How Long Planned Cities Actually Take: Navi Mumbai, Shenzhen, Songdo

“When will Dholera be a real city?” is the question underneath almost every other Dholera question. It cannot be answered directly, but it can be bounded — because planned cities have been attempted many times, and their timelines are a matter of record.

Three are worth looking at. One succeeded slowly, one succeeded spectacularly, and one is a cautionary tale that Dholera material never mentions.

Navi Mumbai: the slow success

Conceived in 1971 to relieve pressure on Mumbai, developed by CIDCO on a large greenfield site across the harbour.

The timeline: land acquisition and early infrastructure through the 1970s. Meaningful population growth from the 1980s. Vashi and the early nodes became genuinely established in the 1990s. Today it is a major urban centre of millions — and the airport that was part of the original vision took roughly five decades to arrive.

What it says about Dholera: the closest Indian analogue, and the message is patience. Twenty years from notification to “obviously working.” Early investors did extremely well. Early investors also waited a very long time, and had to survive a couple of decades in which the project looked stalled to anyone not paying close attention.

Navi Mumbai also had an advantage Dholera does not: it was adjacent to one of the world’s densest cities, absorbing demand that already existed. Dholera has to generate its own.

Shenzhen: the outlier

Designated a Special Economic Zone in 1980, when it was a modest settlement of a few hundred thousand. Within three decades, a city of over ten million and one of the world’s major manufacturing and technology centres.

What it says about Dholera: honestly, very little — and it is the comparison Dholera marketing reaches for most often, which should itself tell you something.

Shenzhen’s conditions were extraordinary and largely unrepeatable: adjacency to Hong Kong at the exact moment Hong Kong’s manufacturing was seeking a lower-cost hinterland; a national policy shift of historic magnitude; unrestricted internal migration into a labour market with no competition; and a state capable of acting without the land-acquisition constraints of a democracy.

Shenzhen is the ceiling of what planned urbanisation can achieve. It is not a base case, and treating it as one is how people end up with unrealistic horizons.

Songdo: the one nobody mentions

Songdo International Business District, South Korea. Built on reclaimed land near Incheon from the early 2000s, at a reported cost above $40 billion, as a purpose-built “smart city” with the sensors, the integrated systems, and the vision.

The outcome: it works, but it has consistently underperformed its projections. Occupancy took far longer than planned, the international business tenants who were the entire premise arrived in smaller numbers than projected, and it acquired a reputation — somewhat unfairly, but persistently — as a place that was built and then partly waited for.

What it says about Dholera: this is the failure mode that matters, and it is not fraud. It is a competently executed, genuinely well-built city that does not attract enough people or businesses fast enough to justify its cost. The infrastructure gets delivered exactly as promised, and the demand does not follow on schedule.

Songdo is the scenario Dholera investors should model as their downside — not “the government abandons the project,” but “the government delivers the project and it fills up slower than anyone assumed.”

The pattern across all three

Twenty years is the realistic unit. Navi Mumbai took about two decades to be obviously successful. Shenzhen took roughly three to reach its peak trajectory. Songdo is two decades in and still filling. Nothing here supports a five-year thesis.

Infrastructure precedes demand by a long gap. In every case, roads, utilities, and buildings existed years before the people did. Empty infrastructure is what a planned city looks like mid-build. It is not, by itself, evidence of failure — nor is it evidence of imminent success.

An anchor helps enormously. Shenzhen had Hong Kong’s manufacturing overflow. Navi Mumbai had Mumbai’s congestion. Songdo’s anchor — international businesses relocating to Korea — was the weakest of the three, and Songdo is the weakest outcome.

Dholera’s anchor is the Tata semiconductor fab and the industrial cluster around it. That is a real anchor, and it is more concrete than Songdo’s. It is also narrower than “an adjacent megacity that has run out of room,” which is what the two successes had. We look at what fabs specifically do to a region in what semiconductor cities teach us.

Early investors in the successes did well; early investors in the disappointments waited. In both cases, they waited. The difference in outcome showed up after a decade, not before it — which means the information you would need to distinguish the two is not available at the time you have to decide.

The honest implication for a holding period

If you are considering Dholera, the historical evidence supports:

• A 10–20 year horizon as the realistic frame for the region maturing.
• Position sizing that assumes illiquidity throughout. Not “I could sell if I needed to” — assume you could not, at any price you would accept.
• A downside case of Songdo, not of fraud. The most likely bad outcome is delivered-but-slow, and that produces years of flat prices and thin resale rather than a dramatic loss.
• Treating five-year projections as marketing. No planned city in modern history has done what those projections describe.

None of this argues against investing. Navi Mumbai made patient early investors a great deal of money, and Dholera has advantages Navi Mumbai lacked at the same stage — a named anchor tenant, an operational expressway, and a runway.

It argues for matching your horizon to the asset. If a ten-to-twenty year, illiquid, patient position is something you can genuinely hold, the case deserves serious examination. If it is not, no amount of infrastructure progress makes it the right investment for you — and our checklist deliberately starts with that question, before it gets anywhere near a plot.

Also worth reading: why Dholera has taken so long, which covers the specific mechanics of the delay.