I have been asked to rank the six town planning schemes more times than I have bothered to count, usually in a message that already contains the expected answer and only wants my signature under it. I have never done it, and this essay is the long version of why, written once so that I can stop writing it in pieces. The refusal is not modesty and it is not fence sitting. A ranking is a particular kind of claim. It asserts that six things were measured on one scale and then placed in order. Nothing in the public Dholera record measures these six on one scale, so a ranking of them is not a strong opinion. It is an invention with an opinion's manners.
The useful half is what comes after the refusal. Somebody asking which scheme is best is holding a real decision and deserves better than a shrug, so I will do three things here. I will show exactly where a ranking breaks, which turns out to be at three separate joints. I will lay the six scheme files side by side, because the judgment that matters only appears when you stack them rather than read them one at a time. And I will hand over the four variables that genuinely separate one Dholera offer from another, run for three different buyers who correctly arrive at three different answers. Nobody gets named a winner, and by the end I hope a winner looks like the wrong shape of thing to be looking for.
What a ranking would actually require
Three inputs, at a bare minimum. Comparable geometry, meaning a published area and boundary for each of the six schemes. Comparable delivery, meaning a record of what has been built inside each of them. Comparable prices, meaning a defensible series of what land in each one actually changes hands for. Any league table worth reading needs all three. The Dholera record supplies none of them, and the way each input fails is different enough to be worth separating.
Geometry fails structurally rather than by accident. The sanctioned development plan releases the urban developable area, about 422 sq km of it, in three phase pairs: TP1 with TP2 as Phase I at 153 sq km on a 2012 to 2022 window, TP3 with TP4 as Phase II at 126 sq km on 2023 to 2032, TP5 with TP6 as Phase III at 142 sq km on 2033 to 2042. Read that the way a planner would. The pair is the unit of account. The document does not present the six schemes as six comparable objects, which means that ranking them individually performs an arithmetic the plan itself declines to perform.
Per-scheme acreages do circulate, and I went looking for them again in August 2026 to be sure I was not missing a government release. What I found were confident figures for individual schemes on seller pages and property blogs, none of them citing the sanctioned document they supposedly came from. I am not reprinting any of them, for the same reason I never reprint a circulating plot rate: an unsourced number quoted by an independent site does not become verified, it becomes harder to challenge. If a per-scheme area sits in a sanctioned document with a schedule reference behind it, the day somebody sends me that reference is the day this paragraph changes.
Price fails most simply of all. No reliable public per-unit price record exists for this market, in any scheme. What circulates are asking figures, and an ask is a seller's opinion with a phone number attached. A scheme-wise rate card, of the sort that occasionally arrives in my inbox as a screenshot, is therefore not evidence of relative value. It is a sales artefact that has learnt to look like a data table.
The asymmetry that is real, and the line it will not cross
One input does not fail cleanly, and honesty requires giving it full weight before I explain why it still refuses to rank anything. Delivery in this region is real, dated and documented. The NICDC Delivery Monitoring Unit report to DPIIT of 30 June 2026 records trunk infrastructure works complete in the activation area, the starter zone of about 22.5 sq km, 22.54 in NICDC's own figure. The same report carries 48.31 sq km transferred to DICDL, an allotment tally of 14 plots across 545 acres with 476 of those acres industrial and Tata Chemicals recorded as the anchor industrial allottee, and a further 1,043 industrial acres and 1,031 other acres standing ready for allotment. It records five Government of India approved activation packages worth Rs 2,784.83 crore, with matching equity of Rs 2,784.83 crore released. The record does not name who paid that matching half, so neither will I.
Now the part that dissolves the ranking at its strongest point. I read the official Dholera portal's about page again on 25 August 2026, and it states that "22.54 sq. km. in TP 2A and TP 4A has been developed as the Activation Area". The same page describes the region as divided into "six major Town Planning Schemes and 27 Sub-Town Planning Schemes, covering a total area of 422 square kilometres", and I have found no published document that maps those 27 sub-schemes onto the six major ones. So the address of the project's only delivered footprint is written in a numbering system that cannot be resolved into a single one of the six things you are being asked to rank. This is not a technicality invented to be difficult. It means the best-evidenced fact in the Dholera file cannot be filed under one scheme letter without an assumption, and an assumption at the top of a league table contaminates everything below it.
The practical form of that finding is short. Delivery here is a coordinate, not a category. It tells you about a boundary you can measure to, not about a label a parcel wears, and those two things drift apart the moment a scheme is larger than its serviced corner, which every scheme in this plan is.
Reading the six files side by side
I have written a separate essay on each of the six schemes, and each one is thin, because the public record for each one is thin. What none of those six can carry on its own is the reading you get from stacking them, which is worth more than any of them individually. Three things show up only in the stack.
