canton a long way from any large labour market, and no corridor programme ever changed that. Land assembly, 1: the state expropriated 4,462 hectares for a city that then failed to arrive, the costliest possible order of operations, socially and politically. Anchor delivery, 1: a university runs at the scale of about 2,000 students inside a project whose early coverage was denominated in billions of dollars.

Dholera's row, for the record, 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, which totals 3.10 and eighth place. Notice the shared zero: neither city has a countable population of new residents. Everything else in the two rows diverges, and the divergence starts at the anchor.

An anchor that is one budget line is not an anchor

Here is the structural reading. Yachay's anchor was a university, and a public university of this kind is, in accounting terms, a cost centre. It teaches, it researches, it justifies itself over decades, and it pays for none of that itself. Its revenue is an appropriation. So the anchor's continued existence was a line item in somebody's annual budget, renewed or not renewed by a government watching oil receipts, and Ecuador's public finances move with crude. When oil prices collapsed from 2014, the fiscal room collapsed with them, and the project met austerity while still mostly a construction site. The academic literature that now studies Yachay, including Jose David Gomez-Urrego's 2019 paper in the journal Tapuya on infrastructure, expectations and breakdown at the site, describes the sequence in those terms: momentum, then stagnation, then downscaling, then dispute.

Politics finished what prices started. The project became an object of contention between the Correa administration and its successors, and each subsequent government held the same pen the first one had used. Decree 1060 of 19 May 2020 sent the delivery company into liquidation under President Lenin Moreno, with deadline extensions running into 2021, as Primicias and El Comercio reported. Decree 639 of January 2023 disposed of the wider project under President Lasso. At no point was a court defeat, a market crash or an engineering failure required. An omission would have been enough, and in the end even the omission was formalised on letterhead.

That is what I mean when I say an anchor that depends on one budget line is not an anchor. It has nothing to do with whether the institution is real, and Yachay Tech is real; I have no interest in mocking a working university that survived its own parent project. The problem is failure geometry. An anchor is supposed to be the thing that holds when sentiment turns. An appropriation-funded anchor inverts that: it is precisely the thing that lets go when sentiment turns, because sentiment is the raw material appropriations are made from.

What an anchor with its own balance sheet looks like

Dholera's anchor is a semiconductor fab, and the structural contrast is the entire reason this pairing is worth an essay. Tata Electronics and Taiwan's PSMC received Cabinet approval on 29 February 2024 for India's first major commercial fab inside the SIR, an investment of Rs 91,000 crore, with the Fiscal Support Agreement signed on 5 March 2025 citing Rs 91,526 crore. The plant is designed for up to 50,000 wafers a month on 300 mm silicon, construction is reported past 50 percent civil work by mid 2026 with cleanroom fit-out under way, first silicon is targeted around December 2026, and commercial production is reported for mid 2028. No chip has been produced yet, which is why anchor delivery scores a 2 in Dholera's row, and why I have argued at length that the fab is the ballgame.

The difference from Yachay is not size, although the sizes are not close. The difference is that this anchor is a factory, and a factory is a balance sheet with a roof. It exists to make products that customers pay for, its owners commit capital expecting a return, and once tens of thousands of crores are sunk into concrete and cleanrooms, walking away costs the owner more than finishing. A university funded by decree can be ended by a signature that costs the signer nothing. A half-built fab punishes its own abandonment. The incentive gradient points toward completion, and that gradient is the most valuable property an anchor can have.

The complications I owe you

A comparison this convenient needs stress testing, so let me argue against my own table for three paragraphs. First, Dholera's city shell is budget-line money too. The trunk infrastructure was built with Government of India approved activation packages worth Rs 2,784.83 crore and matching equity of the same amount, recorded in the NICDC Delivery Monitoring Unit report of 30 June 2026, and Gujarat's 2026-27 budget reportedly carries a further Rs 610 crore line for trunk and logistics work. Appropriations laid the roads here exactly as they laid Yachay's. I have mapped the public-money machine separately in how Dholera is funded, and the honest summary is that the shell is sovereign while the anchor is not, which is the reverse of Ecuador's arrangement and the better way around.

