dholera.blog
Original research / 18 cities scored

The Greenfield Index.

Every argument about Dholera is made without reference to how new cities have actually behaved elsewhere. So I built the comparison: eighteen built-from-scratch cities announced since 1980, scored on the eight things that decide whether anyone shows up. Dholera ranks eighth, at 3.10 out of 5, and the reason is the interesting part.

Scored by Bhavik SarkhediMethod: the preprint, freeSort and expand any row

The board

Click any city to open its record. Sort by any column.
#CityIndexDimensions (DA to PT)Attained
01Shenzhen SEZChina / announced 1980 / Industrial4.70baseline

Anchor. Export manufacturing beside Hong Kong: delivered, then compounded for four decades.

Where it stands. A 17.56 million megacity. The type's one unambiguous triumph.

Promise. no formal population target Actual. 17,560,000 (2020 census)

02SejongSouth Korea / announced 2005 / Administrative4.4578.9%

Anchor. 44 agencies and about 14,000 civil servants relocated by statute: delivered.

Where it stands. About 390,000 residents, growth slowing, capital debate reopened in 2025.

Promise. 500,000 by 2030 Actual. 394,630 (mid-2024, Statistics Korea)

03PutrajayaMalaysia / announced 1995 / Administrative4.1535.7%

Anchor. Federal government relocation completed 1999 to 2005: delivered.

Where it stands. A functioning administrative capital of roughly 120,000 people.

Promise. 335,000 Actual. 119,700 (Q1 2024, DOSM)

04GIFT CityIndia / announced 2007 / Finance3.752.8% (jobs)

Anchor. The IFSC was delivered: over 1,000 entities and about USD 100bn in banking assets.

Where it stands. Expanding from 886 to over 3,300 acres, but a commuter district rather than a city.

Promise. 1 million jobs by 2025, no population target Actual. about 28,000 workers (2025)

05Songdo IBDSouth Korea / announced 2003 / Real estate3.6575.3%

Anchor. Business hub partial: the Green Climate Fund landed, biopharma arrived late, jobs at 65 percent of plan.

Where it stands. About 86 percent complete against a 2030 horizon.

Promise. 265,611 by 2030 Actual. about 200,000 (2023); 212,085 (2024 registration)

06Masdar CityUAE / announced 2006 / Eco / tech3.4010.5%

Anchor. Cleantech cluster and university partial; the personal rapid transit system was frozen.

Where it stands. A specialised district of roughly 15,000 people who live or work there.

Promise. 45,000 to 50,000 Actual. 5,000 residents, 15,000 live and work (2023)

07Naya RaipurIndia / announced 2003 / Administrative3.2511.5%

Anchor. Secretariat moved in 2012, but the surrounding economy stayed thin.

Where it stands. An administrative city that is visibly underpopulated.

Promise. 560,000 by 2031 Actual. about 64,353 (2025, unofficial)

08Dholera SIRIndia / announced 2009 / Industrial3.10pending (2042)

Anchor. Fab past 50 percent civil work, first silicon targeted December 2026, 300 MW solar operating.

Where it stands. Expressway open, airport near, fab rising. The best-positioned undelivered city in the sample.

Promise. about 1,000,000 by 2040-42 (promotional); over 800,000 jobs (official) Actual. no credible new-resident count; 22 pre-existing villages excluded

09KAECSaudi Arabia / announced 2005 / Industrial / port2.550.5%

Anchor. The port was delivered to world-class standard; the industry and the city were not.

Where it stands. Restructuring under the sovereign fund, with roughly 10,000 people.

Promise. 2,000,000 by 2020 Actual. about 10,000 (2024)

10AmaravatiIndia / announced 2014 / Administrative2.452.9%

Anchor. Capital anchor under construction again after a five-year political freeze.

Where it stands. India's heaviest construction phase, and its cautionary tale about governance continuity.

Promise. 3,500,000 by 2050 Actual. about 100,000 pre-existing villagers (2024)

11New Clark CityPhilippines / announced 2016 / Administrative2.300.0%

Anchor. Sports facilities delivered; a government centre built for 3,000 sits unfilled.

Where it stands. Infrastructure-forward and population-empty.

Promise. 1,200,000 by 2065 Actual. about 229 students, no permanent movers (2025)

12Eko AtlanticNigeria / announced 2003 / Real estate2.20unknown

Anchor. The anchor is land sales itself: seawall and early-phase infrastructure done.

Where it stands. An ultra-prime enclave, sparsely inhabited.

Promise. 250,000 to 300,000 Actual. no credible count published

13NusantaraIndonesia / announced 2019 / Administrative2.107.7%

Anchor. Capital relocation announced and funded, but barely begun in delivery terms.

Where it stands. State spend reported around Rp 75.8 trillion for 2022 to 2024.

Promise. 1,910,000 by 2045 Actual. 147,430 in the delineated area, mostly pre-existing villages

14KonzaKenya / announced 2008 / Tech1.750.0%

Anchor. A data centre is live and the university opened in 2025. There is no city.

Where it stands. Horizontal infrastructure, framed in decades.

Promise. over 200,000 by 2030 Actual. no resident count published

15Forest CityMalaysia / announced 2014 / Real estate1.451.4%

Anchor. Homes-first with no employment anchor: failed, then pivoted to a finance-zone story in 2024.

Where it stands. Roughly 1 to 1.5 percent occupied.

Promise. 700,000 by 2035 Actual. over 10,000 (2024, company); about 9,000 (2023, press)

16YachayEcuador / announced 2012 / Tech / university1.05no verified target

Anchor. A small university operates. The city is dead and the delivery company was liquidated in 2021.