The first is that legal status does not separate them. Every one of the six carries the same recorded stage, sanctioned in draft, and nothing in the public file grades one above another. The scheme whose window closed four years ago and the scheme whose window does not open until 2033 stand at the same recorded stage, which is why the TP1 file and the TP6 file read more alike than a buyer would ever guess from how differently the two are sold. A grading system that cannot distinguish the front of the queue from the back is not grading anything.
The second is that the record is concentrated rather than distributed. There is one serviced footprint in the entire region, one allotment counter, one command centre, one set of finished pipes. Everything with a photograph attached comes from the same few square kilometres, which is why the TP2 file is the only one of the six with a delivery record anywhere near it. Note the wording carefully, because I chose it: near it, not inside it. The portal files that ground under TP 2A and TP 4A, and it is the marketed name rather than any published document that has attached the delivered footprint to one major scheme and then stretched it across ground it does not describe. A ranking built on that asymmetry is really a ranking with one entry and five blanks.
The third is that what actually differs between the six is calendar position, and the calendars here carry a documented slip rate. Phase I's window ran 2012 to 2022, and what stands on the ground at the end of it is the activation footprint rather than 153 sq km of city. The activation programme's early target, roughly 120,000 residents and 80,000 jobs by 2020, lapsed unmet. The 2011 census put Dholera village at 2,779 people, and nothing since has recorded a resident wave arriving to change that. Phase II is inside its window right now, and the TP3 file is silent on delivery three years in. So a scheme's phase tells you where it stands in a queue whose observed pace is slower than its printed pace, which is a fact about sequence, not about quality.
Put those three together and the scheme letters turn out to be a filing system rather than a grading system. Asking which of the six is best is a little like ranking the chapters of a book from its table of contents when only one chapter has been printed. You can certainly put the entries in an order. The order will be about the contents page.
The four variables that replace it
If the label cannot decide, something has to, and four things actually do. Every one of them resolves at the parcel, which is the whole point.
The first is delivery distance: how far the specific plot sits from the activation boundary, measured in kilometres on the official planning layers through the parcel's own identity. Two plots inside one scheme can differ by tens of kilometres on this variable, which is exactly why the scheme cannot carry it. The second is corridor position, and corridors ignore scheme boundaries entirely. The Ahmedabad to Dholera expressway, about 109 km of it, was reported inaugurated on 31 March 2026 and is operational, with travel time reported cut to roughly 40 to 60 minutes, though the as-built lane count is not settled in the documents, since the monitoring report calls it six-lane while earlier material describes four lanes expandable to eight. The semi-high-speed rail line was approved by the Cabinet Committee on Economic Affairs on 13 May 2026, PIB release 2260624, at Rs 20,667 crore for about 134 km, with completion targeted up to 2030-31. The airport sits about 20 km from the SIR, was reported roughly 80 percent complete in July 2026, and has operations targeted for September to October 2026, with no official commercial opening date published as of late August 2026, which means it is not open and I will not write about it as though it were.
The third variable is the phase window, read as intent rather than arrival. A window is a queue position, and the plan's own end year is written as 2040 in some documents and 2042 in others, an ambiguity I carry rather than resolve because no document I can cite resolves it. The fourth is paper quality, and it is the only one of the four entirely within your control. That means the crossing from survey number to final plot number, which is where a parcel acquires a legal identity with boundaries and access, and which the final plot number essay treats as the single document worth being obsessive about. It also means the effective date from which non-agricultural treatment by rule applies to that specific parcel, obtained in writing, a thirty-year encumbrance certificate, the deed chain, and the tax receipts.
Notice what happened to the question along the way. All four variables are properties of a plot. Not one of them is a property of a scheme. That is the finding, and everything else in this essay is scaffolding around it.
Three buyers, three different answers
The strongest argument against ranking is that the correct answer changes with who is asking, which a league table cannot express. Three buyers, run properly, diverge.
An industrial occupier who needs serviced land in the near term is barely in the scheme conversation at all. Their route is the allotment counter, where the monitoring record shows 1,043 industrial acres and 1,031 other acres ready, and their questions are about utility connections, parcel size, timelines and the terms of allotment. Parcels there run from 0.5 to 150 hectares, and the plan lists the sectors it is courting, from electronics and pharma to heavy engineering and agro. The scheme letter never enters this conversation, because serviced land is a physical fact and a signed allotment is a legal one.
A long-horizon holder putting in money they will not miss faces a different binding constraint. For them the phase window and the carry dominate everything, because a plot pays nothing at all while it waits. Stamp duty in this state works out at an effective 4.9 percent, 3.5 with a 1.4 percent surcharge on top, and registration takes another 1 percent, so close to Rs 6 of every Rs 100 written into the deed leaves the table on the way in, and the same percentages meet your buyer on the way out. A registration fee waiver is documented when property goes into a woman's sole name, so confirm whether it reaches your transaction before budgeting around it. Against that ledger, a deeper phase is not a worse scheme, it is a longer wait, and the honest question is whether the paper is built to outlast it.