Second, corporate anchors can walk, and Dholera carries the scar to prove it. The Vedanta and Foxconn venture, a memorandum valued at roughly USD 19.5 billion in 2022, dissolved when Foxconn withdrew on 10 July 2023, before any committed capital reached the ground. A balance sheet protects a city only after real money is sunk; before that moment, a corporate promise is no sturdier than a decree. Dholera's position in 2026 differs from its position in 2022 because the sinking has visibly happened, and it differs from Yachay's because the party doing the sinking is not the state.

Third, the fab's economics lean on policy. The India Semiconductor Mission covers half of the project's eligible cost, so a government line runs through this anchor too, and pretending otherwise would be the same flattery I have been accusing brochures of. The distinction is between subsidy support and appropriation dependence. Yachay's anchor had no economics without the budget; the fab has customers, products and co-investors with their own money at risk, and the subsidy improves the return on an activity that exists commercially rather than constituting the activity. Beneath both projects sits a legal difference I weigh heavily. Dholera exists under the Gujarat SIR Act of 2009, a statute passed by a legislature, delivered through a company owned 51 percent by the state and 49 percent by the centre. Yachay was born, renamed and buried entirely by executive decree. Statutes are slower to make and slower to unmake, and slowness in both directions is what a thirty-year project needs.

Three questions this comparison hands you

Strip the two cases down and you are left with a portable test for any greenfield anchor anywhere, including every future announcement Dholera itself will generate. Ask, first, whether the anchor earns revenue of its own or is a cost centre attached to a sponsor. Ask, second, who loses committed private capital if it stops; if the honest answer is nobody, or only taxpayers, you are looking at a policy preference with buildings. Ask, third, whether stopping would require an act, meaning a default or a decision someone must own in public, or merely an omission, meaning a line quietly absent from next year's budget.

Yachay fails all three. The university earned nothing, no private party lost committed capital when the city stalled, and the stopping needed no decision at all until decrees tidied up what neglect had already accomplished. The fab, as of August 2026, passes the second test cleanly, passes the third structurally, and passes the first only in the future tense: revenue begins when chips ship, which is reported for mid 2028, with first silicon targeted this December. That future tense is why Dholera's anchor delivery score is a 2 rather than a 4, and why this essay must not be read as reassurance. The architecture is better. The building is not finished. Both sentences are true at once.

One more axis, because the wider sample keeps forcing it on me. An anchor also needs somewhere to draw people from. Ahmedabad sits roughly 100 km from Dholera, cut to a reported 40 to 60 minutes by the expressway opened on 31 March 2026, and that adjacency is why Dholera scores a 3 on proximate metro gravity. Yachay scored a 1, a campus in a rural canton with no large labour market within practical reach. I have set out the full eighteen-city pattern in the lessons essay; the compressed version is that anchor architecture decides whether a city can live, and adjacency decides how fast.

Where this leaves the verdict

Eighth against sixteenth, 3.10 against 1.05, and nearly all of the gap is visible on paper before a single outcome arrives: what the anchor is, who pays for it, what stopping would cost the payer, and what sits next door. That is worth something. It is also worth less than a delivered chip, and the index will not move Dholera's 2 until one exists. Yachay's quieter lesson, the one nobody prices, is that projects can die politely: no scandal, no crash, just a sequence of thin budgets and then a decree acknowledging what had already happened. The counter-programme is public, dated evidence of an anchor that cannot afford to leave, which is exactly what the next two years of fab milestones will either provide or fail to provide.

Funding cadence, the neighbouring failure mode, deserves its own file rather than a paragraph here. Konza in Kenya is the index's study in what interrupted cashflow does even when nobody formally cancels anything, and I read it against Dholera's funded-package model in the Konza comparison. And if this essay nudges you toward or away from a plot, the homework does not change either way: verify the scheme's GUJRERA registration on the state portal where it applies, and satisfy yourself of clear, marketable title inside the SIR boundary before you part with a rupee, because no anchor architecture has ever repaired a defective title.

I will keep the score honest in both directions. If the fab slips past its reported mid 2028 commercial date with silence where explanations should be, the 2 can fall, and the 3.10 with it. If silicon ships anywhere near the stated schedule, Dholera will hold the one credential Yachay never printed: an anchor that pays its own way. Cities are not kept alive by the people who announce them. They are kept alive by whoever cannot afford to let them stop.

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