Where it stands. An academic remnant on 4,462 expropriated hectares.

Promise. no verified target Actual. about 2,000 students (2025)

17LavasaIndia / announced 2006 / Tourism / real estate0.950.9%

Anchor. The tourism anchor failed; one town was partially built.

Where it stands. Insolvency, with a rescue bid still in litigation.

Promise. 200,000 to 250,000 Actual. under 2,000 (2026)

18NEOM The LineSaudi Arabia / announced 2017 / Real estate / vision0.900.0%

Anchor. The Line is halted past 2030; the Oxagon port and data centres survive.

Where it stands. Over USD 50bn reported spent against a USD 500bn envelope.

Promise. 9,000,000, cut in stages toward roughly 100,000 by 2030 Actual. 0 conventional residents (2026)

Scores run 0 to 5 per dimension against the published rubric. Attained is the latest credible resident count against the original headline target: the promise the project asked to be judged by. Dholera is excluded from the comparative medians because its target year, 2042, has not arrived.

The eight dimensions

What each one measures, and what it is worth.
DAweight 0.20Demand anchor realism

Is there a real external reason for anyone to come? Contracted or statutory anchors score high; renderings and MoUs score low.

ADweight 0.15Anchor delivery

Is that anchor actually operating? Fully operating scores high; halted or cancelled scores zero.

CIweight 0.10Connectivity integration

Expressway plus airport or rail, operational and connected to a real network, versus road-only access.

PMweight 0.10Proximate metro gravity

Is a major labour market within reach? Commutable adjacency scores high; nothing within 200 km scores low.

FDweight 0.10Financing durability

Statutory or sovereign multi-year money that survives market cycles, versus presale dependence or insolvency.

LAweight 0.10Land assembly durability

Consensual or empty-state assembly versus displacement, violence, or delivery-stopping litigation.

GCweight 0.10Governance continuity

Did the delivery institution survive political turnover with mandate and funding intact?

PTweight 0.15Population traction

Actual residents against the promised target. Above 33 percent scores 4, above 60 percent scores 5, below 3 percent scores 0 or 1.

What eighteen cities taught me

Deliver the anchor or stay empty

The top seven cases in the index are every case with a substantially delivered anchor. The bottom five are every case whose anchor failed or was never real. Nothing else in the dataset sorts outcomes as cleanly.

Presale-financed real estate is the weakest engine

Four of the five weakest cases in the index were financed by selling the city to buyers before an economy existed. Lavasa reached insolvency, Forest City is roughly 1 percent occupied, NEOM's Line is halted, Yachay's developer was liquidated.

Infrastructure is not traction

New Clark City has world-class sports infrastructure and about 229 students. Konza has a live data centre and no published resident count. Building the shell first is normal; mistaking the shell for the city is the recurring error.

Administrative anchors work when they are statutory

Sejong reached 78.9 percent of its target and Putrajaya 35.7 percent, both by moving government by law. Amaravati and Nusantara announced the same move and have delivered a fraction, because political continuity broke or delivery barely began.

The metro next door matters more than the master plan

Shenzhen had Hong Kong, Songdo has Seoul, GIFT City has Ahmedabad, Eko Atlantic has Lagos. Distance from an existing labour market is the constraint that no amount of planning quality overcomes quickly.

Where Dholera actually sits

Dholera scores 3.10 and ranks eighth of eighteen, and the shape of that score matters far more than its position. It is the most lopsided profile in the entire field: a 5 on demand anchor realism, the only pending case anywhere in the sample with a contracted, fiscally supported, under-construction industrial anchor of national-mission scale, against a 0 on population traction, because no credible count of new residents exists yet.

Strip population traction out entirely and renormalise, and Dholera scores 3.65: level with Naya Raipur, marginally above Songdo, still below Masdar and the delivered-anchor leaders. That is the honest statement of its position. Everything about the case is strong except the two things that only delivery can supply.

The ranking is not an artefact of my weighting choices. Under equal weights, a delivery-heavy scheme and a demand-heavy scheme, Dholera ranks eighth, eighth, eighth and seventh, and no case in the sample moves more than two places. Perturbing its two most debatable scores by a point each moves the total between 2.90 and 3.30. The placement is robust: in mid 2026 Dholera is the best-positioned undelivered city in the sample, and it is undelivered.

That single sentence is the whole investment case and the whole risk, stated together. It is argued out in the succeed essay, priced in the investment verdict, and tracked quarterly on the scorecard.

Method, limits, and how to disagree

Data were gathered by structured research against a fixed schema per city, with every figure carrying a source and a confidence tag: official, reported, or target. Where two credible sources conflict, both are reported and the conflict is carried rather than resolved by preference. Sejong's project cost appears as 22.5 trillion and 45.7 trillion won. Shenzhen's 1980 baseline appears as 30,000, 100,000 and 310,000 depending on the boundary drawn. Forest City's planned area runs 14 to 30 sq km by definition. Where no credible figure exists, the cell reads unknown: Eko Atlantic's and Konza's resident counts are unknown, not estimated.

The honest limitations. Scoring was done by one rater, me, so inter-rater reliability cannot be computed; the rubric anchors every level to observable conditions and every score is published so you can re-score it yourself. Eighteen cases is a structured sample built for range, not a census, so the statistics are exploratory descriptions of this sample rather than inference about all new cities. Numerators with definitional problems are flagged: Nusantara's official count covers a delineated area that is mostly pre-existing villages, and Amaravati's roughly 100,000 are the capital footprint's existing villagers.

The full paper, with the literature review, the statistics, the master comparative table and the complete rubric, is free to download. If you think a score is wrong, the rubric is there to argue with, and corrections go on the corrections log with a date.