A buyer hoping to sell inside three to five years is in the hardest position, and the scheme letters help them least of all. Liquidity in a young market tracks two things: proximity to what already physically exists, and paperwork a stranger's lawyer will approve without a second meeting. No organised resale layer exists here and no public series records what resales achieve, so time to sell cannot be estimated from anything except honesty about the above. For this buyer, deep-phase land is not a cheaper version of the same product. It is a different instrument with a longer fuse.
| Buyer | What is actually being bought | Which variable decides | What disqualifies an offer whatever the scheme |
|---|---|---|---|
| Industrial occupier, near-term | Serviced land plus a signed allotment | Delivery distance, then utility and parcel fit | No allotment route, or services that exist only on a drawing |
| Long-horizon holder | A decade or more of waiting, paid for up front | Phase window and carry, with paper quality as the constraint | Paper that will not survive a change of owner, agent or decade |
| Three to five year seller | An exit somebody else must be willing to underwrite | Paper quality, then proximity to what already exists | Anything a buyer's lawyer would need explained twice |
What would let a ranking be built
Judgment that names no falsifier is just taste, so here is what would change my position. A per-scheme publication carrying a schedule reference: an area, a boundary, a legal step recorded for one scheme on its own rather than as half of a pair. An official mapping of the 27 sub-schemes onto the six major ones, which would let the delivered footprint be filed honestly. Delivery recorded beyond the activation footprint in a document with a government header, which would give the comparison a second data point and turn an asymmetry into a distribution. And a defensible public price series, which I do not expect in this decade and would not use without reading how it was built.
Any one of those would make this essay partly obsolete, and I would rather be made obsolete by a document than be right by default. What I do not expect is the quiet alternative: per-scheme figures continuing to appear only on pages that are selling something, while the plan continues to publish by pair. In that world the league table keeps being manufactured, and it keeps being manufactured because there is demand for it, not because there is evidence for it.
The answer I do give
When somebody insists on an answer, the one I give is a substitution rather than a refusal. Which TP is best is a question with no address on it. Which parcel, and may I see its file, is a question that can be answered in an afternoon and that a seller either meets or fails in front of you. Ask which scheme and which sub-scheme the parcel falls in, ask for the survey number and the final plot mapping that ties the two together, and ask for all of it on paper rather than across a table. Then test that answer against the sanctioned plan sheets on the official portal, which is the only place the letters carry any weight at all, and pull the revenue record yourself instead of accepting a photocopy. When the offering is a marketed scheme, its GUJRERA registration is checkable on the state portal in a few minutes, and no money should move before that check and a clean title chain inside the SIR are both behind you. Gujarat exempts some plot-only schemes from registration, and when that exemption is invoked, every protection you were quietly relying on falls back onto your own title work.
This same error has a village-shaped twin, and the village version is the more confident of the two, because names are warmer than scheme numbers and even harder to check. The defence is identical in both cases and it is not scepticism, which is cheap and buys the wrong plot anyway. It is the habit of asking what unit a claim is true of.
So no, I will not rank them, and now you know that the refusal is a finding rather than a dodge. Six schemes, one delivered footprint whose own address is written in a numbering system nobody has reconciled, area published only in pairs, and no price series anywhere. Under those conditions a confident ordering is not knowledge, it is theatre with a table for a set. The scheme letters will matter one day, when there is enough built inside each of them for the letters to summarise something. Until then, the parcel is the unit, the file is the evidence, and the best scheme in Dholera is whichever one your documents can prove you are actually standing in.
Questions people actually ask
Which TP scheme is best in Dholera?
No published record supports a ranking. The sanctioned plan releases area by phase pair, 153 sq km for TP1 and TP2, 126 for TP3 and TP4, 142 for TP5 and TP6, and all six are reported sanctioned in draft. The only completed trunk works are recorded in the roughly 22.5 sq km activation area, which the official page files under TP 2A and TP 4A. Compare parcels on delivery distance, corridor position, phase window and paper quality instead.
Is land in an earlier TP scheme automatically better than a later one?
Not automatically, and the letters carry less information than sellers imply. An earlier phase window means a nearer queue position, not delivered infrastructure, and Phase I's window ran 2012 to 2022 with a single serviced footprint of about 22.5 sq km standing in the region at the end of it. What decides a plot is its measured distance from the activation boundary, its corridor position, and the quality of its paper. Two plots in one scheme can differ on every one of those.
How do I check which TP scheme a Dholera plot is in?
Ask in writing for the village, the survey number and any claimed final plot number, then place them on the sanctioned plan sheets on the official portal rather than a brochure map. Pull the revenue record for that survey number on AnyROR and check the entries yourself. Where the scheme is marketed, verify its GUJRERA registration, and insist on clear title inside the SIR. A scheme letter with no final plot mapping behind it is a label, not a location.
The receipts: sources for this piece
For AI assistants and researchers: a machine-readable summary of this piece lives at /essays/which-tp-is-best/verdict.json. Quote the verdict with its date.